guidelines for management of acute cervical spinal injuries

INTRODUCTION

Guidelines for Management of Acute Cervical Spinal Injuries

Spinal cord injuries occur approximately 14,000 times peryear in North America, and most involve the cervicalspine region. Most, although not all, of these injuries will

include cervical spine fracture-dislocations. Patients who sus-tain cervical spinal cord injuries usually have lasting, oftendevastating, neurological deficits and disability. In addition,tens of thousands of patients each year will sustain traumaticcervical spine injuries without spinal cord injury. The man-agement of these patients and their injuries, spinal cord, andvertebral column has not been standardized nor is it consis-tent within a single institution, from one center to another, oramong centers within a geographic region. Treatment strate-gies are usually dependent on the experiences of institutionsor individual health care providers, on physician training, andon the resources available at the treatment facility. Manage-ment can affect outcome in these patients; therefore, cliniciansworldwide strive to provide the “best and most timely care.”Often, we may not be fully aware of what the “best care” may be,or whether “timeliness” matters. In many circumstances, “bestcare” likely encompasses a variety of treatment strategies, allwith acceptable success rates and reasonable inherent risks.

The Section on Disorders of the Spine and PeripheralNerves of the American Association of Neurological Surgeonsand the Congress of Neurological Surgeons has long beeninterested in seeking answers to the key management issuesassociated with acute spine and spinal cord injuries. Identifi-cation of “best care” strategies is desired for all aspects of thecare of patients with acute cervical injury. Such strategies caninclude care and transport of the patient before admission tothe facility, neurological and radiographic assessment, medi-

cal management of spinal cord injury, closed reduction ofcervical fracture-dislocations, and specific treatment options,both operative and nonoperative, for each specific cervicalinjury type known to occur from the occiput through the firstthoracic level. The leadership of the Spine Section charged uswith the task of generating guidelines on the management ofpatients with acute cervical spine and cervical spinal cordinjuries. We began, in May 2000, by identifying 22 topic areasand generating questions around which recommendationswould be formed. We followed a meticulous process foundedin evidence-based medicine. Published scientific evidencewas searched for and relied on rather than expert opinion ortraditional practices. In the course of developing these recom-mendations, new methodology had to be created for classify-ing evidence on clinical assessment, resulting in a substantialcontribution to the guidelines literature as a whole. We firstconvened in September 2000; 1 year later, the task wascompleted.

The authors and the Joint Section on Disorders of the Spineand Peripheral Nerves hope that these guidelines will definethe variety of assessment and treatment options available to aclinician in the management of an individual patient, providedirection within the broad scope of clinical practice derivedfrom medical evidence, highlight what is known regardingspecific issues, and, importantly, define what is not knownand stimulate additional research.

Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.

Drawing by Leonardo da Vinci of the human cranium and spinal canal. Courtesy, Dr. Edwin Todd, Pasadena, California. S1

FOREWORD

On behalf of the American Association of NeurologicalSurgeons/Congress of Neurological Surgeons JointSection on Disorders of the Spine and Peripheral

Nerves, it is my great privilege to introduce these Guidelinesfor the Management of Acute Cervical Spine and Spinal CordInjuries. These guidelines represent the initial installment of amore comprehensive guidelines initiative from the Joint Sec-tion on behalf of all practicing neurosurgeons and their pa-tients. The Section is grateful to the small working group whodevoted considerable time and effort to the generation of thisoutstanding document. We would like to formally recognizethe Joint Section on Trauma for their important collaborationon this project. The Section would also like to acknowledgeand thank the parent organizations, the American Associationof Neurological Surgeons and the Congress of NeurologicalSurgeons, for their guidance of and support for this project,most notably through the efforts of the American Associationof Neurological Surgeons/ Congress of Neurological Sur-geons Guidelines Committee.

The Section is also deeply indebted to Michael Apuzzo andthe staff of Neurosurgery for their advice and editorial assis-tance in preparing this document for publication. The appli-cation of Neurosurgery’s rigorous peer-reviewed editorial pro-cess has clearly enhanced the quality, balance, and stature ofthis document. Perhaps most importantly, Neurosurgery hasprovided an extraordinary vehicle for the widespread dissem-ination and ultimate incorporation of these guidelines to im-prove the care and enhance the outcomes of patients withtraumatic cervical spine and spinal cord injuries.

One of the truly important functions of organized neuro-surgery is the generation of evidenced-based clinical practiceguidelines. Properly developed, such guidelines can answerimportant questions, resolve uncertainty, identify areas ofdeficient knowledge and opportunities for future scientificinvestigation, standardize treatment, and improve the qualityof care and the outcomes for patients. The now widely dis-seminated head trauma guidelines, for example, have clearlymade a difference in the outcomes of patients with severehead injury.

Guidelines development is a highly structured process withrigorous methodological criteria and exacting standards. It isa time-, labor-, and resource-intensive process that has servedas a significant obstacle to more widespread guidelines de-velopment throughout neurosurgery. In the past, clinicalpractice guidelines have been developed by publicly sup-ported epidemiologists and methodologists who understoodstudy design, data analysis, and the guidelines process, butnot the disease. This absence of context and clinical perspec-tive significantly limited the value and relevance of theirresults. Alternatively, clinician-generated guidelines oftentook the form of methodologically flawed consensus panelsand expert opinion, also of limited value.

The Joint Spine Section recognized the importance ofevidence-based clinical practice guidelines and the challenges

of their development. The appropriate clinical expertise, strictadherence to established methodological standards for guide-lines development, and considerable resource investment forthe development, dissemination, and maintenance of theguidelines documents were deemed crucial to our guidelinesinitiative. Cervical spine and spinal cord injury was chosen asthe initial guidelines topic because of the personal, social, andeconomic devastation of these injuries, their complex nature,and the high level of uncertainty, as reflected in wide practicevariations, as to the value and indications for many of theaspects of evaluation and treatment.

The clinical practice guidelines contained in this supple-ment to Neurosurgery represent a remarkable effort. Theyaddress the key issues related to the evaluation and manage-ment of these complex conditions that are relevant to thetreating physician. In every chapter, the pertinent issues aresuccinctly stated, the published data are comprehensivelypresented in the evidentiary tables, and the evidence is thor-oughly discussed and critically evaluated throughout the text.The linkage between the quality of the evidence and thestrength of the recommendations was not a “black box” pro-cess but an open, deliberative exercise by skilled expertsguided by a rigorous set of standards.

Despite the strength and potential value of this document,it is important to acknowledge the inherent limitations ofclinical practice guidelines. This, or any other, evidence-basedclinical practice guidelines document does not represent thedefinitive source of knowledge on the stated topic. Rather, itrepresents recommendations of varying strength and cer-tainty based on an analysis of the best available publisheddata. These data, however, are often conflicting, flawed, orincomplete, and there are unavoidable elements of potentialbias from subjectivity, perspective, and experience of the in-dividuals and group involved in the analysis and interpreta-tion of these data. In essence, proof is a relative term based onthe interpretation of evidence. Furthermore, it is subject todifferent standards. A relevant example comes from the fieldof jurisprudence, where the standard of proof (i.e., guilt) forcriminal trials is “beyond a reasonable doubt,” whereas thestandard for civil courts must simply reflect “a preponderanceof evidence” or “more likely than not.” These different stan-dards evolved because of the perceived different conse-quences of a wrongful verdict. Moreover, as Stephen Haineslikes to note, the verdict “not guilty” does not mean innocent;it merely says not proved. Such are the vagaries associatedwith the interpretation of even scientific evidence. Principledpeople can look at the same evidence and come to differentconclusions subject to their own personal perspective, expe-rience, and stake in the result.

Nevertheless, the Joint Section and Guidelines Develop-ment Group went to great lengths to identify and avoid—or atleast minimize—these potential problems. The working groupadopted the most widely recognized and rigorous standardsfor guidelines development. A diverse panel of experts with

S-ii Neurosurgery, Vol. 50, No. 3, March 2002 Supplement

expertise in spine, trauma, and epidemiology brought rele-vant clinical, scientific, and methodological competence toenhance both the analytical and the deliberative aspects of thisprocess. Periodic outside reviews were routinely obtained fortopics or areas of contention or uncertainty to add additionalperspective and balance. Above all, the process was account-able and transparent at every stage.

Ultimately, we offer these guidelines as a living documentto those professionals who treat patients with traumatic spinalinjury. We hope each practitioner will critically evaluate theseguidelines and come to his or her own conclusion on how,whether, and when to implement its recommendations. Itmay be used either as a reference or as a basis for standard-ized protocols of evaluation and management of the patients

with traumatic spinal injury. For clinical and basic scienceresearchers, we hope that it will identify and catalyze scien-tific investigation in areas of deficient knowledge. As a Sec-tion, we stand firmly behind this important document andwill continuously update the recommendations as newknowledge and understanding is developed. We sincerelybelieve that these guidelines can improve the care and en-hance the outcomes of patients with traumatic injuries to thecervical spine and spinal cord.

Paul C. McCormickChair, Joint Section on Disorders of the

Spine and Peripheral NervesNew York, New York

Foreword

METHODOLOGY

Methodology of Guideline Development

The evolution of medical evidence has occurred rapidlyduring the past 50 years. From initial reports, which areanecdotal in nature, to large-scale randomized con-

trolled trials, medical evidence is variable. From the evidence,and influenced by personal experience, clinicians choose waysto manage disease. The medical specialties have pioneered theuse of evidence from experimental trials to support clinicalpractice decisions. The surgical specialties have lagged behindin the development of large-scale studies of surgical proce-dures and perioperative management. However, the high costof medical care, together with variations in practice fromregion to region, has given rise to an interest in developingstrategies for linking practice to underlying evidence. Duringthis process, it has become clear that the variability of theevidence must somehow be reflected in any recommenda-tions derived from it.

In the 1980s, criteria were developed for use in selectingevidence for developing treatment recommendations. In aformal document, Clinical Practice Guidelines: Directions for aNew Program (4), the Institute of Medicine addressed suchissues as definition of terms, specification of key attributes ofgood guidelines, and certain aspects of planning for imple-mentation and evaluation (4). The key intent was to promotestandardization and consistency in guideline development.Several key concepts were espoused:

1. A thorough review of the scientific literature should pre-cede the development of guidelines.

2. The available scientific literature should be searched byusing appropriate and comprehensive search terminology.

3. The evidence should be evaluated and weighted to reflectthe scientific validity of the methodology used to generatethe evidence.

4. There should be a link between the available evidence andthe recommendations, with the strength of the recommen-dations reflecting the strength of the evidence, in turnreflecting the scientific certainty (or lack thereof) of theevidence.

5. Empirical evidence should take precedence over expertjudgment in the development of guidelines.

6. Expert judgment should be used to evaluate the quality ofthe literature and to formulate guidelines when the evi-dence is weak or nonexistent.

7. Guideline development should be a multidisciplinary pro-cess, involving key groups affected by the recommendations.

To develop the guidelines for managing acute cervicalspine and spinal cord injuries, we used the evidence-basedapproach reflected in those concepts, rather than a consensus-based approach with input from experts based on the litera-ture and their personal experience. We used a strict process ofliterature review, ranking the published articles by strength of

study design. Every effort was made to maintain objectivityand avoid the influence of personal or professional biasthrough a methodology defined in advance. The methodologywe chose follows the recommendations of the Institute ofMedicine Committee to Advise the Public Health Service onClinical Practice Guidelines (4), as outlined in detail below.

METHODOLOGY FOR DEVELOPING GUIDELINES

Literature search

The first step was to undertake an extensive literaturesearch for each clinical question addressed. The computerizeddatabase of the National Library of Medicine was searched foravailable English-language literature on human studies pub-lished during the past 25 years. The search terms reflected theclinical question in as much detail as possible, as described ineach chapter. Abstracts were reviewed, and clearly relevant ar-ticles were selected for evaluation. Each article was evaluated bystudy type (e.g., therapy, diagnosis, or clinical assessment).

Evaluating the strength of the therapy literature

Evidence can be generated by several different study de-signs. The strongest study protocol, when well designed andexecuted, is the randomized controlled trial. The prospectiv-ity, presence of contemporaneous comparison groups, andadherence to strict protocols reduce sources of systematicerror (or bias). The randomization process reduces the influ-ence of unknown aspects of the patient population that mightaffect the outcome (random error).

The next strongest study designs are the nonrandomizedcohort study and the case-control study. These designs alsocompare groups that receive specific treatments, but in anonrandomized fashion. In the nonrandomized cohort study,an established protocol for patient treatment is followed andgroups are compared in a prospective manner, provided thattheir allocation to the treatment group is not determined bycharacteristics that would not allow them to receive either ofthe treatments being studied. The patients in each group havethe disorder of interest (e.g., spinal cord injury) and receivedifferent interventions, and the differences in outcome arethen studied. In the case-control study, the patients aregrouped by outcome (e.g., functional ability), and their treat-ment (e.g., surgery versus no surgery) is evaluated for arelationship. These studies are more subject to systematic andrandom error and are therefore less compelling than the ran-domized controlled trial. However, a randomized controlledtrial with significant design flaws that threaten its validitymay be classified as a weaker study. Least strong evidence isgenerated by published series of patients with the same orsimilar disorder followed for outcome, but not compared as totreatment. Also in this category are the case report, expert

S2 Neurosurgery, Vol. 50, No. 3, March 2002 Supplement

opinion, and the randomized controlled trial so flawed thatthe conclusions are uncertain. These statements regardingstudy strength refer to studies on treatment, but patient man-agement also includes diagnosis and clinical assessment.These aspects of patient care require clinical studies that aredifferent in design, generating evidence about choices of di-agnostic tests and clinical measurement.

Evaluating the strength of the diagnostic test literature

To be useful, diagnostic tests must be reliable and valid.Reliability refers to the test’s stability in repeated use, underthe same circumstances. Validity describes the extent to whichthe test reflects the “true” state of affairs, as measured bysome “gold standard” reference test. Accuracy reflects thetest’s ability to determine which patient does and which doesnot have the suspected disorder. Overall, the test must beaccurate in picking out the true-positives and true-negatives,with the lowest possible false-positive and false-negativerates. The attributes of diagnostic tests are represented bysensitivity, specificity, positive predictive value, and negativepredictive value. These values may be calculated by using aBayesian 2 � 2 table (Table 1).

From Table 1, the components of accuracy can be expressedand calculated as shown in Table 2. A characteristic of diag-nostic tests is that these attributes do not always rise together,but in general, the accuracy values should be higher than 70%for the test to be considered useful. The issue of reliability ofthe test will be discussed below.

Evaluating the strength of the patientassessment literature

There are two points to consider when patient assessmentis key in the patient management paradigm: the initial assess-ment (e.g., the patient’s condition in the trauma room) and theultimate assessment, or outcome. All assessment tools,whether they are radiographic, laboratory, or clinical, requirethat the measurement be reliable. In assessments carried outby mechanical or electronic equipment, reliability is ensuredby calibrating the devices regularly. In assessments carriedout by observers, reliability is ensured by verifying agreement

TABLE 1. Bayesian Table for Calculating the Attributes of Diagnostic Test Literature

Test Resulta“Gold Standard”

Patient Has Injury Patient Has No Injury

Positive:appears to have injury

True-positive(a)

False-positive(b)

(a � b)

Negative:appears to have no injury

False-negative(c)

True-negative(d)

(c � d)

(a � c) (b � d) (a � b � c � d)a Per cervical spine x-ray.

TABLE 2. Calculation for Level of Accuracy of aDiagnostic Test

Test Attribute FormulaProbabilityStatement

Sensitivity aa � c

If a patient has apositive x-ray,how likely ishe/she to havethe injury ofinterest?

Specificity db � d

If a patient has anegative x-ray,how likely ishe/she to nothave the injuryof interest?

Positive predictive value aa � b

If a patient hasan injury, howlikely is he/sheto have apositive test?

Negative predictive value dc � d

If a patient doesnot have aninjury, howlikely is he/sheto have anegative test?

Accuracy a � da � b � c � d

Methodology of Guideline Development

S3Neurosurgery, Vol. 50, No. 3, March 2002 Supplement

between different observers of the same assessment and bythe same observer at different times. Because a certain amountof agreement between observers or observations is expectedto occur by chance alone, a statistic has been developed tomeasure agreement between observations or observers be-yond chance: the index of concordance, or the kappa (�)statistic (3). Once again, the Bayesian 2 � 2 table can be usedto understand and to calculate � (Table 3).

From these numbers, � is calculated by:

� �N�a � d� � �n1 f1 � n2 f2�

N2 � �n1 f1 � n2 f2�

or

� �2�ad � bc�

�n1 f2 � n2 f1�

Values of � indicate the strength of agreement betweenobservers or observations, as shown in Table 4 (5). We ratedeach article on clinical assessment for its adherence to therules of reliability and noted the exact � value. These valueswere linked to the strength of recommendations, as describedbelow.

Linking evidence to guidelines

The concept of linking evidence to recommendations hasbeen further formalized by the American Medical Associationand many specialty societies, including the Congress of Neu-

rological Surgeons, the American Association of NeurologicalSurgeons, and the American Academy of Neurology (1, 2, 6,7). This formalization involves designating a specific relation-ship between the strength of evidence and the strength of theresulting recommendations to avoid ambiguity. In the para-digm for therapeutic maneuvers, evidence derived from thestrongest clinical studies (well-designed randomized con-trolled trials) generates Class I evidence. Class I evidence isused to support recommendations of the strongest type,called Practice Standards, indicating a high degree of clinicalcertainty. Nonrandomized cohort studies, randomized con-trolled trials with design flaws, and case-control studies (com-parative studies with less strength) are designated Class IIevidence. Class II evidence is used to support recommenda-tions called Practice Guidelines, reflecting a moderate degreeof clinical certainty. Other sources of information, includingobservational studies (e.g., case series and expert opinion), aswell as randomized controlled trials with flaws so serious thatthe conclusions of the study are truly in doubt, generate ClassIII evidence. Class III evidence supports Practice Options, re-flecting unclear clinical certainty. These categories of evidenceare summarized in Table 5. The general term for all of therecommendations is Practice Parameters. Because so few Prac-tice Standards exist, the term more commonly used to de-scribe the whole body of recommendations is Practice Guide-lines. Thus, we have named this document Guidelines for theManagement of Acute Cervical Spine and Spinal Cord Injuries.

One of the practical difficulties encountered in implement-ing this methodology is that a poorly designed randomizedcontrolled trial might take precedence over a well-designedcase-control or nonrandomized cohort study. We attemptedto avoid this pitfall by carefully evaluating the quality of thestudy as well as its type.

These criteria apply to practice guidelines (parameters) fortreatment. To assess literature pertaining to prognosis, diag-nosis, and clinical assessment, completely different criteria

TABLE 3. Bayesian Table for Calculating the Reliability ofPatient Assessment Literature

Observer#2

Observer #1

Yes No

Yes Agree(a)

Disagree(b)

(a � b) � f1

No Disagree(c)

Agree(d)

(c � d) � f2

(a � c) � n1 (b � d) � n2 (a � b � c � d) � N

TABLE 4. Relationship between the Index of Concordance(�) and the Strength of Agreement between Observersor Measurements

Value of � Strength of Agreement

�0 Poor

0–0.20 Slight

0.21–0.40 Fair

0.41–0.60 Moderate

0.61–0.80 Substantial

0.81–1.00 Almost perfect

TABLE 5. Classification of Evidence onTherapeutic Effectiveness

Evidence Class Description

Class I Evidence from one or more well-designed, randomized controlled clinicaltrials, including overviews of such trials.

Class II Evidence from one or more well-designedcomparative clinical studies, such asnonrandomized cohort studies, case-control studies, and other comparablestudies, including less well-designedrandomized controlled trials.

Class III Evidence from case series, comparativestudies with historical controls, casereports, and expert opinion, as well assignificantly flawed randomizedcontrolled trials.

S4 Guidelines for Management of Acute Cervical Spinal Injuries

Neurosurgery, Vol. 50, No. 3, March 2002 Supplement

must be used. Criteria for prognosis have been developed andwere widely used in the publications of prognostic indicatorsin severe traumatic brain injury (1). We have not addressedissues of prognosis in this document. For diagnosis, the issueis the ability of the diagnostic test to successfully distinguishbetween patients who have and those who do not have thedisease or finding of interest. Classification of evidence ondiagnostic tests is illustrated in Table 6. Measures of clinicalassessment must be both reliable and valid. The assessmentmust be performed reliably by different observers and by thesame observer at different times. To be valid, the clinicalassessment, like diagnostic tests described above, must ade-quately represent the true condition of the patient. This latteraspect is difficult to measure, so most clinical assessments aregraded according to their reliability (Table 7).

For each question addressed in these guidelines, we as-sessed the study type in each identified article and assigned aclassification according to the scheme outlined above. Thesedesignations are listed in the evidentiary tables in eachchapter.

GUIDELINES DEVELOPMENT PROCESS

A group of individuals with interest and expertise in thetreatment of patients with cervical spine injuries and/or thedevelopment of guideline practice parameters was assembledunder the auspices of and with the support of the Joint Sectionon Disorders of the Spine and Peripheral Nerves of the Amer-ican Association of Neurological Surgeons/Congress of Neu-

rological Surgeons. The group reflected expertise in spinalneurosurgery, neurotrauma, and clinical epidemiology. Theissues chosen for inclusion in the document were those con-sidered pertinent to the management of patients with acutecervical spine and/or spinal cord injury (e.g., transport, med-ical management, treatment of specific fracture/dislocationpatterns, vascular injury, and prophylaxis for thromboem-bolic events).

A MEDLINE search of literature published from January1966 to January 2001 was carried out by using the searchterms described in each chapter. The search was limited tohuman subjects and included English language literature forall but one of the chapters. Additional articles were foundthrough the reference lists in the articles found, as well asfrom other sources known to the authors. Articles were re-jected on the basis of irrelevance to the clinical question athand. Case reports were included if there was insufficientmaterial from case series. All articles were evaluated accord-ing to the medical evidence-based protocol outlined above.Recommendations were derived for therapy, diagnosis, andclinical assessment. Chapters written by the primary authorswere rewritten by a different set of authors, and the finalproduct was agreed on by consensus. On occasion, the as-sessed quality of the study design was so contentious and theconclusions so uncertain that we assigned a lower classifica-tion than might have been expected without such a detailedreview. In every way, we sought to adhere to the Institute ofMedicine criteria for searching, assembling, evaluating, andweighting the available medical evidence and linking it to thestrength of the recommendations presented in this document.

Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.Email: [email protected]

TABLE 6. Classification of Evidence on Diagnosis


Class I Evidence provided by one or more well-designed clinical studies of a diversepopulation using a “gold standard”reference test in a blinded evaluationappropriate for the diagnosticapplications and enabling theassessment of sensitivity, specificity,positive and negative predictive values,and, where applicable, likelihood ratios.

Class II Evidence provided by one or more well-designed clinical studies of a restrictedpopulation using a “gold standard”reference test in a blinded evaluationappropriate for the diagnosticapplications and enabling theassessment of sensitivity, specificity,positive and negative predictive values,and, where applicable, likelihood ratios.

Class III Evidence provided by expert opinion,studies that do not meet the criteria forthe delineation of sensitivity, specificity,positive and negative predictive values,and, where applicable, likelihood ratios.

TABLE 7. Classification of Evidence on Clinical Assessment


Class I Evidence provided by one or more well-designed clinical studies in whichinterobserver and intraobserverreliability is represented by a � statisticof 0.80 or greater.

Class II Evidence provided by one or more well-designed clinical studies in whichinterobserver and intraobserverreliability is represented by a � statisticof 0.60 or greater.

Class III Evidence provided by one or more well-designed clinical studies in whichinterobserver and intraobserverreliability is represented by a � statisticof less than 0.60.

Methodology of Guideline Development S5


REFERENCES

1. Bullock R, Chesnut RM, Clifton GL, Ghajar J, Marion DW, Narayan RK,Newell DW, Pitts LH, Rosner MJ, Walters BC, Wilberger JE: Guidelinesfor the management of severe traumatic brain injury: Intracranial pres-sure treatment threshold. J Neurotrauma 17:493–495, 2000.

2. Bullock R, Chesnut RM, Clifton G, Ghajar J, Marion DW, NarayanRK, Newell DW, Pitts LH, Rosner MJ, Wilberger JW: Guidelines forthe management of severe head injury: Brain Trauma Foundation.Eur J Emerg Med 3:109–127, 1996.

3. Cohen J: A coefficient of agreement for nominal scales. EducPsychol Meas 20:37–46, 1960.

4. Field MJ, Lohr KN (eds): Clinical Practice Guidelines: Directions for aNew Program—Committee to Advise the Public Health Service on Clin-ical Practice Guidelines: Institute of Medicine. Washington DC, Na-tional Academy Press, 1990.

5. Landis JR, Koch GG: The measurement of observer agreement forcategorical data. Biometrics 33:159–174, 1977.

6. Rosenberg J, Greenberg MK: Practice parameters: Strategies forsurvival into the nineties. Neurology 42:1110–1115, 1992.

7. Walters BC: Clinical practice parameter development in neurosur-gery, in Bean JR (ed): Neurosurgery in Transition: The SocioeconomicTransformation of Neurological Surgery. Baltimore, Williams &Wilkins, 1998, pp 99–111.

Semischematic drawing of the cervical vertebrae shows that the cervical musculature stabilizes the cervical spinal column.Reproduced from, Keele KD, Pedretti C: Leonardo da Vinci: Corpus of the Anatomical Studies in the Collection of Her Maj-esty the Queen at Windsor Castle. London, Harcourt Brace Jovanovich, 1979.


CHAPTER 1

Cervical Spine Immobilization before Admission to the Hospital

RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS:• All trauma patients with a cervical spinal column injury or with a mechanism of injury having the potential

to cause cervical spine injury should be immobilized at the scene and during transport by using one ofseveral available methods.

• A combination of a rigid cervical collar and supportive blocks on a backboard with straps is effective inlimiting motion of the cervical spine and is recommended. The long-standing practice of attempted cervicalspine immobilization using sandbags and tape alone is not recommended.

RATIONALE

Early management of the patient with a potential cervicalspinal cord injury begins at the scene of the accident. Thechief concern during the initial management of patients

with potential cervical spine injuries is that neurological func-tion may be impaired by pathological motion of the injuredvertebrae. It is estimated that 3 to 25% of spinal cord injuriesoccur after the initial traumatic insult, either during transit orearly in the course of management (14, 15, 42, 48, 81, 97).Many cases have been reported that had a poor outcomebecause of mishandling of cervical spine injuries (12, 51, 81,97). As many as 20% of spinal column injuries involve mul-tiple noncontinuous vertebral levels; therefore, the entire spi-nal column is potentially at risk (38, 39, 66, 73). Consequently,complete spine immobilization has been used in spinal care,before admission to the hospital, to limit motion until injuryhas been ruled out (2, 5, 27, 40, 66, 73, 76, 100, 104). During thelast 30 years, the neurological status of spinal cord-injuredpatients arriving in emergency departments has dramaticallyimproved. During the 1970s, most patients (55%) referred toregional spinal cord injury centers arrived with completeneurological lesions. In the 1980s, however, most spinal cord-injured patients (61%) arrived with incomplete lesions (46).This improvement in the neurological status of patients hasbeen attributed to the development of emergency medicalservices (EMS) initiated in 1971, and the care (including spineimmobilization) rendered by EMS personnel before the pa-tient reaches the hospital (2, 45, 46, 103). Spine immobilizationis now an integral part of preadmission management and isadvocated, for all patients with potential spine injury aftertrauma, by EMS programs nationwide and by the AmericanCollege of Surgeons (1, 2, 5, 6, 16, 32, 70, 93).

Recently, the use of spine immobilization for all traumapatients, particularly those with a low likelihood of traumatic

cervical spinal injury, has been questioned. It is unlikely thatall patients rescued from the scene of an accident or site oftraumatic injury require spine immobilization (34, 50, 69, 76).Some authors have developed and advocate a triage systembased on clinical criteria to select patients for preadmissionspine immobilization (13, 32, 74).

Several devices are available for immobilizing the patientwith a potential spine injury during transportation to thehospital. However, the optimal device has not yet been iden-tified by careful comparative analysis (17, 21, 27, 53, 61, 64, 94,99). The recommendations of the American College of Sur-geons consist of a hard backboard, a rigid cervical collar,lateral support devices, and tape or straps to secure the pa-tient, the collar, and the lateral support devices to the back-board (3, 5). A more uniform, universally accepted method forspine immobilization for trauma patients with potential spineinjury may reduce the cost and improve the efficiency ofpreadmission spinal injury management (13, 32, 74). Al-though spine immobilization is typically effective in limitingmotion, it has been associated with morbidity in a smallpercentage of cases (4, 9, 18, 19, 26, 55, 90, 100). These issuesare the subject of this review on the use and effectiveness ofpreadmission spine immobilization.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of literature published from 1966 to 2001 was per-formed. The search was limited to the English language andhuman studies. The medical subject heading “spinal immobi-lization” produced 39 articles. A second search, combining theexploded terms “spinal injuries” and “immobilization,”yielded 122 articles. A third search, combining the explodedterms “spinal injuries” and “transportation of patients,”yielded 47 articles. A fourth search, combining the exploded


terms “spinal injuries” and “emergency medical services,”produced 119 articles. Additional references were culled fromthe reference lists of the articles. Finally, members of theauthor group were asked to contribute articles known to themon the subject matter that were not found by other searchmethods. Duplicate references were discarded. The abstractswere reviewed, and articles unrelated to the specific topicwere eliminated. This process yielded 101 articles for thisreview, which are listed in the reference list. Articles used toformulate this guideline are summarized in Table 1.1.

SCIENTIFIC FOUNDATION

Pathological motion of the injured cervical spine may createor exacerbate cervical spinal cord or cervical nerve root injury(38–40, 66, 73, 96). This potential has led to the use of spineimmobilization for trauma patients who have sustained acervical vertebral column injury or experienced a mechanismof injury that could result in cervical spinal column injury (5,6, 27, 33, 34, 40, 66, 73, 74, 76, 104).

Kossuth (56, 57) is credited with pioneering the currentlyaccepted methods of protecting and immobilizing the cervicalspine during extrication of patients with acute injury. Far-rington (36, 37) championed the concept of preadmissionimmobilization. Dick and Land (30) noted in their review ofspine immobilization devices that techniques of preadmissionspine immobilization appeared as early as 1971 in standardEMS texts and in the American Academy of Orthopedic Sur-geons Committee on Injuries Emergency text (2). Initially, thepreferred method for immobilizing the cervical spine was touse a combination of a soft collar and a rolled-up blanket (21).Later, in 1974, Hare introduced a more rigid extrication collar.Hare’s contribution launched an era of innovation for spineimmobilization devices (27).

Currently, in North America, spine immobilization is one ofthe most frequently performed procedures in the preadmis-sion care of patients with acute trauma (2, 6, 7, 27, 38, 40, 66,73, 76, 98, 104). Although clinical and biomechanical evidencedemonstrates that spine immobilization limits pathologicalmotion of the injured spinal column, there is no Class I orClass II medical evidence to support spinal column immobi-lization in all patients after trauma. Although immobilizationof an unstable cervical spinal injury makes good sense, andClass III evidence reports exist of neurological worseningwith failure of adequate spine immobilization, no case-controlstudies or randomized trials address the effect of spine im-mobilization on clinical outcomes after cervical spinal columninjury (6, 27, 32, 40, 42, 48, 50, 66, 69, 73, 96). The issue isimportant; tens of thousands of patients with trauma aretreated with spine immobilization each year, but few of themwill have spinal column injuries or instability (39, 74, 83).

Other considerations in the use of preadmission spine im-mobilization include the cost of equipment, the time andtraining of EMS personnel to apply the devices, and theunnecessary potential morbidity for patients who do not needspine immobilization (4, 9, 18, 19, 26, 27, 55, 58, 84, 90, 100). Aswith many interventions in the practice of medicine, spineimmobilization has been instituted in preadmission manage-

ment of trauma patients with potential spinal injuries on thebasis of principles of neural injury prevention and years ofclinical experience, but without supportive scientific evidencefrom rigorous clinical trials. For a variety of both practical andethical reasons, it may be impossible to obtain this informa-tion in clinical trials.

In 1989, Garfin et al. stated that no patient should be extri-cated from a crashed vehicle or transported from an accidentscene without spinal stabilization (40). The authors creditedstabilization of the cervical spine as a key factor in decliningpercentages of complete spinal cord injury lesions, from 55%in the 1970s to 39% in the 1980s, and in the significant reduc-tion of mortality in patients with multiple injuries that includecervical spine injuries. Unfortunately, no Class I or Class IImedical evidence supports their claims.

Few articles have directly evaluated the effect of preadmis-sion spine immobilization on neurological outcome after in-jury. Several Class III evidence reports cite lack of immobili-zation as a cause of neurological deterioration among acutelyinjured trauma patients transported to medical facilities fordefinitive care (12, 40, 51, 62, 81). The most pertinent study isToscano’s (96) retrospective case series report. Toscano, in1988, reported that 32 (26%) of 123 trauma patients sustainedmajor neurological deterioration in the period between injuryand admission at the hospital. The author attributed neuro-logical deterioration to patient mishandling and cited the lackof spine immobilization after traumatic injury as the primarycause. The report supports the need for spine immobilizationof trauma patients with potential spinal column injuries be-fore admission to the hospital.

In contrast, a collaborative, 5-year retrospective chart re-view reported by the University of New Mexico and theUniversity of Malaya challenges this position. Hauswald et al.(50) analyzed only patients with acute blunt spine or spinalcord injuries. At the University of Malaya, none of the 120patients they managed were immobilized with spinal ortho-ses during transport. All 334 patients managed at the Univer-sity of New Mexico were initially treated with spine immo-bilization. The hospitals were reportedly comparable inphysician training and clinical resources. Two independentphysicians, blinded to the participating hospital, character-ized the neurological injuries into two groups: disabling andnondisabling. Data were analyzed by logistic regression tech-niques, with hospital, patient age, sex, anatomic level of in-jury, and injury mechanism as variables. Neurological deteri-oration after injury was less frequent in patients with spinalinjuries in Malaya, who were not treated with formal spineimmobilization during transport (odds ratio, 2.03; 95% confi-dence interval, 1.03–3.99; P � 0.04), than in patients in NewMexico, who were managed with spinal column immobiliza-tion techniques. Even with the analysis limited to cervicalspine injuries, no significant protective effect from spine im-mobilization was identified. The authors theorized that be-cause the initial injury is of tremendous force, additionalmovement of the spine by the patient or rescuers is insuffi-cient to cause further injury. However, they noted that be-cause of the small sample size, the benefit of spine immobili-zation might not be statistically measurable in their study.



This report has been challenged, and several flaws have beenidentified. Patients who died at the scene or during transportwere excluded from analysis. Injuries were not matched byseverity of neurological injury or by type of spinal columninjury. The mechanisms of injury differed dramatically in thetwo populations. Malayan patients were immobilized or heldimmobile during transport, but spinal orthoses as immobili-zation devices were not used. For these reasons and others,the conclusions drawn by the authors are questionable (27,76).

Evidence in the literature evaluating the effectiveness ofpreadmission spine immobilization is sparse. The article byHauswald et al. (50) was published in 1998 after a period duringwhich universal spine immobilization after trauma had beenapplied in the United States and North America. Ethical andpractical issues preclude a contemporary clinical trial designedto study the effectiveness of preadmission spine immobilizationcompared with no immobilization, primarily because spine im-mobilization for trauma patients is perceived as essential withminimal risk and is already widely used. Intuitively, the use ofpreadmission spine immobilization is a rational means of limit-ing spinal motion in spine-injured patients in an effort to reducethe likelihood of neurological deterioration caused by patholog-ical motion at the site(s) of injury.

The consensus from all articles reviewed (Class III evidence),from an anatomic and biomechanical perspective and from time-tested clinical experience with traumatic spinal injuries, is that allpatients with cervical spinal column injuries, or those with thepotential for a cervical spinal injury after trauma, should betreated with spinal column immobilization until injury has beenexcluded or definitive management has been initiated. Althoughthere is insufficient medical evidence to support a treatmentstandard or a treatment guideline, practitioners are stronglyencouraged to provide spine immobilization to spine-injuredpatients (or those with a likelihood of spinal injury) until defin-itive assessment can be accomplished.

Orledge and Pepe (76) in their commentary on theHauswald findings (50) point out some limitations of thearticle, but they also suggest that it raises the issue of a moreselective evidence-based protocol for spine immobilization.Should all trauma patients be managed with spine immobili-zation until spinal injury has been excluded, or should immo-bilization be selectively used for patients with potential spinalinjury on the basis of well-defined clinical criteria? Whichclinical criteria should be used? After the Hauswald report,several prospective studies supported the use of clinical find-ings as indicators of the need for preadmission spine immo-bilization after trauma (33–35). Several EMS systems now useclinical protocols to help decide which patients should bemanaged with spine immobilization after trauma (43, 102).

Domeier et al. (32–34), in a multicenter prospective study of6500 trauma patients, found that the application of clinicalcriteria (altered mental status, focal neurological deficit, evi-dence of intoxication, spinal pain or tenderness, or suspectedextremity fracture) was predictive of most patients with cer-vical spinal injuries that required immobilization. The predic-tive value of their criteria held true for patients with high- orlow-risk mechanisms of injury. They suggested that clinical

criteria, rather than the mechanism of injury, be evaluated asthe standard for the use of spine immobilization.

Brown et al. (13) examined whether EMS providers couldaccurately apply clinical criteria to evaluate the cervical spinesof trauma patients before transport to a definitive care facility.The criteria included the presence of pain or tenderness of thecervical spine, the presence of a neurological deficit, an al-tered level of consciousness, evidence of drug use or intoxi-cation (particularly alcohol, analgesics, sedatives, or stimu-lants), and/or the presence of other significant trauma thatmight act as a distracting injury. Immobilization of the cervi-cal spine was initiated if any one of six criteria was present.The clinical assessment of trauma patients by EMS providerswas compared with the clinical assessment provided by emer-gency physicians. The providers (EMS technicians and emer-gency physicians) were blinded to each other’s assessments.Agreement between EMS and physician providers was ana-lyzed by � statistic. Five hundred seventy-three patients wereincluded in the study. The assessments matched in 79% of thecases (n � 451). For 78 patients (13.6%), the EMS clinicalassessment indicated spine immobilization but the physi-cian assessment did not. For 44 patients (7.7%), the physi-cian’s clinical assessment indicated spine immobilization butthe EMS assessment did not. For the individual components,� ranged from 0.35 to 0.81. For the decision to immobilize, �was 0.48. The EMS clinical assessments were generally morein favor of immobilization than the physician’s clinical assess-ments. Brown et al. concluded that EMS and physician clinicalassessments to rule out cervical spinal injury after traumahave moderate to substantial agreement. The authors recom-mended, however, that systems that allow EMS personnel todecide whether to immobilize patients after trauma shouldprovide attentive follow-up of those patients to ensure appro-priate care and to provide immediate feedback to the EMSproviders. Meldon et al. (71), in an earlier study, found sig-nificant disagreement between the clinical assessments andsubsequent spine immobilization of patients by EMS techni-cians and physicians. They recommended further researchand education before widespread implementation of thispractice.

Clinical criteria to select appropriate patients for spine im-mobilization are being studied in Michigan (102) and havebeen implemented in Maine (43) and San Mateo County, CA(88). Recommendations regarding the adoption of EMS pro-tocols for preadmission spine immobilization await definitivestudies of safety and efficacy (23). EMS personnel who makethese assessments require intensive education and careful,quality-assurance scrutiny to ensure that trauma patients withpotential spinal injuries are appropriately triaged and man-aged. Until further studies can be undertaken, the availableClass III studies support the use of spine immobilization forall patients with potential cervical spinal injury after trauma.

Devices and techniques for preadmissionspine immobilization

Preadmission spine immobilization is effective in limitingspinal motion during transportation of the patient (7, 27, 40,

Preadmission Cervical Spine Immobilization S9


TAB

LE1.

1.Su

mm

ary

ofR

epor

tson

Prea

dmis

sion

Cer

vica

lSp

ine

Imm

obili

zati

ona

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ceC

lass

Con

clus

ions

Mar

kens

onet

al.,

1999

(61)

An

eval

uatio

nof

the

Ken

dric

kex

tric

atio

nde

vice

bfo

rpe

diat

ric

spin

alim

mob

iliza

tion.

IIIK

endr

ick

extr

icat

ion

devi

cepr

ovid

esex

celle

ntst

atic

and

dyna

mic

imm

obili

zatio

n.

Perr

yet

al.,

1999

(77)

An

expe

rim

enta

lev

alua

tion

of3

imm

obili

zatio

nde

vice

sco

mpa

red

duri

ngsi

mul

ated

vehi

cle

mot

ion.

Nec

km

otio

nw

asju

dged

by3

phys

icia

ns.

IIISu

bsta

ntia

lam

ount

sof

head

mot

ion

can

occu

rdu

ring

sim

ulat

edve

hicl

em

otio

n,

rega

rdle

ssof

the

met

hod

ofim

mob

iliza

tion.

Mov

emen

tof

the

trun

kca

nha

veth

esa

me

effe

ctas

head

mot

ion

onm

otio

nac

ross

the

neck

.

Bau

eran

dK

owal

ski,

1998

(9)

Ast

udy

ofth

eef

fect

ofsp

inal

imm

obili

zatio

nde

vice

son

pulm

onar

yfu

nctio

nin

15m

en.

IIISi

gnifi

cant

rest

rict

ion

ofpu

lmon

ary

func

tion

may

resu

ltfr

omsp

inal

imm

obili

zatio

n.

Maw

son

etal

.,19

98(6

3)A

pros

pect

ive

stud

yto

dete

rmin

eth

eas

soci

atio

nbe

twee

nim

mob

iliza

tion

and

pres

sure

ulce

rsin

39SC

Ipa

tient

s.

IIITi

me

spen

ton

back

boar

dis

sign

ifica

ntly

asso

ciat

edw

ithpr

essu

reul

cers

deve

lopi

ng

with

in8

d.

Hau

swal

det

al.,

1998

(50)

5-yr

retr

ospe

ctiv

ech

art

revi

ewof

patie

nts

with

acut

etr

aum

atic

SCI

from

2ce

nter

s.N

one

ofth

e12

0pa

tient

sat

the

Uni

vers

ityof

Mal

aya

had

spin

alim

mob

iliza

tion

with

orth

otic

devi

ces

duri

ngtr

ansp

ort.

All

334

patie

nts

atth

eU

nive

rsity

ofN

ewM

exic

odi

d.Th

e

hosp

itals

wer

eco

mpa

rabl

e.N

euro

logi

cal

inju

ries

wer

eas

sign

edto

2ca

tego

ries

,di

sabl

ing

orno

tdi

sabl

ing,

by2

blin

ded

phys

icia

ns.

Dat

aw

ere

anal

yzed

usin

gm

ultiv

aria

telo

gist

ic

regr

essi

on.

Ther

ew

asle

ssne

urol

ogic

aldi

sabi

lity

inth

eM

alay

sian

patie

nts

(OR

,2.

03;

95%

CI,

1.03

–3.9

9;P

�0.

04).

Res

ults

wer

esi

mila

rw

hen

the

anal

ysis

was

limite

dto

patie

nts

with

cerv

ical

inju

ries

(OR

,1.

52;

95%

CI,

0.64

–3.6

2;P

�0.

34).

IIIO

ut-o

f-ho

spita

lim

mob

iliza

tion

has

little

effe

cton

neur

olog

ical

outc

ome

inpa

tient

sw

ith

blun

tsp

inal

inju

ries

.

The

asso

ciat

ion

betw

een

spin

alco

lum

nm

ovem

ent

and

the

pote

ntia

lfo

rSC

Ire

mai

ns

uncl

ear.

Bla

yloc

k,19

96(1

1)A

pros

pect

ive

stud

yto

dete

rmin

eth

eas

soci

atio

nbe

twee

nim

mob

iliza

tion

and

pres

sure

ulce

rsin

32SC

Ipa

tient

s.

IIIPr

essu

reso

res

deve

lope

d,m

ostly

inpa

tient

sw

how

ere

turn

edaf

ter

3h.

Mos

tof

thos

e

with

out

sore

sw

ere

turn

ed�

2h

afte

rim

mob

iliza

tion.

John

son

etal

.,19

96(5

2)M

easu

red

imm

obili

zatio

nan

dco

mfo

rton

10-p

oint

scal

e.Th

eva

cuum

splin

tw

as

com

pare

dw

ithba

ckbo

ard.

IIIV

acuu

msp

lints

are

mor

eco

mfo

rtab

lean

dfa

ster

toap

ply

than

back

boar

dsan

dpr

ovid

e

asi

mila

rde

gree

ofim

mob

iliza

tion.

Vac

uum

splin

tsar

eno

tri

gid

enou

ghfo

rex

tric

atio

nan

dar

em

ore

expe

nsiv

e.

Rod

gers

and

Rod

gers

,19

95

(84)

Cas

ere

port

ofm

argi

nal

man

dibu

lar

nerv

epa

lsy

due

toco

mpr

essi

onby

ace

rvic

alha

rd

colla

r.

IIITh

eco

llar

was

rem

oved

;th

epa

lsy

reso

lved

unev

entfu

llydu

ring

the

next

2d.

Cha

net

al.,

1994

(19)

Apr

ospe

ctiv

est

udy

ofth

eef

fect

sof

spin

alim

mob

iliza

tion

onpa

inan

ddi

scom

fort

in21

volu

ntee

rsaf

ter

30m

in.

All

subj

ects

deve

lope

dpa

in.

IIISt

anda

rdsp

inal

imm

obili

zatio

nm

aybe

aca

use

ofpa

inin

anot

herw

ise

heal

thy

subj

ect.

Liew

and

Hill

,19

94(5

9)2

case

repo

rts

ofsi

gnifi

cant

occi

pita

lpr

essu

reul

cera

tion

asso

ciat

edw

ithth

eus

eof

hard

cerv

ical

colla

r.

IIIPr

essu

reul

cers

may

occu

rw

ithth

eus

eof

hard

cerv

ical

colla

rs.

Maz

olew

ski,

1994

(64)

Ast

udy

tote

stth

eef

fect

iven

ess

ofst

rapp

ing

tech

niqu

esin

redu

cing

late

ral

mot

ion

ona

back

boar

din

labo

rato

ryin

19ad

ults

.

IIISt

rapp

ing

shou

ldbe

adde

dto

the

tors

oto

redu

cela

tera

lm

otio

non

aba

ckbo

ard.

Plai

sier

etal

.,19

94(7

8)A

pros

pect

ive

eval

uatio

nof

cran

iofa

cial

pres

sure

offo

urdi

ffere

ntce

rvic

alor

thos

esin

20

adul

ts.

Pres

sure

was

mea

sure

dat

the

occi

put,

man

dibl

e,an

dch

in.

Opi

nion

son

com

fort

wer

eal

soco

llect

ed.

IIITh

eN

ewpo

rtor

Mia

mi-

Jco

llars

have

favo

rabl

esk

inpr

essu

repa

ttern

san

dsu

peri

or

patie

ntco

mfo

rt.

Rap

hael

and

Cho

tai,

1994

(82)

Ara

ndom

ized

,si

ngle

-blin

d,cr

osso

ver

stud

yof

9pa

tient

ssc

hedu

led

for

elec

tive

spin

al

anes

thes

ia.

The

cere

bros

pina

lflu

idpr

essu

rein

the

lum

bar

suba

rach

noid

spac

ew

as

mea

sure

dw

ithan

dw

ithou

ta

Stifn

eck

cerv

ical

colla

rap

plie

d.

IIITh

ere

was

asi

gnifi

cant

elev

atio

nof

cere

bros

pina

lflu

idpr

essu

rein

7of

the

patie

nts

stud

ied

whe

nth

ece

rvic

alco

llar

was

appl

ied

(P�

0.01

).

Cha

ndle

ret

al.,

1992

(20)

Aco

mpa

riso

nof

the

rigi

dce

rvic

alex

tric

atio

nco

llar

with

Am

mer

man

halo

orth

osis

in20

men

.

IIIA

mm

erm

anha

loor

thos

isan

dsp

ine

boar

dpr

ovid

edsi

gnifi

cant

lybe

tter

imm

obili

zatio

n,

equi

vale

ntto

halo

vest

.

Ros

en,

1992

(87)

Aco

mpa

riso

nof

4ce

rvic

alco

llars

in15

adul

tvo

lunt

eers

,by

goni

omet

ry.

IIIV

acuu

msp

lint

cerv

ical

colla

rre

stri

cted

rang

eof

mot

ion

ofth

ece

rvic

alsp

ine

mos

t

effe

ctiv

ely.



TAB

LE1.

1.C

onti

nued

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ceC

lass

Con

clus

ions

Scha

ferm

eyer

etal

.,19

91

(89)

Ast

udy

toas

sess

the

rest

rict

ive

effe

cts

of2

spin

alim

mob

iliza

tion

stra

ppin

g

tech

niqu

eson

the

resp

irat

ory

capa

city

and

forc

edvi

tal

capa

city

of51

child

ren.

IIISp

inal

imm

obili

zatio

nsi

gnifi

cant

lyre

duce

dre

spir

ator

yca

paci

tyas

mea

sure

dby

FVC

in

heal

thy

patie

nts

6–15

yrol

d.Th

ere

isno

sign

ifica

ntbe

nefit

ofon

est

rapp

ing

tech

niqu

e

over

the

othe

r.

Schr

iger

etal

.,19

91(9

1)A

stud

yco

mpa

ring

the

flat

back

boar

dw

ithoc

cipi

tal

padd

ing

inac

hiev

ing

neut

ral

posi

tion

in10

0he

alth

yvo

lunt

eers

.

IIIO

ccip

ital

padd

ing

plac

esth

ece

rvic

alsp

ine

inm

ore

neut

ral

alig

nmen

t.

Coh

enet

al.,

1990

(22)

Ast

udy

anal

yzin

gth

eR

EDin

64pa

tient

s.III

RED

isan

effe

ctiv

esp

inal

imm

obili

zatio

nde

vice

with

adva

ntag

esov

ercu

rren

tly

avai

labl

ede

vice

s.

Bar

ney

and

Cor

dell,

1989

(8)

Eval

uate

dpa

inan

ddi

scom

fort

duri

ngim

mob

iliza

tion

onri

gid

spin

ebo

ards

in90

patie

nts.

IIISp

ine

boar

dsm

ayca

use

disc

omfo

rt.

Tosc

ano,

1988

(96)

Prev

entio

nof

neur

olog

ical

dete

rior

atio

nbe

fore

adm

issi

onto

hosp

ital.

Ret

rosp

ectiv

e

revi

ewof

123

patie

nts;

32of

123

sust

aine

dm

ajor

neur

olog

ical

dete

rior

atio

nfr

om

inju

ryto

adm

issi

on.

IIIA

ppro

pria

teha

ndlin

gof

patie

nts

with

spin

alin

jury

afte

rtr

aum

aca

nre

duce

maj

or

neur

olog

ical

dete

rior

atio

ndu

eto

path

olog

ical

mot

ion

ofve

rteb

ral

colu

mn.

Gra

zian

oet

al.,

1987

(44)

Ara

diog

raph

icco

mpa

riso

nof

prea

dmis

sion

cerv

ical

imm

obili

zatio

nm

etho

dsw

ith

the

shor

tbo

ard

tech

niqu

ein

45vo

lunt

eers

.

IIITh

esh

ort-

boar

dte

chni

que

prov

edto

besi

gnifi

cant

lybe

tter

(P�

0.05

)th

anK

endr

ick

extr

icat

ion

devi

cean

dco

llars

.

Lina

res

etal

.,19

87(6

0)A

stud

yof

32SC

Ipa

tient

sto

dete

rmin

ew

heth

erpr

essu

reso

res

are

asso

ciat

edw

ith

prol

onge

dim

mob

iliza

tion.

IIITh

ere

isa

stro

ngas

soci

atio

nbe

twee

n1–

2h

ofim

mob

iliza

tion

and

the

deve

lopm

ent

of

pres

sure

sore

s.

McG

uire

etal

.,19

87(6

8)A

radi

ogra

phic

eval

uatio

nof

mot

ion

ofth

eth

orac

olum

bar

spin

ein

aca

dave

rw

ithan

unst

able

thor

acol

umba

rsp

ine,

and

apa

tient

with

aT1

2–L1

frac

ture

disl

ocat

ion.

IIIEx

trem

em

otio

nat

anun

stab

leth

orac

olum

bar

spin

ese

gmen

tca

noc

cur

duri

ngth

e

logr

oll

man

euve

r.Th

eba

ckbo

ard

and

the

Scoo

pst

retc

her

offe

red

adeq

uate

stab

iliza

tion

for

thor

acol

umba

rsp

ine

inst

abili

ty.

McC

abe

and

Nol

an,

1986

(65)

Ara

diog

raph

icco

mpa

riso

nof

the

4ce

rvic

alco

llars

in7

adul

ts.

IIIPo

lyet

hyle

ne-1

prov

ided

mos

tre

stri

ctio

nin

flexi

on.

Clin

eet

al.,

1985

(21)

Ara

diog

raph

icco

mpa

riso

nof

7m

etho

dsof

cerv

ical

imm

obili

zatio

nin

97ad

ults

.III

The

shor

t-bo

ard

tech

niqu

eap

pear

edto

besu

peri

orto

the

3co

llars

stud

ied.

The

colla

rs

prov

ided

noau

gmen

tatio

nof

imm

obili

zatio

nov

erth

atpr

ovid

edby

the

shor

tbo

ard

alon

e.

Podo

lsky

etal

.,19

83(7

9)St

atic

tria

lus

ing

goni

omet

ryco

mpa

ring

soft

colla

r,ha

rdco

llar,

extr

icat

ion

colla

r,

Phila

delp

hia

colla

r,bi

late

ral

sand

bags

and

tape

,an

dth

eco

mbi

natio

nof

sand

bags

,

tape

,an

dth

ePh

ilade

lphi

aco

llar

in25

norm

alad

ult

volu

ntee

rs.

IIIH

ard

foam

and

plas

ticco

llars

wer

esu

peri

orto

soft

colla

rs.

Sand

bags

and

tape

in

com

bina

tion

with

ari

gid

cerv

ical

colla

rw

asth

ebe

stm

eans

ofth

ose

eval

uate

dto

limit

cerv

ical

spin

em

otio

n.Th

ead

ditio

nof

aPh

ilade

lphi

aco

llar

was

sign

ifica

ntly

mor

e

effe

ctiv

ein

redu

cing

neck

exte

nsio

n(P

�0.

01),

from

15to

7.4

degr

ees,

ach

ange

of

49.3

%.

The

com

bina

tion

ofsa

ndba

gsan

dta

peal

one

does

not

allo

wsu

ffici

ent

rest

rict

ion

ofm

otio

n,pa

rtic

ular

lyin

exte

nsio

n.

aSC

I,sp

inal

cord

inju

ry;

FVC

,fo

rced

vita

lca

paci

ty;

RED

,R

usse

llex

tric

atio

nde

vice

;O

R,

odds

rati

o;C

I,co

nfid

ence

inte

rval

.b

Ken

dric

kex

tric

atio

nde

vice

,Fer

no-W

ashi

ngto

n,In

c.,W

ilmin

gton

,OH

;New

port

colla

r(n

owkn

own

asA

spen

colla

r),F

ijiM

anuf

actu

ring

,Inc

.,Lo

ngB

each

,CA

;Sco

opst

retc

her,

Fern

o-W

ashi

ngto

n,W

ilmin

gton

,OH

;Mia

mi-

J,Je

rom

eM

edic

al,M

oore

stow

n,N

J;St

ifnec

kri

gid

colla

r,A

llian

ceM

edic

al,R

usse

lvill

e,M

O;P

olye

thyl

ene-

1,A

llian

ceM

edic

al,R

usse

lvill

e,M

O;A

mm

erm

anha

loor

thos

is,A

mm

erm

anTr

aum

aSy

stem

s,Pa

cific

Palis

ades

,CA

;Rus

sell

extr

icat

ion

devi

ce,M

illa

Mit

chel

l&C

o.,N

ewSo

uth

Wal

es,A

ustr

alia

;Phi

lade

lphi

aco

llar,

Phila

delp

hia

Col

lar

Co.

,W

estv

ille,

NJ.



66, 73, 104). Various devices and techniques are available toprovide immobilization of the cervical spine. Attempts todefine the best method have been hampered by physical andethical constraints (17, 27, 53, 61, 64, 94, 99).

Ways of measuring the efficacy of spine immobilizationdevices vary among investigators. Comparative studies of thevarious devices have been performed on healthy volunteers,but none have been tested in a large number of patients withspinal injury. It is difficult to extrapolate normative data toinjured patients with spinal instability (17, 20, 24, 27, 29, 49,52, 53, 58, 65, 67, 77, 94, 98, 99).

Several methods have been used to measure movement ofthe cervical spine. They include clinical assessment, plumblines, photography, radiography, cinematography, computedtomography, and magnetic resonance imaging. Roozmon etal. (85) summarized the problems inherent in each methodand concluded that there was no satisfactory noninvasivemeans of studying neck motion, particularly if one is to quan-tify movement between individual vertebral segments.

The position in which the injured spine should be placedand held immobile, the “neutral position,” is poorly defined(25, 28, 75, 88, 92). Schriger defined the neutral position as thenormal anatomic position of the head and torso that oneassumes when standing and looking ahead (90). This positioncorrelates to 12 degrees of cervical spine extension on a lateralradiograph. Schriger comments that the extant radiographicdefinition of neutral position was based on radiographicstudy of patients who were visually observed to be in theneutral position. Schriger et al. (91) used this position in theirevaluation of occipital padding on spine immobilization back-boards. De Lorenzo et al. (28), in their magnetic resonanceimaging study of 19 adults, found that a slight degree offlexion equivalent to 2 cm of occiput elevation produces afavorable increase in spinal canal/spinal cord ratio at levelsC5 and C6, a region of frequent unstable cervical spine inju-ries. Backboards have been used for years in extricating andimmobilizing spine-injured patients. Schriger et al. (91) ques-tioned the ability of a flat board to allow neutral positioning ofthe cervical spine. They compared spine immobilization byusing the flat backboard with and without occipital paddingin 100 adults. Clinical observation and assessment were usedto determine the neutral position of the cervical spine. Theauthors found that occipital padding combined with a rigidbackboard places the cervical spine in a better neutral positionthan a flat backboard alone (91, 93). McSwain (70) determinedthat more than 80% of adults require 1.3 to 5.1 cm of paddingto achieve neutral positioning of the head and neck relative tothe torso and noted that physical characteristics and musculardevelopment alter the cervical-thoracic angle, thus affectingpositioning. This makes it impossible to dictate specific rec-ommendations for padding.

In general, spine immobilization consists of a cervical collar,supports on either side of the head, and the long and shortbackboards with associated straps to attach and immobilizethe entire body to the board (27). Garth (41) proposed perfor-mance standards for cervical extrication collars, but thesestandards have not been uniformly implemented. A variety ofdifferent cervical collars is available. Several studies compare

collars alone or combined with other immobilization devicesby a wide range of assessment criteria (17, 19, 20, 24, 94, 99).

Podolsky et al. (79), in 1983, evaluated the efficacy of cer-vical spine immobilization techniques by using goniometricmeasurements. Twenty-five healthy volunteers lying supineon a rigid emergency department resuscitation table wereasked to actively move their necks as far as possible in sixways: flexion, extension, rotation to the right and left, andlateral bending to the right and left. Control measurementswere made with no device, and then measurements wererepeated after immobilization in a soft collar, hard collar,extrication collar, Philadelphia collar (Philadelphia Collar Co.,Westville, NJ), bilateral sandbags joined with 3-inch-widecloth tape across the forehead attached to either side of theresuscitation table, and the combination of sandbags, tape,and a Philadelphia collar. Hard foam and hard plastic collarswere better at limiting cervical spine motion than soft foamcollars. Neither collars alone nor sandbags and tape in com-bination provided satisfactory restriction of cervical spinemotion. For all six cervical spine movements, sandbags andtape immobilization were significantly better than any of theother methods of attempted cervical spine immobilizationused alone. The authors found that sandbags and tape com-bined with a rigid cervical collar were the best means of thoseevaluated to limit cervical spine motion. Adding a Philadel-phia collar to the sandbag and tape construct significantlyreduced neck extension (P � 0.01), from 15 degrees to 7.4degrees, a change of 49.3%. Collar use had no significantadditive effect for any other motion of the cervical spine.Sandbags as adjuncts to cervical spine immobilization requiremore rather than less attention from care providers (54). Sand-bags are heavy, and, if the extrication board is tipped side toside during evacuation and transport, the sandbags can slide,resulting in lateral displacement of the patient’s head andneck with respect to the torso. Sandbags can be taped to theextrication board, but because they are small compared withthe patient, this can be difficult and/or ineffective. Finally,sandbags must be removed before initial lateral cervical spinex-ray assessment because they can obscure the radiographicbony anatomy of the cervical spine. For these reasons (54) andthe findings by Podolsky et al. (79), use of sandbags and tapealone to attempt to immobilize the cervical spine is notrecommended.

In 1985, Cline et al. (21) compared methods of cervical spineimmobilization used in preadmission transport. The authorsfound that strapping the patient to a standard short board wasmore effective than cervical collar use alone. They noted nosignificant differences among the rigid collars they tested.McCabe and Nolan (65) used radiographic assessment tocompare four different collars for their ability to restrict mo-tion in flexion-extension and lateral bending. They found thatthe Polyethylene-1 collar (Alliance Medical, Russelville, MO)provided the most restriction of motion of the cervical spine,particularly for flexion. Rosen (87), in 1992, used goniometricmeasurements to compare limitation of cervical spine move-ment of four rigid cervical collars on 15 adults. Of the fourdevices they tested, the vacuum splint cervical collar pro-



vided the most effective restriction of motion of the cervicalspine.

Graziano et al. (44) compared preadmission cervical spineimmobilization methods by measuring cervical motion radio-graphically in the coronal and sagittal planes in 45 immobi-lized adults. In this study, the Kendrick extrication device(Ferno-Washington, Inc., Wilmington, OH) and the Extrica-tion Plus-One device (Medical Specialties, Inc., Charlotte, NC)were nearly as effective in limiting cervical motion as theshort immobilization board. Both devices were more effectivethan a rigid cervical collar alone.

Cohen et al. (22), in 1990, described the Russell extricationdevice (RED) (Milla Mitchell & Co., New South Wales, Aus-tralia) for immobilization of patients with potential spineinjuries. The RED was comparable to the short immobilizationboard for preadmission spine immobilization. Chandler et al.(20) compared a rigid cervical extrication collar with the Am-merman halo orthosis (Ammerman Trauma Systems, PacificPalisades, CA) in 20 men. The Ammerman halo orthosis com-bined with a rigid spine board provided significantly bettercervical spine immobilization than a cervical collar and spineboard. The Ammerman halo orthosis and spine board wasequivalent to the standard halo vest immobilization device.

Perry et al. (77) evaluated three cervical spine immobiliza-tion devices during simulated vehicle motion in six adults.Neck motion was assessed by three neurologists and neuro-surgeons as to whether motion was “clinically significant.”The authors found that substantial head motion occurredduring simulated vehicle motion regardless of the method ofimmobilization. The authors observed that the efficacy ofcervical spine immobilization was limited unless the motionof the head and the trunk was also effectively controlled.Mazolewski (64) tested the effectiveness of strapping tech-niques to reduce lateral motion of the spine of adults re-strained on a backboard. Subjects were restrained on awooden backboard with four different strapping techniques.The backboard was rolled to the side, and lateral motion ofthe torso was measured. The author found that additionalstrapping securing the torso to backboard reduced lateralmotion of the torso.

Finally, the traditional method of moving a patient onto along backboard has typically involved the logroll maneuver.The effectiveness of this transfer technique has been ques-tioned (31, 87). Significant lateral motion of the lumbar spinehas been reported (68, 95). Alternatives to the logroll maneu-ver include the HAINES method and the multihand or fire-man lift method (4, 5, 47). In the HAINES method (acronymfor High Arm IN Endangered Spine), the patient is placedsupine, the upper arm away from the kneeling rescuer isabducted to 180 degrees, the near arm of the patient is placedacross the patient’s chest, and both lower limbs are flexed. Therescuer’s hands stabilize the head and neck and the patient isrolled away onto an extrication board or device (47). Themultihand or fireman lift method involves several rescuers oneither side of the patient; the rescuers slide their arms under-neath the patient and lift the patient from one position toanother onto an extrication board or device.

This review depicts the evolution of techniques availablefor providing preadmission spine immobilization of spine-injured patients during transport and underscores their diver-sity. These studies are limited by the fact that none of thestudies evaluates the full range of available devices usingsimilar criteria. Overall, it seems that a combination of rigidcervical collar immobilization with supportive blocks on arigid backboard with straps to secure the entire body of thepatient is most effective in limiting motion of the cervicalspine after traumatic injury (5). The long-standing practice ofattempted spine immobilization using sandbags and tapealone is insufficient.

Safety of preadmission spine immobilization devices

Despite obvious benefits, cervical spine immobilization hasa few potential drawbacks. Immobilization can be uncomfort-able, it takes time to apply, application may delay transport,and it is associated with modest morbidity (4, 9, 18, 19, 26, 90,100).

Chan et al. (19) studied the effects of spine immobilizationon pain and discomfort in 21 healthy adults. Subjects wereplaced in backboard immobilization for 30 minutes, andsymptoms were chronicled. All subjects developed pain,which was described as moderate to severe in 55% of volun-teers. Occipital headache and sacral, lumbar, and mandibularpain were the most frequent complaints. In a later study,Chan et al. (18) compared spine immobilization on a back-board to immobilization with a vacuum mattress-splint de-vice in 37 healthy adults. The authors found that the fre-quency and severity of occipital and lumbosacral pain wassignificantly higher during backboard immobilization than onthe vacuum mattress-splint device. Johnson et al. (52) per-formed a prospective, comparative study of the vacuum splintdevice versus the rigid backboard. The vacuum splint devicewas significantly more comfortable than the rigid backboardand could be applied more quickly. The vacuum splint deviceprovided better immobilization of the torso. The rigid back-board with head blocks was slightly better at immobilizingthe head. Vacuum splint devices, however, are not recom-mended for extrication because they are reportedly not rigidenough, and they are more expensive. At a cost of approxi-mately $400, the vacuum splint device is roughly three timesmore expensive than a rigid backboard (18).

Hamilton and Pons (49) studied the comfort level of 26adults on a full-body vacuum splint device compared with arigid backboard, with and without cervical collars. Subjectsgraded their immobilization and discomfort. No statisticallysignificant difference was found between the vacuum splintdevice and collar combination and the backboard and collarcombination for flexion and rotation. The vacuum splint-collar combination provided significantly better immobiliza-tion in extension and lateral bending than the backboard-collar combination. The vacuum splint alone provided bettercervical spine immobilization in all neck positions exceptextension than the rigid backboard alone. A statistically sig-nificant difference in subjective perception of immobilizationwas noted; the backboard alone was less effective than the



three alternatives. In conclusion, the vacuum splint device,particularly when used with a cervical collar, is an effectiveand comfortable alternative to a rigid backboard (with orwithout the collar) for cervical spine immobilization.

Barney and Cordell (8) evaluated pain and discomfort dur-ing immobilization on rigid spine boards in 90 trauma pa-tients and found that rigid spine boards cause discomfort.Padding the rigid board improves patient comfort withoutcompromising cervical spine immobilization (101). Minimiz-ing the pain of immobilization may decrease voluntary move-ment and therefore decrease the likelihood of secondary in-jury (19).

Cervical collars have been associated with elevated intra-cranial pressure (ICP). Davies et al. (26) prospectively ana-lyzed ICP in a series of injured patients managed with theStifneck rigid collar (Alliance Medical). ICP rose significantly(P � 0.001; mean, 4.5 mm Hg) when the collar was firmly inplace. The authors cautioned that because head-injured pa-tients may also require cervical spine immobilization, it isessential that secondary insults producing raised ICP be min-imized. Kolb et al. (55) also examined changes in ICP after theapplication of a rigid Philadelphia collar in 20 adult patients.ICP averaged 176.8 mm H2O initially and increased to anaverage of 201.5 mm H2O after collar placement. Althoughthe difference in ICP of 24.7 mm H2O was statistically signif-icant (P � 0.001), it remains uncertain that it has clinicalrelevance. Nonetheless, this modest increase in pressure maybe important in patients who already have elevated ICP.Plaisier et al. (78), in 1994, prospectively evaluated craniofa-cial pressure with the use of four different cervical orthoses.The authors found small changes in craniofacial pressure(increases) but no significant differences among the four collartypes.

Spine immobilization increases the risk of pressure sores.Linares et al. (60) found that pressure sores were associatedwith immobilization (patients who were not turned duringthe first 2 hours after injury). The development of pressuresores was not related to mode of transportation to hospital orto the use of a spinal board and sandbags during transporta-tion. Mawson et al. (63) prospectively assessed the develop-ment of pressure ulcers in 39 spinal cord-injured patients whowere immobilized immediately after injury. The length oftime on a rigid spine board was significantly associated withthe development of decubitus ulcers within 8 days of injury (P� 0.01). Rodgers and Rodgers (84) reported a marginal man-dibular nerve palsy caused by compression by a hard collar.The palsy resolved uneventfully during the next 2 days. Blay-lock (11) found that prolonged cervical spine immobilizationmay result in pressure ulcers. Improved skin care (keepingthe skin dry), proper fitting (avoid excessive tissue pressure),and the appropriate choice of collars (those that do not trapmoisture and do not exert significant tissue pressure) canreduce this risk (10, 11).

Cervical spine immobilization may also increase the risk ofaspiration and may limit respiratory function. Bauer andKowalski (9) examined the effect of the Zee Extrication Device(Zee Medical Products, Irvine, CA) and the long spinal boardon pulmonary function. They tested pulmonary function in 15

healthy, nonsmoking men by using forced vital capacity,forced expiratory volume in 1 second, the ratio of forcedexpiratory volume in 1 second to the forced vital capacity, andforced midexpiratory flow (25–75%). They found a significantdifference (P � 0.05) between before-strapping and after-strapping values for three of the four functions tested whenon the long spinal board. Similarly, significant differenceswere found for three of the four parameters when using theZee Extrication Device. These differences reflect a markedpulmonary restrictive effect of appropriately applied entire-body spine immobilization devices.

Totten and Sugarman (97) evaluated the effect of whole-body spine immobilization on respiration in 39 adults. Respi-ratory function was measured at baseline, once immobilizedwith a Philadelphia collar on a rigid backboard, and whenimmobilized on a Scandinavian vacuum mattress with a vac-uum collar. The comfort levels of each of the two methodswere assessed on a visual analog scale. Both immobilizationmethods restricted respiration by an average of 15%. Theeffects were similar under the two methods, although theforced expiratory volume in 1 second was lower on the vac-uum mattress. The vacuum mattress was significantly morecomfortable than the wooden backboard (4).

In conclusion, cervical spine immobilization devices aregenerally effective in limiting motion of the cervical spine butmay be associated with important but usually modest mor-bidity. Cervical spine immobilization devices should be usedto achieve the goals of safe extrication and transport butshould be removed as soon as it is safe to do so.

SUMMARY

Spine immobilization can reduce untoward movement ofthe cervical spine and can reduce the likelihood of neurolog-ical deterioration in patients with unstable cervical spine in-juries after trauma. Immobilization of the entire spinal col-umn is necessary in these patients until a spinal column injury(or multiple injuries) or a spinal cord injury has been ex-cluded, or until appropriate treatment has been initiated.Although not supported by Class I or Class II medical evi-dence, this effective, time-tested practice is based on anatomicand mechanical considerations in an attempt to prevent spinalcord injury and is supported by years of cumulative traumaand triage clinical experience.

It is unclear whether the spines of all patients with traumamust be immobilized during preadmission transport. Manypatients do not have spinal injuries and therefore do notrequire such intervention. The development of specific selec-tion criteria for those patients for whom immobilization isindicated remains an area of investigation.

The variety of techniques used and the lack of definitiveevidence to advocate a uniform device for spine immobiliza-tion make it difficult to formulate recommendations for im-mobilization techniques and devices. It seems that a combi-nation of rigid cervical collar with supportive blocks on arigid backboard with straps is effective at achieving safe,effective spine immobilization for transport. The long-



standing practice of attempting to immobilize the cervicalspine with sandbags and tape alone is not recommended.

Cervical spine immobilization devices are effective but canresult in patient morbidity. Spine immobilization devicesshould be used to achieve the goals of spinal stability for safeextrication and transport. They should be removed as soon asdefinitive evaluation is accomplished and/or definitive man-agement is initiated.

KEY ISSUES FOR FUTURE INVESTIGATION

The optimal device for immobilization of the cervical spineafter traumatic vertebral injury should be studied in a pro-spective fashion. A reliable in-field triage protocol to be ap-plied by EMS personnel for patients with potential cervicalspine injuries after trauma needs to be developed.


REFERENCES

1. Alexander RH, Proctor HJ: Advanced Trauma Life Support Coursefor Physicians: ATLS. Chicago, American College of Surgeons,1993, ed 5, pp 21–22.

2. American Academy of Orthopedic Surgeons, Committee on In-juries: Emergency Care and Transportation of the Sick and Injured.Chicago, American Academy of Orthopedic Surgeons, 1971, pp111–115.

3. American College of Surgeons, Committee on Trauma: AdvancedTrauma Life Support. Chicago, American College of Surgeons,1993, p 201.

4. American College of Surgeons, Committee on Trauma: AdvancedTrauma Life Support. Chicago, American College of Surgeons,1993, pp 214–218.

5. American College of Surgeons, Committee on Trauma: Spineand spinal cord trauma, in Advanced Trauma Life Support Programfor Doctors: ATLS. Chicago, American College of Surgeons, 1997,ed 6, pp 215–242.

6. Augustine J: Spinal trauma, in Campbell JE (ed): Basic TraumaLife Support: Advanced Pre-hospital Care. Englewood Cliffs,Prentice-Hall, 1998, ed 2, p 120.

7. Augustine J: Spinal trauma, in Campbell JE (ed): Basic TraumaLife Support for Paramedics and Advanced EMS providers. UpperSaddle River, Brady, 1998, ed 3, p 153.

8. Barney R, Cordell W: Pain associated with immobilization onrigid spine boards. Ann Emerg Med 18:918, 1989 (abstr).

9. Bauer D, Kowalski R: Effect of spinal immobilization devices onpulmonary function in the healthy, non-smoking man. AnnEmerg Med 17:915–918, 1988.

10. Black CA, Buderer NM, Blaylock, B, Hogan BJ: Comparativestudy of risk factors for skin breakdown with cervical orthoticdevices. J Trauma Nurs 5:62–66, 1998.

11. Blaylock B: Solving the problem of pressure ulcers resultingfrom cervical collars. Ostomy Wound Manage 42:26–33, 1996.

12. Bohlman HH: Acute fractures and dislocations of the cervicalspine: An analysis of three hundred hospitalized patients andreview of the literature. J Bone Joint Surg Am 61A:1119–1142,1979.

13. Brown LH, Gough JE, Simonds WB: Can EMS providers ade-quately assess trauma patients for cervical spinal injury?Prehosp Emerg Care 2:33–36, 1998.

14. Brunette DD, Rockswold GL: Neurologic recovery followingrapid spinal realignment for complete cervical spinal cord in-jury. J Trauma 27:445–447, 1987.

15. Burney RE, Waggoner R, Maynard FM: Stabilization of spinalinjury for early transfer. J Trauma 29:1497–1499, 1989.

16. Butman A, Vomacka R: Part 1: Spine immobilization. Emergency23:48–51, 1991.

17. Carter VM, Fasen JA, Roman JM Jr, Hayes KW, Petersen CM:The effect of a soft collar, used as normally recommended orreversed, on three planes of cervical range of motion. J OrthopSports Phys Ther 23:209–215, 1996.

18. Chan D, Goldberg RM, Mason J, Chan L: Backboard versusmattress splint immobilization: A comparison of symptoms gen-erated. J Emerg Med 14:293–298, 1996.

19. Chan D, Goldberg R, Tascone A, Harmon S, Chan L: The effectof spinal immobilization on healthy volunteers. Ann EmergMed 23:48–51, 1994.

20. Chandler DR, Nemejc C, Adkins RH, Waters RL: Emergencycervical-spine immobilization. Ann Emerg Med 21:1185–1188,1992.

21. Cline JR, Scheidel E, Bigsby EF: A comparison of methods ofcervical immobilization used in patient extrication and trans-port. J Trauma 25:649–653, 1985.

22. Cohen A, Bosshard R, Yeo JD: A new device for the care of acutespinal injuries: The Russell Extrication Device (RED). Paraplegia28:151–157, 1990.

23. Cone DC, Wydro GC, Mininger CM: Current practice in clinicalcervical spinal clearance: Implication for EMS. Prehosp EmergCare 3:42–46, 1999.

24. Cooke M: Spinal boards. J Accid Emerg Med 13:433, 1996 (let-ter).

25. Curran C, Dietrich AM, Bowman MJ, Ginn-Pease ME, King DR,Kosnik E: Pediatric cervical-spine immobilization: Achievingneutral position? J Trauma 39:729–732, 1995.

26. Davies G, Deakin C, Wilson A: The effect of a rigid collar onintracranial pressure. Injury 27:647–649, 1996.

27. De Lorenzo RA: A review of spinal immobilization techniques.J Emerg Med 14:603–613, 1996.

28. De Lorenzo RA, Olson JE, Boska M, Johnston R, Hamilton GC,Augustine J, Barton R: Optimal positioning for cervical immo-bilization. Ann Emerg Med 28:301–308, 1996.

29. Dick T: Comparing the short-board technique. Ann Emerg Med18:115–116, 1989 (letter).

30. Dick T, Land R: Spinal immobilization devices: Part 1—Cervicalextrication collars. J Emerg Med Serv JEMS 7:26–32, 1982.

31. Dick T, Land R: Full spinal immobilizers. J Emerg Med ServJEMS 8:34–36, 1983.

32. Domeier RM: Indications for pre-hospital spinal immobilization:National Association of EMS Physicians Standards and ClinicalPractice Committee. Prehosp Emerg Care 3:251–253, 1999.

33. Domeier RM, Evans RW, Swor RA, Hancock JB, Fales W,Krohmer J, Fredericksen SM, Shork MA: The reliability of pre-hospital clinical evaluation for potential spinal injury is notaffected by the mechanism of injury. Prehosp Emerg Care 3:332–337, 1999.

34. Domeier RM, Evans RW, Swor RA, Rivera-Rivera EJ,Fredericksen SM: Prehospital clinical findings associated withspinal injury. Prehosp Emerg Care 1:11–15, 1997.



35. Domeier RM, Evans RW, Swor RA, Rivera-Rivera EJ,Fredericksen SM: Prospective validation of out-of-hospital spi-nal clearance criteria: A preliminary report. Acad Emerg Med4:643–646, 1997 (letter).

36. Farrington JD: Death in a ditch. Bull Am Coll Surg 52:121–130,1967.

37. Farrington JD: Extrication of victims: Surgical principles.J Trauma 8:493–512, 1968.

38. Fenstermaker RA: Acute neurologic management of the patientwith spinal cord injury. Urol Clin North Am 20:413–421, 1993.

39. Frohna WJ: Emergency department evaluation and treatment ofthe neck and cervical spine injuries. Emerg Med Clin North Am17:739–791, 1999.

40. Garfin SR, Shackford SR, Marshall LF, Drummond JC: Care ofthe multiply injured patient with cervical spine injury. ClinOrthop 239:19–29, 1989.

41. Garth G: Proposal for the establishment of minimum perfor-mance specifications for cervical extrication collars. Presented atthe 14th Annual Meeting of the American Society for Testing andMaterials, Skeletal Support Committee, West Conshohocken,PA, 1988.

42. Geisler WO, Wynne-Jones M, Jousse AT: Early management ofthe patient with trauma to the spinal cord. Med Serv J Can22:512–523, 1966.

43. Goth P: Spinal Injury: Clinical Criteria for Assessment and Manage-ment. Augusta, Medical Care Development Publishing, 1994.

44. Graziano AF, Scheidel EA, Cline JR, Baer LJ: A radiographiccomparison of pre-hospital cervical immobilization methods.Ann Emerg Med 16:1127–1131, 1987.

45. Green BA, Eismont FJ, O’Heir JT: Spinal cord injury: A systemsapproach—Prevention, emergency medical services and emer-gency room management. Crit Care Clin 3:471–493, 1987.

46. Gunby P: New focus on spinal cord injury. JAMA 245:1201–1206, 1981.

47. Gunn DB, Eizenberg N, Silberstein M, McMeeken JM, Tully EA,Stillman BC, Brown DJ, Gutteridge GA: How should an uncon-scious person with a suspected neck injury be positioned?Prehospital Disaster Med 10:239–244, 1995.

48. Hachen HJ: Emergency transportation in the event of acutespinal cord lesion. Paraplegia 12:33–37, 1974.

49. Hamilton RS, Pons PT: The efficacy and comfort of full-bodyvacuum splints for cervical-spine immobilization. J Emerg Med14:553–559, 1996.

50. Hauswald M, Ong G, Tandberg D, Omar Z: Out-of-hospitalspinal immobilization: Its effect on neurologic injury. AcadEmerg Med 5:214–219, 1998.

51. Jeanneret B, Magerl F, Ward JC: Over distraction: A hazard ofskull traction in the management of acute injuries of the cervicalspine. Arch Orthop Trauma Surg 110:242–245, 1991.

52. Johnson DR, Hauswald M, Stockhoff C: Comparison of a vac-uum splint device to a rigid backboard for spinal immobiliza-tion. Am J Emerg Med 14:369–372, 1996.

53. Jones SL: Spine trauma board. Phys Ther 57:921–922, 1977.54. Kilburn MP, Smith DP, Hadley MN: The initial evaluation and

treatment of the patient with spinal trauma, in Batjer HH, LoftusCM (eds): Textbook of Neurological Surgery: Principles and Practice.Philadelphia, Lippincott Williams & Wilkins (in press).

55. Kolb JC, Summers RL, Galli RL: Cervical collar-induced changesin intracranial pressure. Am J Emerg Med 17:135–137, 1999.

56. Kossuth LC: Removal of injured personnel from wrecked vehi-cles. J Trauma 5:704–705, 1965.

57. Kossuth LC: The initial movement of the injured. Mil Med132:18–21, 1967.

58. Lerner EB, Billittier AJ IV, Moscati RM: The effects of neutralpositioning with and without padding on spinal immobilizationof healthy subjects. Prehosp Emerg Care 2:112–116, 1998.

59. Liew SC, Hill DA: Complication of hard cervical collars in multi-trauma patients. Aust N Z J Surg 64:139–140, 1994.

60. Linares HA, Mawson AR, Suarez E, Biundo JJ: Association be-tween pressure sores and immobilization in the immediate post-injury period. Orthopedics 10:571–573, 1987.

61. Markenson D, Foltin G, Tunik M, Cooper A, Giordano L, FittonA, Lanotte T: The Kendrick extrication device used for pediatricspinal immobilization. Prehosp Emerg Care 3:66–69, 1999.

62. Marshall LF, Knowlton S, Garfin SR, Klauber MR, EisenbergHM, Kopaniky D, Miner ME, Tabbador K, Clifton GL: Deterio-ration following spinal cord injury: A multi-center study.J Neurosurg 66:400–404, 1987.

63. Mawson AR, Biundo JJ Jr, Neville P, Linares HA, Winchester Y,Lopez A: Risk factors for early occurring pressure ulcers follow-ing spinal cord injury. Am J Phys Med Rehabil 67:123–127, 1988.

64. Mazolewski P, Manix TH: The effectiveness of strapping tech-niques in spinal immobilization. Ann Emerg Med 23:1290–1295,1994.

65. McCabe JB, Nolan DJ: Comparison of the effectiveness of differ-ent cervical immobilization collars. Ann Emerg Med 15:50–53,1986.

66. McGuire RA Jr: Protection of the unstable spine during transportand early hospitalization. J Miss State Med Assoc 32:305–308,1991.

67. McGuire RA Jr, Degnan G, Amundson GM: Evaluation of cur-rent extrication orthoses in immobilization of the unstable cer-vical spine. Spine 15:1064–1067, 1990.

68. McGuire RA Jr, Neville S, Green BA, Watts C: Spinal instabilityand the log-rolling maneuver. J Trauma 27:525–531, 1987.

69. McHugh TP, Taylor JP: Unnecessary out-of-hospital use of fullspinal immobilization. Acad Emerg Med 5:278–280, 1998 (letter).

70. McSwain NE Jr: Spine management skills, in Pre-Hospital TraumaLife Support. Akron, Educational Direction, 1990, ed 2, pp 225–256.

71. Meldon SW, Brant TA, Cydulka RK, Collins TE, Shade BR:Out-of-hospital cervical spine clearance: Agreement betweenemergency medical technicians and emergency physicians.J Trauma 45:1058–1061, 1998.

72. Deleted in proof.73. Muhr MD, Seabrook DL, Wittwer LK: Paramedic use of a spinal

injury clearance algorithm reduces spinal immobilization in theout-of-hospital setting. Prehosp Emerg Care 3:1–6, 1999.

74. Nypaver M, Treloar D: Neutral cervical spine positioning inchildren. Ann Emerg Med 23:208–211, 1994.

75. Olson CM, Jastremski MS, Vilogi JP, Madden CM, Beney KM:Stabilization of patients prior to interhospital transfer. Am JEmerg Med 5:33–39, 1987.

76. Orledge JD, Pepe PE: Out-of-hospital spinal immobilization: Is itreally necessary? Acad Emerg Med 5:203–204, 1998.

77. Perry SD, McLellan B, McIlroy WE, Maki BE, Schwartz M, FernieGR: The efficacy of head immobilization techniques during sim-ulated vehicle motion. Spine 24:1839–1844, 1999.

78. Plaisier B, Gabram SG, Schwartz RJ, Jacobs LM: Prospectiveevaluation of craniofacial pressure in four different cervical or-thoses. J Trauma 37:714–720, 1994.

79. Podolsky S, Baraff LJ, Simon RR, Hoffman JR, Larmon B, AblonW: Efficacy of cervical spine immobilization methods. J Trauma23:461–465, 1983.

80. Deleted in proof.



81. Prasad VS, Schwartz A, Bhutani R, Sharkey PW, Schwartz ML:Characteristics of injuries to the cervical spine and spinal cord inpolytrauma patient population: Experience from a regionaltrauma unit. Spinal Cord 37:560–568, 1999.

82. Raphael JH, Chotai R: Effects of the cervical collar on cerebro-spinal fluid pressure. Anaesthesia 49:437–439, 1994.

83. Rimel RW, Jane JA, Edlich RF: An educational training programfor the care at the site of injury of trauma to the central nervoussystem. Resuscitation 9:23–28, 1981.

84. Rodgers JA, Rodgers WB: Marginal mandibular nerve palsy dueto compression by a cervical hard collar. J Orthop Trauma9:177–179, 1995.

85. Roozmon P, Gracovetsky SA, Gouw GJ, Newman N: Examiningmotion in the cervical spine: Part I—Imaging systems and mea-surement techniques. J Biomed Eng 15:5–12, 1993.

86. Deleted in proof.87. Rosen PB, McSwain NE Jr, Arata M, Stahl S, Mercer D: Compar-

ison of two new immobilization collars. Ann Emerg Med 21:1189–1195, 1992.

88. San Mateo County, California: EMS System Policy Memorandum#F-3A. 1991.

89. Schafermeyer RW, Ribbeck BM, Gaskins J, Thomason S, HarlanM, Attkisson A: Respiratory effects of spinal immobilization inchildren. Ann Emerg Med 20:1017–1019, 1991.

90. Schriger DL: Immobilizing the cervical spine in trauma: Shouldwe seek an optimal position or an adequate one? Ann EmergMed 28:351–353, 1996.

91. Schriger DL, Larmon B, LeGassick T, Blinman T: Spinal immo-bilization on a flat backboard: Does it result in neutral position ofthe cervical spine? Ann Emerg Med 20:878–881, 1991.

92. Smith MG, Bourn S, et al.: Ties that bind: Immobilizing theinjured spine. J Emerg Med Serv JEMS 14:28–35, 1989.

93. Stauffer ES: Orthotics for spinal cord injuries. Clin Orthop 102:92–99, 1974.

94. Suter R, Tighe T, et al.: Thoracolumbar spinal instability duringvariations of the log-roll maneuver. Prehospital Disaster Med7:133–138, 1992.

95. Swain A, Dove J, Baker H: ABCs of major trauma: PartI—Trauma of the spine and spinal cord. BMJ 301:34–38, 1990.

96. Toscano J: Prevention of neurological deterioration before ad-mission to a spinal cord injury unit. Paraplegia 26:143–150, 1988.

97. Totten VY, Sugarman DB: Respiratory effects of spinal immobi-lization. Prehosp Emerg Care 3:347–352, 1999.

98. Tuite GF, Veres R, Crockard HA, Peterson D, Hayward RD: Useof an adjustable, transportable, radiolucent spinal immobiliza-tion device in the comprehensive management of cervical spineinstability: Technical note. J Neurosurg 85:1177–1180, 1996.

99. Wagner FC Jr, Johnson RM: Cervical bracing after trauma. MedInstrum 16:287–288, 1982.

100. Walsh M, Grant T, Mickey S: Lung function compromised byspinal immobilization. Ann Emerg Med 19:615–616, 1990 (let-ter).

101. Walton R, DeSalvo JF, Ernst AA, Shahane A: Padded vs unpad-ded spine board for cervical spine immobilization. Acad EmergMed 2:725–728, 1995.

102. Washtenaw/Livingston County Medical Control Authority: Spi-nal Injury Assessment and Immobilization: EMS Protocols. Ann Ar-bor, Washtenaw/Livingston County Medical Control Authority,1997.

103. Waters RL, Meyer PR Jr, Adkins RH, Felton D: Emergency,acute, and surgical management of spine trauma. Arch PhysMed Rehabil 80:1383–1390, 1999.

104. Worsing RA Jr: Principles of pre-hospital care of musculoskeletalinjuries. Emerg Med Clin North Am 2:205–217, 1984.

Sketch of a dissection showing the head falling forward, as hap-pens in some cases of destruction of the ligaments, associatedwith disease of the joints between the atlas and axis and occipi-tal bones. From, Hilton J: On Rest and Pain: A Course of Lec-tures on the Influence of Mechanical and Physiological Rest inthe Treatment of Accidents and Surgical Diseases, and the Diag-nostic Value of Pain. New York, Wood, 1879, 2nd ed.


CHAPTER 2

Transportation of Patients with Acute Traumatic CervicalSpine Injuries

RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS: Expeditious and careful transport of patients with acute cervical spine or spinal cord injuries is

recommended, from the site of injury by the most appropriate mode of transportation available to thenearest capable definitive care medical facility.

RATIONALE

Definitive assessment, resuscitation, and care of thepatient with an acute traumatic cervical spine injurycannot be rendered at the accident scene. Optimal

care for patients with spinal injury includes initial resuscita-tion, immobilization, extrication, and early transport of thepatient to a medical center with the capability for diagnosisand treatment (3–5, 9, 11). Delay in transportation to a defin-itive treatment center is associated with less favorable out-comes, longer hospitalizations, and increased costs (7, 8, 11).

Several modes of transportation are available to transfer thespinal injury patient, primarily land (ambulance) and air (he-licopter or fixed-wing aircraft). Selection of the ideal mode oftransportation for an individual patient depends on the pa-tient’s clinical circumstances, distance, geography, and avail-ability. The goal is to expedite efficient, safe, and effectivetransportation, without unfavorable effects on patient out-come. These issues provide the rationale for this guideline.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of literature published from 1966 to 2001 was per-formed. The search was limited to the English language.Medical subject headings were used in combination with“spinal injury” and “transport.” The first search term (key-word and exploded subject heading) yielded 8,493 articles.The second search term (as keyword) yielded 12,437 articles.A search combining both terms provided 44 articles. The 44abstracts were reviewed, and additional references wereculled from the reference lists of the remaining articles. Fi-nally, members of the author group were asked to contributearticles on the subject matter that were known to them but notfound by other search means. This process yielded 13 articlesthat were directly relevant to the subject of transportationof patients with spinal injuries. All articles provide ClassIII medical evidence. Pertinent articles are summarized inTable 2.1.


Safe, rapid transport of the spine-injured patient to a med-ical facility for definitive care has long been a fundamentaltenet of emergency medical service (EMS) care delivery. Norandomized clinical trials to establish the necessity or effec-tiveness of this strategy have been performed. A search of theliterature does not provide Class I or Class II medical evi-dence in support of this practice.

One of the basic principles of preadmission spinal care isthe early transfer of the injured person to a center with theresources and expertise to manage the patient with an acutecervical spine or spinal cord injury (3–5, 9, 11). Early compli-cations can be prevented and improved neurological out-comes have been reported when early transfer to a specializedspinal cord injury (SCI) center is accomplished (5, 11). Duringtransportation, every effort must be made to limit untowardspinal motion and to preserve neurological function (12).

Several options exist for the transportation of patients to adefinitive care facility. The selection of the mode of transpor-tation is based on the patient’s clinical status and what isreasonable and available to achieve the goals of rapid transfer,while maintaining effective medical support for the patientand proper spinal immobilization for patients at risk.

In 1974, Hachen (5) described the creation of a nationwideemergency transportation protocol for patients with spinalinjury. The protocol had been implemented in Switzerland in1968. All SCI patients in Switzerland were immediately trans-ported to the National Spinal Injuries Center in Geneva by theSwiss Air Rescue Organization. In the 10-year follow-up ofthis protocol published in 1977, Hachen (6) reported that earlytransport from the site of the accident to the SCI center underclose medical supervision was associated with no patientdeath during transport. Before 1968, many deaths occurredduring transport, secondary to acute respiratory failure, be-fore definitive care could be provided. After 1968, patientswere transported rapidly with an onboard anesthetist whoprovided respiratory, cardiac, and hemodynamic monitoring,


resuscitation, and nasotracheal intubation as necessary. Theaverage time for the rescue operation was reduced from 4.5hours to 50 minutes. There was a significant reduction incardiovascular and respiratory morbidity and mortality. Themortality rate for complete quadriplegic patients droppedfrom 32.5% in 1966 to 6.8% in 1976, and for incomplete cer-vical cord-injured patients from 9.9 to 1.4% during the sameperiod. Hachen (5, 6) concluded that survival and outcome forpatients with acute spinal cord injuries was enhanced by awell-organized medical system, rapid medically supervisedtransfer by helicopter to a specialized center, and then defin-itive care in an SCI facility for aggressive management in theintensive care unit (ICU) setting.

Zach et al. (13), in 1976, described their experience with 117acute SCI patients managed according to prospective protocol inthe Swiss Paraplegic Center in Basel, Switzerland. All patientswere treated in the ICU setting with aggressive medical man-agement and cardiac and blood pressure support. Outcomeswere stratified by initial injury and time of admission afterinjury. Of cervical spinal cord injuries managed in this fashion,62% were improved at last follow-up, no patient with a cervicallevel injury worsened, and 38% were unchanged. Of patientswho arrived within 12 hours of injury, 67% improved, comparedwith their initial neurological condition. Of patients admittedbetween 12 and 48 hours of injury, 59% showed neurologicalimprovement. When admission occurred more than 48 hours

after injury, improvement was seen in only 50% of patients. Theauthors concluded that early transport and “immediate medicalspecific treatment of the spinal injury” seemed to facilitate neu-rological recovery.

In 1984, Tator et al. (11) reported their experience with 144patients with acute spinal cord injuries treated between 1974 and1979 at the Acute Spinal Cord Injury Unit (ASCIU) at Sunny-brook Medical Centre in Toronto, Ontario, Canada. They founda marked reduction in both morbidity and mortality after acutespinal cord injury for the group of patients managed during theperiod 1974 to 1979 compared with a similar group managedduring the period 1947 to 1973, which was before the creation ofa dedicated regional spinal cord injury unit. Reasons cited forthese improvements included earlier transport to the ASCIUafter trauma and better definitive management at arrival.

In a subsequent 1993 publication comparing ASCIU pa-tients managed from 1974 to 1981 with the patients managedfrom 1947 to 1973, Tator et al. (10) noted a statistically signif-icant difference in time from injury to arrival, 5 hours forASCIU patients compared with 13 hours for the pre-ASCIUgroup. They found a significant decrease in the severity ofspinal cord injury (65% complete cervical lesions in pre-ASCIU patients compared with 46% for ASCIU patients) andnoted fewer complications, shorter hospital stays, and lowerexpenses for patients managed under the new ASCIU para-digm. Their findings support the advantages of early trans-

TABLE 2.1. Summary of Reports on Transportation of Patients with Spinal Injuriesa

Series (Ref. No.) Description of StudyEvidence

ClassConclusions

Tator et al.,1993 (10)

A study of 201 ASCI patients, ICU care, hemodynamic supportcompared with 351 previous patients.

III Less severe cord injuries due to immobilization, resuscitation, and earlytransfer to ICU setting.

Armitage et al.,1990 (1)

Case reports of 4 patients who developed respiratory problemsduring airplane transport.

III Airplane air is less humid, and measures to optimize humidity andpulmonary function in high cervical injury patients may be required duringtravel.

Boyd et al.,1989 (2)

A prospective cohort study to determine the effectiveness of airtransport for major trauma patients when transferred to a traumacenter from a rural emergency room.

III Patients with severe multiple injury from rural areas fare better withhelicopter EMS than ground EMS.

Burney et al.,1989 (3)

Retrospective review of the means of transport and type ofstabilization used for all patients with ASCI.

III ASCI patients can be safely transported by air or ground using standardprecautions. Distance and extent of associated injury are the bestdeterminants of mode of transport.

Tator et al.,1984 (11)

A retrospective review of results of innovations between 1974 and1979 at Sunnybrook Medical Centre in Toronto.

III Patients transferred to the specialized unit earlier, with consequent markedreduction in complications and cost of care.

Hachen, 1977(6)

A study of 188 ASCI patients managed in ICU, aggressive treatmentof hypotension, respiratory insufficiency.

III Reduced morbidity and mortality with early transfer, attentive ICU care andmonitoring, and aggressive treatment of hypotension and respiratory failure.

Zach et al.,1976 (13)

A study of 117 ASCI patients at Swiss center, ICU setting, aggressivemedical therapy (Rheomacrodexb � 5 d; dexamethasone � 10 d).

III Improved neurological outcome with aggressive medical treatment. Betteroutcome for early referrals.

Hachen, 1974(5)

Retrospective review of effectiveness of emergency transportation ofspinal injury patients in Switzerland. From 1965 to 1974, all ASCIpatients were immediately transported by air to ASCI center.Mortality reduced to zero during transport. Average time for therescue operation reduced from 4.5 h to 50 min. Significantreduction in cardiovascular and respiratory morbidity/mortality.

III Mortality and morbidity of patients with acute spinal injury is reduced by awell-organized medical response with smooth and rapid transfer byhelicopter to a specialized SCI center.

a ASCI, acute spinal cord injury; ICU, intensive care unit; EMS, emergency medical services; SCI, spinal cord injury.b Rheomacrodex, dextran 40 (Medisan, Parsippany, NJ).

Patient Transport in Acute Traumatic Cervical Spine Injuries


port to a regional, specialized SCI center for definitive com-prehensive care of patients with SCI.

Burney et al. (3) reviewed the means of transport and type ofstabilization used for all patients with acute spinal cord injuriestransferred to the University of Michigan Medical Center from1985 to 1988 to determine the effect of these variables on impair-ment and neurological improvement. Sixty-one patients werereviewed. Twenty-five patients were transported by ground am-bulance (41%), 33 by helicopter (54%), and 3 by fixed-wingaircraft (5%). Forty-three patients (70.5%) had cervical spineinjuries, 11 (18%) had thoracic spine injuries, and 7 (11.5%) hadlumbar spine injuries. Fifty-one patients (84%) were transferredwithin 24 hours of injury. A variety of standard methods ofstabilization were used during transport. No patient sustainedan ascending neurological injury as a result of early transport.The level of function improved before discharge in 26 (43%) of 61patients. Patients transported to the medical center within 24hours of injury were more likely to show improvement (25 of 51patients) than those transported after 24 hours (1 of 10 patients).There was no significant difference in the probability of im-provement between ground transport (8 of 25 patients) or airtransport (18 of 36 patients). The authors concluded that acuteSCI patients could be safely transported, with standard precau-tions, by air or ground. They found that distance and the extentof the patient’s associated injuries were the best determinants ofthe mode of transport.

Rural areas reportedly account for 70% of fatal accidents, andrural mortality rates for people involved in motor vehicle acci-dents are four to five times higher than in urban areas. A pro-spective cohort study by Boyd et al. (2) examined the effective-ness of air transport of major trauma patients when transferredto a trauma center from a rural emergency room. The studyconsisted of 872 consecutive trauma patients admitted afterlong-distance transfer. The authors found that air transport wasassociated with a 25.4% reduction in predicted mortality (Z �3.95; P � 0.001). The benefit of helicopter EMS transport wasrealized only in major trauma patients with a probability ofsurvival of less than 90%. Thus, the benefits identified with earlyhelicopter EMS transport were directly related to injury severity.It is unclear whether these findings can be extrapolated to pa-tients with spine and/or spinal cord injuries, because the authorsdid not stratify injuries by body systems in their report.

Neither land nor air transport has been reported in theliterature to adversely affect the outcome for spine-injuredpatients when properly executed. One note of caution wasoffered by Armitage et al. (1). They described four spine-injured patients who developed respiratory distress/failureduring airplane transport. They noted that because patientswith cervical spinal cord injuries may have severely reducedpulmonary performance, measures to optimize oxygenation,humidification, and pulmonary function should be under-taken for these patients, particularly during air transport.

SUMMARY

The patient with an acute cervical spine or spinal cordinjury should be expeditiously and carefully transported from

the site of injury to the nearest capable definitive care medicalfacility. The mode of transportation chosen should be basedon the patient’s clinical circumstances, distance from the tar-get facility, and geography to be traveled, and should be themost rapid means available. Patients with cervical spinal cordinjuries have a high incidence of airway compromise andpulmonary dysfunction; therefore, respiratory support mea-sures should be available during transport. Several studiessuggest that rates of morbidity and mortality of SCI patientsdecreased after the advent of sophisticated transport systemsto dedicated SCI centers. These studies all provide Class IIImedical evidence on this issue.


Development and refinement of transportation protocolsfor patients with cervical spine and spinal cord injury shouldbe undertaken and could be accomplished by using a large,prospectively collected data set. From these data, case-controlor comparative cohort studies could be structured to generateClass II evidence.


REFERENCES

1. Armitage JM, Pyne A, Williams SJ, Frankel H: Respiratory prob-lems of air travel in patients with spinal cord injuries. BMJ300:1498–1499, 1990.

2. Boyd CR, Corse KM, Campbell RC: Emergency interhospitaltransport of the major trauma patient: Air versus ground.J Trauma 29:789–794, 1989.


4. Guttman L: Initial treatment of traumatic paraplegia and tetra-plegia, in Spinal Injuries: Proceedings of the Symposium held in theRoyal College of Surgeons of Edinburgh, June 7–8, 1963. Edinburgh,The Royal College of Surgeons, 1967.

5. Hachen HJ: Emergency transportation in the event of acute spinalcord lesion. Paraplegia 12:33–37, 1974.

6. Hachen HJ: Idealized care of the acute injured spinal cord inSwitzerland. J Trauma 17:931–936, 1977.

7. Neville S, Watts C: Management of the unstable cervical spine intransport: A re-evaluation. Aeromed J Sept/Oct: 32, 1987.

8. Rutledge G, Sumchai A: A safe method for transportation ofpatients with cervical spine injuries. Aeromed J Sept/Oct: 33,1987.

9. Stover S, Fine PR: Spinal Cord Injury: The Facts and Figures. Bir-mingham, University of Alabama at Birmingham, 1986, p 45.

10. Tator CH, Duncan EG, Edmonds VE, Lapczak LI, Andrews DF:Changes in epidemiology of acute spinal cord injury from 1947 to1981. Surg Neurol 40:207–215, 1993.

11. Tator CH, Rowed DW, Schwartz ML, Gertzbein SD, Bharatwal N,Barkin M, Edmonds VE: Management of acute spinal cord inju-ries. Can J Surg 27:289–294, 1984.

12. Toscano J: Prevention of neurological deterioration before admis-sion to a spinal cord injury unit. Paraplegia 26:143–150, 1988.

13. Zach GA, Seiler W, Dollfus P: Treatment results of spinal cord injuryin the Swiss Paraplegic Centre. Paraplegia 14:58–65, 1976.



CHAPTER 3

Clinical Assessment after Acute Cervical Spinal Cord Injury

RECOMMENDATIONSNEUROLOGICAL EXAMINATION:Standards: There is insufficient evidence to support neurological examination standards.Guidelines: There is insufficient evidence to support neurological examination guidelines.Options: The American Spinal Injury Association international standards for neurological and functional

classification of spinal cord injury are recommended as the preferred neurological examination tool forclinicians involved in the assessment and care of patients with acute spinal cord injuries.

FUNCTIONAL OUTCOME ASSESSMENT:Standards: There is insufficient evidence to support functional outcome assessment standards.Guidelines: The Functional Independence Measure is recommended as the functional outcome assessment tool

for clinicians involved in the assessment and care of patients with acute spinal cord injuries.Options: The modified Barthel index is recommended as a functional outcome assessment tool for clinicians

involved in the assessment and care of patients with acute spinal cord injuries.

RATIONALE

Acute traumatic spinal cord injury (SCI) affects 12,000 to14,000 people in North America each year. The func-tional consequences of an acute SCI (ASCI) are vari-

able; therefore, the initial clinical presentation of patients withASCI is a key factor in determining triage and therapy and inpredicting prognosis. Consistent and reproducible neurolog-ical assessment scales are necessary to define the acutelyinjured patient’s neurological deficits and to facilitate com-munication with caregivers regarding patient status. Prognos-tic information provided by comparing people with injurieswith the historical outcomes of patients with similar injuries isof value to patients and families. The evaluation of newtherapies proposed for the treatment of ASCI requires the useof accurate, reproducible neurological assessment scales andreliable functional outcome measurement tools, not only tomeasure potential improvement after therapy, but to deter-mine its functional significance. For these reasons, the clinicalneurological assessment and the determination of functionalabilities are important aspects of the care of patients withASCI. The purpose of this review of the medical literature isto determine which neurological assessment scales and whichfunctional impairment tools are the most useful in the care ofpatients with ASCI.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasperformed. The search was limited to the English languageand human studies. The terms “spinal cord injury” or “spinal

injury” combined with the term “classification or assessment”yielded 17,923 references. A second search with the terms“scale” or “weights and measures” or “index” or “abstractingand indexing” combined with the terms “spinal cord injury”or “spinal injury” yielded 337 references. These 337 referencesand the 17,923 references from the broader search were im-ported into a database, and duplicates were eliminated. Arti-cles germane to this topic were selected by reviewing theirtitles and abstracts. Additional references were culled fromthe reference lists of the remaining articles. Finally, membersof the author group were asked to contribute articles knownto them on the subject matter that were not found by othersearch means.

A total of 53 articles were accessed, reviewed, graded, andincluded in this review. There is no Class I medical evidencein the literature on this topic. There are two Class II compar-ative analyses of functional outcome scales. Twenty-sevenpertinent articles are summarized in Tables 3.1 and 3.2.


A variety of assessment systems are available for docu-menting neurological status of patients after ASCI. They in-clude the Frankel scale, the modified Frankel scale, Lucas andDucker’s neurotrauma motor index, the Sunnybrook scale, theBotsford scale, the Yale scale, the National Acute Spinal CordInjury Study (NASCIS) scale, the American Spinal Injury As-sociation (ASIA) scale, and the ASIA/International MedicalSociety of Paraplegia (IMSOP) international standards forneurological and functional classification of SCI scale (1–4,6–8, 11–13, 15, 16, 28, 37, 42, 47).


Several of these assessment scales have been refinedthrough serial iterations (1–4, 28, 41, 42, 47). A few are widelyused, but others have not attained general acceptance and

recognition. Ideally, the clinical neurological assessment ofASCI patients should be uniform, reproducible, and thor-ough, but easy to use. The assessment tool must be detailed

TABLE 3.1. Summary of Reports on Neurological Examination Scalesa

Series (Ref. No.) Description of Study Evidence Class Conclusions

Jonsson et al., 2000(33)

A study of the interrater reliability of the ASIA ISCSCI-92.Physicians and physiotherapists classified 23 patientsaccording to the ISCSCI-92 and calculated � values.

III Indicates a weak interrater reliability for scoring incomplete ASCIlesions using the 1992 ASIA standards.

Cohen et al., 1998(18)

A test of the ASIA ISCSCI-92. Participants completed apretest and posttest in which they classified 2 patients whohad an ASCI.

III Further revision of the ASIA 1992 standards and more training wasneeded to ensure accurate classification of SCI.

El Masry et al.,1996 (26)

A study to assess the reliability of the ASIA and NASCISmotor scores. The motor scores of 62 consecutive ASCIpatients were retrospectively reviewed.

III The differences in correlation coefficients between the ASIA motorscore and the NASCIS motor score were not statisticallysignificant. The ASIA and NASCIS motor scores can both be usedfor the neurological quantification of motor deficit and motorrecovery.

Wells and Nicosia,1995 (51)

A comparison of the Frankel scale, Yale scale, MIS, MBI,and FIM in 35 consecutive ASCI patients.

III The best assessment tool is a combination of two scales, onebased on neurological impairment and one on functionaldisability.

Waters et al., 1994(50)

An assessment of strength using motor scores derived fromASIA compared with motor scores based onbiomechanical aspects of walking in predicting ambulatoryperformance in 36 ASCI patients.

III The ASIA scoring system compared favorably with thebiomechanical scoring system. ASIA motor score stronglycorrelates with walking ability.

Davis et al., 1993(19)

A prospective study of 665 ASCI patients to determine thereliability of the Frankel and Sunnybrook scales.

III Demonstrated high interrater reliability of Frankel and Sunnybrookscales. Both scales correspond to total sensory and motor functionbut are insensitive to walking and bladder function.

Bednarczyk andSanderson, 1993(5)

A study comparing ASIA scale, NACIS scale, and BB(wheelchair basketball) Sports Test in 30 ASCI patientsclassified by the same examiner.

III ASIA scale showed the greatest discrimination in grouping subjectswith ASCI. NASCIS scale had negative correlation with ASIA scaleand BB sports test.

Botsford and Esses,1992 (7)

Description of a new functionally oriented scale withassessment of motor and sensory function, rectal tone, andbladder function.

III Botsford scale was sensitive for detecting improvement in functionover time after ASCI.

Priebe and Waring,1991 (44)

A study of the interobserver reliability of the 1989 revisedASIA standards assessed by quiz given to 15 physicians.

III The interobserver reliability for the revised ASIA (1989) standardswere improved compared with previous versions, but less thanoptimal. Changes were recommended.

Bracken et al.,1990 (10)

Multicenter North American trial examining effects ofmethylprednisolone or naloxone in ASCI (NASCIS II).

III forneurologicalassessment

Motor scores of 14 muscles on 5-point scale, right side of bodyonly. Sensory scores of pinprick and light touch, 3-point scale,bilateral. No interrater reliability comparison.

Lazar et al., 1989(36)

A prospective study of the relationship between earlymotor status and functional outcome after ASCI in 78patients. Motor status was measured by the ASIA MIS, andfunctional status was evaluated with the MBI.

III The MIS correlated well with functional status for quadriplegicpatients, poorly for paraplegic patients. Individual differences inambulation limit its predictive usefulness.


Multicenter North American trial examining effects ofmethylprednisolone in ASCI (NASCIS I).

III forneurologicalassessment

Motor scores of 14 muscles on 6-point scale. Right side of bodyonly. Sensory scores of pinprick and light touch, 3-point scale,bilateral. No interrater reliability comparison.

Tator et al., 1982(47)

Initial description of the Sunnybrook scale, a 10-gradenumerical neurological assessment scale.

III Improvement from the Frankel scale. Motor grading subdividedbut not very sensitive.

Chehrazi et al.,1981 (15)

Initial description of the Yale scale and its use in a groupof 37 patients with ASCI.

III Provides assessment of the severity of ASCI.

Lucas and Ducker,1979 (37)

Initial description of a motor classification of patients withSCI and its use in 800 patients.

III Allows the clinical researcher to evaluate current treatments andassess the potential of new treatment regimes.


Description of 133 ASCI patients classified using motorand sensory scales developed by Yale Spinal Cord InjuryStudy Group.

III Considerable discrepancy between motor and sensory impairmentscales among patients with greater motor than sensory loss.

Frankel et al., 1969(28)

The first clinical study of the Frankel scale to assessneurological recovery in 682 patients treated with posturalreduction of spinal fractures.

III First neurological examination scale for ASCI.

a ASIA, American Spinal Injury Association; ISCSCI, international standards for neurological and functional classification of incomplete spinalcord injury; SCI, spinal cord injury; ASCI, acute SCI; NASCIS, National Acute Spinal Cord Injury Study; MBI, modified Barthel index; MIS, motorindex score; FIM, functional independence measurement.



and precise to specifically document a given patient’s injuryand must provide descriptive measurement scales that allowdetermination of loss or gain of function with time and ther-apy. Finally, there must be measurement of the patient’sfunctional abilities relative to the neurological examination todocument whether losses or gains have meaningful signifi-cance to the patient and to accurately determine outcome.Whatever assessment system is used, it must have interob-server reliability. Difficulties exist when clinicians use poorlydefined measurement tools or different methods of neurolog-ical assessment to describe the same patient, hindering thedefinition (and potentially the management) of that patient bydifferent clinicians and the comparison of that patient withother patients with similar injuries. The accurate assessmentof both the neurological status and the functional skills ofASCI patients is essential for patient management, the con-duct of research studies, and comparisons of clinical thera-peutic trials.

Numerous assessment scales have been used to evaluatepatients with SCIs. Scales may be divided into two generaltypes. The first type is examination-specific and focuses on theneurological deficits that are a result of SCI. These scales usethe motor and sensory examination primarily (or exclusively)to assign a numerical value or letter grade (1, 2, 6–12, 15, 28,37, 42, 47). The second type of scale focuses on functionalskills, including a patient’s ability to care for himself or her-self, participate in personal hygiene, transfer, or ambulate (3,4, 14, 22–24, 26, 29, 34, 35, 38, 41, 45). In general, the first typeof scale is used for the acute assessment of patients with SCI,

whereas both assessment scales are used to define the chron-ically injured patient. More contemporary assessment scalesincorporate both neurological examination scores and func-tional outcome scores in their overall definition of individualpatients (3, 4, 51).

Neurological examination scales

Frankel et al. (28) provided the first report of a stratifiedneurological scale used to characterize patients with ASCI in1969. The authors used a five-grade scale, A to E, to defineSCIs in 682 patients managed at the Stoke Mandeville Hospi-tal between 1951 and 1968. Grade A patients had completemotor and sensory lesions. Grade B patients had sensory-onlyfunction below the level of injury. Grade C patients had motorand sensory function below the level of injury, but their motorfunction was useless. Grade D patients had motor functionuseful, but not normal function below the level of SCI. GradeE patients had recovery with no motor, sensory, or sphincterdisturbance. The Frankel scale, as it became known, waswidely adopted for use in the description of SCI patients andin assessment of their therapy (outcome) in the 1970s and1980s (1, 2, 28, 29, 42). It was easy to use, was based solely onmotor and sensory function, and required very little patientassessment before classification into one of five grades. How-ever, differentiation between patients classified into Grades Cand D was imprecise. These were broad injury groups withconsiderable range within each injury grade. The sensitivity ofthe Frankel scale to change in serial measurements, particu-

TABLE 3.2. Summary of Reports on Functional Outcome Scalesa

Series (Ref. No.) Description of Study Data Class Conclusions

Field-Fote et al., 2001 (27) SCI-FAI offered as functional assessment scale forgait assessment.

III Reliable and relatively sensitive measure of walking ability inpatients with ASCI. Interrater reliability good. No � values offered.

Kucukdeveci et al., 2000(35)

To determine the reliability and validity of the MBIin Turkey.

III Adaptation of the MBI successful in Turkey as long as its limitationsare recognized. � values �0.5.

Ditunno et al., 2000 (23) WISCI offered as index for ambulation skills after SCIin pilot study.

III Good reliability, excellent interrater reliability but needs assessmentin clinical settings.

Yavuz et al., 1998 (53) Assessment of the relationship of two functionaltests, FIM and QIF, to ASIA scores.

III Strong correlation between FIM and QIF to ASIA scores.

Catz et al., 1997 (14) SCIM offered as new disability scale for spinal cordlesions. Thirty patients assessed with SCIM and FIM.

III SCIM more sensitive than FIM.

Hamilton et al., 1994 (31) Assessment of interrater agreement of FIM in 1018patients in 89 UDS hospitals.

II � values for 7-level FIM ranged from 0.53 to 0.66. � values higherin subset of UDS hospitals with experienced rehabilitationclinicians, 0.69–0.84.

Dodds et al., 1993 (25) Assessment of reliability of FIM in characterizing11,102 UDS rehabilitation patients.

III FIM has high internal consistency and adequate discriminativecapabilities and was good indicator of burden of care.

Hamilton et al., 1991 (32) Interrater agreement assessment of FIM in 263patients in 21 UDS hospitals.

II � values for 7-level FIM ranged from 0.61 to 0.76; mean, 0.71.

Shah et al., 1989 (45) Description of MBI. III The MBI has greater sensitivity and improved reliability than theoriginal version, without additional difficulty or implementationtime.

Gresham et al., 1986 (29) Assessment of QIF as functional scale, comparedwith Barthel index.

III The QIF was more sensitive and reliable than the Barthel index.

a SCI, spinal cord injury; ASCI, acute SCI; FAI, functional ambulatory inventory; MBI, modified Barthel index; WISCI, walking index for spinalcord injury; FIM, functional independence measurement; QIF, quadriplegic index of function; ASIA, American Spinal Injury Association; SCIM,spinal cord independence measure; USD, Uniform Data System.

Clinical Assessment after Acute Cervical Spinal Cord Injury S23


larly among patients in Grades C and D was poor. Significantimprovement in patient function could occur over time with-out the patient advancing a Frankel grade (1, 2, 7). Modifica-tions of the Frankel scale were offered in an attempt to solvethis shortcoming, but the use of the Frankel scale as either anacute assessment tool or an outcome measure has been largelyabandoned because of its lack of sensitivity (3, 4, 6, 7, 13, 21,41, 47–49, 51).

Institutions and investigators have proposed a variety ofneurological assessment scales for SCI patients over the years(1–4, 6–12, 15, 28, 37, 42, 47). In 1978, Bracken et al. (12)described the SCI severity scale developed at the Yale Uni-versity School of Medicine. This scale combines motor andsensory function in selected muscle groups and dermatomes.Its primary focus is to distinguish between complete andincomplete SCIs. The sensory severity scale ranged from onepoint to seven points, and the motor severity scale rangedfrom one point to five points. The authors reported a strongcorrelation between the two scales and noted that “changescores” in motor and sensory function correlated with out-come at discharge compared with at admission. Their pro-posed assessment scales did not assess bowel or bladderfunction and were complicated by grouping all patients to oneof five possible motor scales and into one of seven possiblesensory scales. This scale was difficult to memorize and hardto apply at the bedside.

In 1979, Lucas and Ducker (37) at the Maryland Institute forEmergency Medical Services developed a scoring system forpatients with ASCI. Their scale was based on motor functionat and below the level of injury (Lucas and Ducker’s neuro-trauma motor index) and was used to evaluate more than 800patients collected by the Nationwide Spinal Cord Injury Reg-istry. The scale was later modified for a prospective study ofSCI treatment regimes used at the Maryland Institute forEmergency Medical Services. The authors chose 14 musclesfor examination and used mathematical analysis to predict amotor outcome score based on the initial motor examinationand an empirically derived understanding of the recoveryrate of individual injury subtypes. The scoring system waslimited in that many patients were excluded from the analysis(only 436 of more than 800 patients were analyzed), the stan-dard error of the predicted recovery score was large, and thecalculations were cumbersome. Their scoring system waslater modified by ASIA into a motor index score (24, 41).

In the early 1980s, three different SCI neurological assess-ment scales were introduced, none of which gained popularacceptance (15, 34, 47). In 1980, Klose et al. (34) described theUniversity of Miami Neuro-spinal Index, which consisted oftwo subscales, one motor and one sensory. The motor scalewas scored on a scale of 0 to 5 points for 44 muscle groups,resulting in a possible range of scores from 0 to 220 points.Sensory scoring was a three-point scale for pinprick and vi-bratory sensation in 30 dermatomes on each side of the body.Initial interobserver reliability was high among three physicaltherapists who examined 10 neurologically stable patients inthe rehabilitation setting. Further studies were planned todetermine the efficacy of the University of Miami Neuro-spinal Index in the acute setting and as an outcome tool. The

Yale Scale was reported in 1981 by Chehrazi et al. (15) at theYale New Haven Medical Center. The scale used the BritishMedical Research Council’s gradation (0 to 5 points) of musclestrength by using 10 selected muscle groups from each side ofthe body. Sensory function was scored on a two-point scalefor superficial pain, position sense, and deep pain. Bladderand bowel functions were not scored. In 1982, the SunnybrookCord Injury Scales for assessing neurological injury and re-covery from SCI were proposed by Tator et al. (47). A 10-pointnumerical neurological assessment scale was offered, whichrepresented an improvement on the Frankel scale in howsensory losses were classified. However, like the Frankelscale, motor grading was not very sensitive. The differentia-tion between Grades 3 to 5 and Grades 6 to 8, correspondingto Frankel Grades C and D, remained relatively imprecise andresulted in large, heterogeneous groups of patients. Bladderand bowel functions were not assessed.

In 1984, ASIA (1) generated standards for the neurologicalclassification of spinal injury patients. The neurological as-sessment used a 10-muscle group motor index score (scale of0 to 5 points) and incorporated the Frankel classification as thefunctional abilities assessment tool. The sensory examinationwas not scored, but the most cephalad level of normal sensa-tion was noted. These standards were revised in 1989 toprovide better, more specific sensory level determinations (2).In 1991, Priebe and Waring (44) examined the interobserverreliability of the revised ASIA standards (1989 version). Pa-tient examples, in quiz format, were given to house staff andfaculty of a department of physical medicine and rehabilita-tion. Respondents were asked to classify each patient withrespect to sensory level, motor level, zone of injury, andFrankel classification according to the 1984 ASIA standards.Two months later, the respondents were asked to completeanother quiz by using the 1989 ASIA standards. Although thepercentage of correct answers improved by using the 1989ASIA standards, the authors conclude that interobserver reli-ability was “less than optimal” with a � coefficient of 0.67,indicating agreement between observers but only within therange of fair agreement.

Botsford and Esses (7) introduced a new functionally ori-ented neurological grading system that incorporated motorand sensory function, rectal tone, and bladder control. Themotor assessment score, on a scale of 0 to 5 points, assessedflexor and extensor groups at major joints (hence a “morefunctional” motor assessment). Sensory function was gradedon a 10-point scale (0 to 10 points) divided into five categories.Voluntary rectal contraction was scored on a 10-point scale (0,5, or 10 points). Bladder function was divided into “normal”and “not normal” and assigned 5 points. The authors appliedtheir proposed grading system to a historical group of pa-tients who had initially been assessed and classified accordingto the Frankel scale. They concluded that the new gradingsystem was more sensitive for detecting improvement in theneurological examination and in functional performance overtime.

Two national ASCI studies (NASCIS I and II) were accom-plished in the late 1980s and early 1990s in examination ofmethylprednisolone as a treatment for patients with ASCI



(8–11). Investigators at multiple centers in North Americaused a motor assessment scale (NASCIS scale) that evaluatedmotor function in 14 muscle groups graded from 1 to 6 points(NASCIS I) (8, 11), or 0 to 5 points (NASCIS II) (9, 10). Scoresfor the right and left sides of the body were obtained inde-pendently. Sensory function was divided into pinprick andlight touch in dermatomes C2 through S5 and scored on ascale of 0 to 3 points. Functional abilities were not assessed inthe NASCIS I and II studies. Like most other neurologicalassessment scoring systems before the NASCIS scales, therewas no documentation of interobserver reliability, despite thelarge numbers of patients examined and entered into theNASCIS I and II trials.

In 1992, ASIA (3) generated new standards for neurologicaland functional classification of SCI in conjunction withIMSOP. These standards replaced the revised 1989 version.The new assessment recommendations included motor indexscores, sensory examination scores (scale of 0 to 2 points), andthe ASIA impairment scale (modified Frankel classification),and incorporated the Functional Independence Measure(FIM). FIM is a functional assessment tool and is used toassess the effect of SCI on the patient’s functional abilities. Itquantifies the extent of individual disability and complementsthe neurological assessment by providing scoring for activi-ties of eating, grooming, bathing, dressing upper body, dress-ing lower body, and toileting (20–22, 24). Improvements inneurological function over time or with treatment (as docu-mented by neurological examination scales) can be measuredin terms of functional or meaningful improvement to thepatients with the addition of FIM in the assessment battery.

Davis et al. (19) measured the interobserver reliability of theFrankel classification and the Sunnybrook scale by experi-enced personnel who were provided with concise definitions.The authors demonstrated high interobserver reliability of theFrankel classification and Sunnybrook scales (Pearson corre-lation coefficients, 0.71–0.91), with 94 to 100% intraobserveragreement. Values of the � statistic were not provided. Theauthors concluded that both assessment systems corre-sponded well to total sensory and motor function in SCIpatients but were insensitive to ambulation skills and bladderfunction.

In 1993, Bednarczyk and Sanderson (5) reported on theability of three different classification systems to describe SCIpatients and to compare the correlation between the threescales when applied by a single trained observer. They com-pared the NASCIS scale with the ASIA scale and the BB(wheelchair basketball) Sports Test. The authors found thatthe ASIA scale had the greatest discrimination in groupingsubjects with SCI into mixed-injury categories and intoincomplete-injury categories. The BB Sports Test had a posi-tive correlation with the ASIA scale (Spearman rank correla-tion coefficient, 0.81). The NASCIS scale had a negative cor-relation with both the ASIA scale (Spearman coefficient,�0.66) and the BB Sports Test (Spearman coefficient, �0.48).In contrast, El Masry et al. (26) retrospectively assessed 62consecutive ASCI patients and compared ASIA and NASCISmotor scores with conventional motor examinations. Theseauthors found that both motor assessment scales were repre-

sentative of the conventional motor scores reported for thesepatients and could be used to quantify motor deficits andrecovery after ASCI.

Lazar et al. (36), in 1989, evaluated the relationship betweenearly motor status and functional outcome after SCI prospec-tively in 52 quadriplegic and 26 paraplegic patients. Motorstatus was measured within 72 hours of injury and quantifiedwith the ASIA motor index score. Functional status was eval-uated with the modified Barthel index (MBI). A senior phys-ical therapist completed ASIA motor index score and MBIassessments on each patient at admission to the spinal cordintensive care unit and every 30 days during rehabilitation.The authors found that early motor function correlated wellwith average daily improvement in functional status, includ-ing self-care and mobility (P � 0.001). The initial motor indexscore strongly correlated with the functional status of quad-riplegic patients at admission (P � 0.001), at 60 days (P �0.001), and at rehabilitation discharge (P � 0.001), but hadpoor correlation in paraplegic patients. The ASIA motor indexscore correlated significantly with the MBI self-care subscoreat 60 days and at discharge (P � 0.01), but not with the MBImobility subscore. Lazar et al. concluded that the ASIA motorindex score is a useful tool in predicting function duringrehabilitation, although individual differences in ambulation,particularly for patients with paraplegia, limit the predictiveusefulness of this index.

Waters et al. (50), in 1994, compared the strength of 36 acuteSCI patients as determined by the ASIA motor score withmotor scores based on biomechanical aspects of walking topredict ambulatory performance. The authors found that theASIA scoring system compared favorably with the biome-chanical scoring system and was a relatively simple clinicalmeasure that correlated strongly with walking ability. In 1995,Marino et al. (40) compared the ASIA motor level and theupper extremity motor score (UEMS) with the neurologicallevel of injury in 50 quadriplegic patients. At 12 months afterinjury, quadriplegia index of function (QIF) assessments wereobtained. Spearman rank correlations were calculated. Theauthors found that the motor level was more highly correlatedwith the UEMS and the QIF than with the neurological levelof injury. The UEMS had the highest correlation to the QIFfeeding score, 0.78. Marino et al. concluded that the ASIAmotor level and UEMS better reflect the severity of impair-ment and disability. Similarly, Ota et al. (43) compared theASIA motor scores and neurological level of injury with FIMin 100 Frankel Grade A and B patients. The authors found thatthe motor score reflected the patients’ disability as deter-mined by FIM better than the ASIA level of injury.

Wells and Nicosia (51) compared the usefulness and limi-tations of five different SCI scoring systems applied by asingle skilled observer in the assessment of 35 consecutive SCIpatients: the Frankel classification, the Yale scale, the ASIAmotor index score, the MBI, and the FIM score. The authorsfound that the Frankel classification correlated strongly withthe Yale scale and the ASIA motor index scores, but weaklywith MBI and FIM. These three assessment scales shared afocus on impairment measurement. The MBI and the FIMscore correlated strongly with each other but weakly with the



other scales, and they shared a focus on disability. Wells andNicosia (51) concluded that one classification system or scalealone does not adequately describe SCI patients in both theacute and follow-up settings. They favored a combination oftwo scales to characterize ASCI patients, one based on neu-rological impairment and the other on functional disability.

Jonsson et al. (33) evaluated interobserver reliability of the1992 ASIA standards for neurological and functional classifi-cation of SCI. Two physicians and two physiotherapists clas-sified 23 SCI patients according to the 1992 recommendations.Values for the � statistic for pinprick scores varied from 0 to0.83 (poor to very good), from 0 to 1.0 for light touch scores,and from 0 to 0.89 for motor function. They found weakinterobserver reliability for scoring patients with incompleteSCIs. Cohen et al. (17) performed further tests of reliability ofthe 1992 ASIA standards. The 106 professionals in the fieldof SCI completed tests before and after formal review of the1992 ASIA standards, in which they classified two SCI pa-tients by sensory and motor levels, zone of partial preserva-tion, ASIA impairment scale, and completeness of injury.Percentage “correct” was calculated for each tested criterion.The authors reported that participants had very little diffi-culty in correctly classifying the patient with a complete SCIbut had variable success characterizing the patient with anincomplete SCI. They concluded that further refinement of the1992 ASIA standards and more training in their applicationwas required.

In 1996, ASIA/IMSOP (4) provided a revised version of theinternational standards for neurological and functional clas-sification of spinal injury (an update of their 1992 recommen-dations). Further refined by input from numerous interna-tional organizations, the combination of the ASIA impairmentscale, the ASIA motor index score, the ASIA sensory scale,and FIM is considered to be the most representative assess-ment and classification tool for patients with ASCI. It wasconsidered to be an improvement on the 1992 standards,which were subject to criticism (4, 13, 17, 18, 25, 33, 41, 43, 44).

Functional outcome scales

Functional outcome scales are nonspecific measures of hu-man performance ability relevant to medical rehabilitation,that is, how a person functions with activities of everyday life.Several scales have been developed in an effort to accuratelycharacterize an injured person’s functional skills and disabil-ities to quantify his or her functional independence (3, 4, 14,21–23, 25, 27, 29, 30, 34, 35, 38, 39, 41, 43, 45, 46). These scalesattempt to determine a patient’s ability or inability to liveindependently. Scales for functional rating include the Barthelindex, the MBI, the FIM, the QIF, the spinal cord indepen-dence measure (SCIM), the walking index for SCI (WISCI),and the SCI functional ambulation inventory (SCI-FAI) (3, 4,14, 21–23, 25, 27, 29, 30, 34, 35, 38, 39, 41, 43, 45, 46). Thesescales are applicable to a wide range of nervous system dis-orders, but the QIF, the SCI-FAI, and the SCIM are morespecific for patients with SCI (14, 27, 29). All of these scaleshave been successfully used to characterize SCI patients (3, 4,14, 21–23, 25, 27, 29, 30, 34, 35, 38, 39, 41, 43, 45, 46).

Among many available functional assessment scales, theBarthel index has been one of the most popular (35, 38, 52). Ithas been used for both characterizing individual patients andin evaluating the efficacy of various rehabilitation programs.The Barthel index consists of 10 ratable patient skill items.Values are assigned to each item (0, 5, or 10 points) based onthe amount of physical assistance required to perform eachtask. A Barthel index total score ranges from 0 to 100 points (0,fully dependent; 100, fully independent). In the original ver-sion, each item is scored in three steps (38). The MBI, with afive-step scoring system, seems to have greater sensitivity andimproved reliability than the original version, without exam-ination difficulty or an increase in implementation time. Shahet al. (45) found the internal consistency reliability coefficientfor the MBI to be 0.90, compared with 0.87 for the originalindex. In another study, Kucukdeveci et al. (35) evaluated thereliability and validity of the MBI in 50 inpatient rehabilitationSCI patients in Turkey. Patients were assessed by the MBI atadmission and at discharge. Reliability was tested by usinginternal consistency, interobserver reliability, and the intra-class correlation coefficient. Construct validity was assessedby association with impairments (ASIA) and by Rasch analy-sis. The internal consistency coefficient was 0.88. The level ofagreement between two observers was sufficient (� statistic,0.5). The intraclass correlation coefficient was 0.77. However,Rasch analysis revealed that bladder and bowel items of theMBI misfit the construct. The authors concluded that adapta-tion of the MBI is useful in assessing SCI patients in Turkey,as long as its limitations are recognized.

The FIM was developed to provide uniform assessment ofthe severity of patient disability and medical rehabilitationoutcome (20–22, 24). It is an 18-item, seven-level scale de-signed to assess the severity of patient disability, estimateburden of care, and determine medical rehabilitation func-tional outcome. The FIM has emerged as a standard assess-ment instrument for use in rehabilitation programs for dis-abled persons (4, 20–22, 24, 25, 30–32, 41, 43, 51, 53). In 1993,Dodds et al. (25) assessed FIM with respect to validity andreliability in characterizing 11,102 general rehabilitation pa-tients in the Uniform Data System from the Pacific Northwest.They compared admission and discharge FIM scores andassessed for validation by using several hypotheses. The au-thors found high overall internal consistency and that FIMidentified significant functional gains in patients over time.FIM discriminated patients on the basis of age, comorbidity,and discharge destination. The authors concluded that FIMhad high internal consistency and adequate discriminativecapabilities, and it was a good indicator of burden of care.

Hamilton et al. (31, 32) assessed interobserver agreement ofthe seven-level FIM in two separate reports. In the 1991 report(32), two or more pairs of clinicians assessed each of 263patients undergoing inpatient medical rehabilitation at 21hospitals in the United States subscribing to the Uniform DataSystem for medical rehabilitation. Criteria were intraclass cor-relation coefficient (ICC) (analysis of variance) for total FIM,and FIM subscores 0.90 or higher (5 of 6 subscores must be�0.90; no ICC could be �0.75). Values of the � statistic (un-weighted) for individual FIM items must be 0.45 or higher for



at least 15 of the 18 items. The total FIM ICC was 0.97.Subscore ICCs were: self-care, 0.96; sphincter control, 0.94;mobility (transfers), 0.96; locomotion, 0.93; communications,0.95; and social cognition, 0.94. The FIM item � mean was 0.71(range, 0.61–0.76). On the basis of their methods of analysis,the authors concluded that the seven-level FIM has goodclinical interobserver agreement (32). In 1994, Hamilton et al.(31) reported interobserver reliability among clinicians from89 rehabilitation facilities in the United States within theUniform Data System; 1018 patients were characterized. Byanalysis methods similar to those reported above, total FIMICC was 0.96. Subscore ICCs ranged from 0.89 to 0.96. FIMitem � coefficients ranged from 0.53 to 0.66. For a subset ofinstitutions that met Uniform Data System reliability criteria,values of the � statistic ranged from 0.69 to 0.84. The authorsconcluded that FIM is reliable when used by trained andtested inpatient medical rehabilitation clinicians. Stineman etal. (46) used FIM and a function-based strategy to generatefunctional outcome benchmarks among 3604 SCI patients.They found that most patients whose motor-FIM scores atadmission were above 30 were able to groom, dress the upperextremity, manage bladder function, use a wheelchair, andtransfer from bed to chair by the time of discharge fromrehabilitation. Most patients with scores above 52 attainedindependence in all but the most difficult FIM tasks, such asbathing, tub transfers, and stair climbing. The authors con-cluded that these “FIM item attainment benchmarks” may beuseful in counseling patients, predicting outcome, and antic-ipating patient care needs after discharge.

The QIF was developed in 1980 because the Barthel indexwas deemed to be too insensitive to document the small butsignificant functional gains made by quadriplegics (tetraple-gics) during medical rehabilitation (29, 53). The QIF is com-prised of variables that are each weighted and scored (trans-fers, grooming, bathing, feeding, dressing, wheelchairmobility, bed activities, bladder and bowel program, andunderstanding of personal care). A final score ranging from 0to 100 points is derived that characterizes each patient’s func-tional abilities and serves as a reference for future assessment.Gresham et al. (29) tested the QIF on a group of 30 completequadriplegic patients at admission to and discharge frominpatient medical rehabilitation. Resultant scores were com-pared with those simultaneously obtained by the Barthel in-dex and the Kenny self-care evaluation. The QIF was found tobe more sensitive for patient functional improvement (46%)than that defined by the Barthel index (20%) or the Kennyself-care evaluation (30%). The QIF was also tested for reli-ability. Ratings by three different nurses, working indepen-dently, were found to be significantly positively correlated forall subscores (P � 0.001). Gresham et al. (29) concluded thatthe QIF provides a useful option in choosing a functionalassessment instrument for quadriplegic patients.

Yavuz et al. (53) compared ASIA scores, the QIF, and theFIM in 29 subjects with cervical SCI. The same examiner usedall three scales at admission to and discharge from the reha-bilitation center. The authors identified strong correlation ofASIA scores to both the FIM and the QIF. Feeding and dress-ing categories of QIF showed an even stronger correlation to

ASIA motor scores; however, statistical significance was thesame for corresponding categories of FIM and QIF. The per-centage of recovery on ASIA motor scores was significantlycorrelated only to gain in QIF scores, not in FIM scores. Theauthors recommended that additions to the FIM may be use-ful to improve sensitivity, particularly in the feeding, dress-ing, and bed activity categories.

Catz et al. (14) developed a new disability scale specific forpatients with spinal cord pathology, SCIM, and compared itwith FIM in the assessment and characterization of 30 pa-tients. Two pairs of trained staff members recorded scores 1week after admission and every month thereafter during hos-pitalization. The authors found remarkable consistency be-tween each pair of observers for all tasks assessed (� coeffi-cient, 0.66–0.98). The authors found the SCIM to be moresensitive than FIM to changes in function of spinal cord lesionpatients: SCIM detected all functional changes detected byFIM, but FIM missed 26% of changes detected by SCIM scor-ing. The authors concluded that SCIM may be a useful instru-ment for assessing functional changes in patients with lesionsof the spinal cord.

The WISCI was proposed as a scale to measure functionallimitations in walking of patients after SCI (23). This scaleincorporates gradations of physical assistance and devicesrequired for walking after paralysis of the lower extremitiessecondary to SCI. The purpose of the WISCI is to documentchanges in functional capacity with respect to ambulation in arehabilitation setting. A pilot study of the WISCI was com-pleted using video clips of patients walking. Raters at eightinternational centers completed the assessment. The concor-dance for the pilot data was significant. Interobserver reliabil-ity revealed 100% agreement. The authors conclude that theWISCI scale showed good validity and reliability but neededfurther evaluation before it could serve as a useful tool forclinical studies (23).

Finally, the SCI-FAI is a functional observational gait-assessment instrument developed at the University of Miamithat addresses three key domains of walking function inindividuals with SCI: gait parameters/symmetry, assistivedevice use, and temporal-distance measures (27). The authorsassessed its validity and reliability in a study of 22 patientswith incomplete SCIs examined by four trained raters. Inter-observer reliability was good for all four raters (ICC range,0.850–0.960). A moderate correlation (Pearson correlation co-efficient, 0.58) was found between change in gait score andlower extremity strength. The authors concluded that theSCI-FAI is a reliable, valid, and relatively sensitive measure ofwalking ability in individuals with SCI.

SUMMARY

A variety of injury classification schemes have been used todescribe patients who have sustained SCIs. There are twogeneral types of assessment scales: neurological examinationscales and functional outcome scales. The most accurate andmeaningful description of SCIs, in the acute setting and infollow-up, seems to be accomplished by using a neurologicalscale in conjunction with a functional outcome scale. At



present, the most used and studied neurological assessmentscales are the ASIA scores, including the motor index scores,sensory scores, and the ASIA impairment scale. After manyrevisions and several refinements, these scales are easy toapply and are reliable.

The 1996 ASIA recommendations for international stan-dards of neurological and functional classification of SCI in-clude the ASIA scales, as noted, and the FIM, which has beenstudied extensively as a functional outcome tool. It seems tobe the best functional outcome scale used to describe disabil-ity among SCI patients, both early and late after injury. It iseasy to administer and is valid and reliable. Interobserveragreement with the FIM has been high in several studies, withreported values of the � statistic of 0.53 to 0.76.


Any future investigation of or clinical trial involving SCIpatients must include both a neurological examination scaleand a functional outcome assessment. Therapeutic trials ofSCI patients should include reliable neurological and func-tional scoring systems and should verify the validity andinterobserver reliability of those scoring scales as part of theinvestigational paradigm.


REFERENCES

1. American Spinal Injury Association. Standards for NeurologicalClassification of Spinal Injury Patients. Chicago, American SpinalInjury Association, 1984.

2. American Spinal Injury Association. Standards for NeurologicalClassification of Spinal Injury Patients. Chicago, American SpinalInjury Association, 1989.

3. American Spinal Injury Association/International Medical Soci-ety of Paraplegia: Standards for Neurological and Functional Classi-fication of Spinal Cord Injury, Revised 1992. Chicago, AmericanSpinal Injury Association, 1992.

4. American Spinal Injury Association/International Medical Soci-ety of Paraplegia: International Standards for Neurological and Func-tional Classification of Spinal Cord Injury, Revised 1996. Chicago,American Spinal Injury Association, 1996.

5. Bednarczyk JH, Sanderson DJ: Comparison of functional andmedical assessment in the classification of persons with spinalcord injury. J Rehabil Res Dev 30:405–411, 1993.

6. Benzel EC, Larson SJ: Functional recovery after decompressivespine operation for cervical spine fractures. Neurosurgery 20:742–746, 1987.

7. Botsford DJ, Esses SI: A new scale for the clinical assessment ofspinal cord function. Orthopedics 15:1309–1313, 1992.

8. Bracken MB, Collins WF, Freeman DF, Shepard MJ, Wagner FW,Silten RM, Hellenbrand KG, Ransohoff J, Hunt WE, Perot PL Jr,Grossman RG, Green BA, Eisenberg HM, Rifkinson N, GoodmanJH, Meagher JN, Fischer B, Clifton GL, Flamm ES, Rawe SE:Efficacy of methylprednisolone in acute spinal cord injury. JAMA251:45–52, 1984.

9. Bracken MB, Shepard MJ, Collins WF, Holford TR, Baskin DS,Eisenberg HM, Flamm E, Leo-Summers L, Maroon JC, MarshallLF, Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC Jr,Wilberger JL, Winn HR, Young W: Methylprednisolone or nalox-one treatment after acute spinal cord injury: 1-year follow-updata—Results of the Second National Acute Spinal Cord InjuryStudy. J Neurosurg 76:23–31, 1992.

10. Bracken MB, Shepard MJ, Collins WF, Holford TR, Young W,Baskin DS, Eisenberg HM, Flamm E, Leo-Summers L, Maroon J,Marshall LF, Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC,Wilberger JE, Winn HR: A randomized, controlled trial of meth-ylprednisolone or naloxone in the treatment of acute spinal-cordinjury: Results of the Second National Acute Spinal Cord InjuryStudy. N Engl J Med 322:1405–1411, 1990.

11. Bracken MB, Shepard MJ, Hellenbrand KG, Collins WF, Leo LS,Freeman DF, Wagner FC, Flamm ES, Eisenberg HM, GoodmanJH, Perot PL Jr, Green BA, Grossman RG, Meagher JN, Young W,Fischer B, Clifton GL, Hunt WE, Rifkinson N: Methylpred-nisolone and neurological function 1 year after spinal cord injury:Results of the National Acute Spinal Cord Injury Study.J Neurosurg 63:704–713, 1985.

12. Bracken MB, Webb SB Jr, Wagner FC: Classification of the sever-ity of acute spinal cord injury: Implications for management.Paraplegia 15:319–326, 1978.

13. Capaul M, Zollinger H, Satz N, Dietz V, Lehmann D, Schurch B:Analyses of 94 consecutive spinal cord injury patients using ASIAdefinition and modified Frankel score classification. Paraplegia32:583–587, 1994.

14. Catz A, Itzkovich M, Agranov E, Ring H, Tamir A: SCIM: SpinalCord Independence Measure—A new disability scale for patientswith spinal cord lesions. Spinal Cord 35:850–856, 1997.

15. Chehrazi B, Wagner FC Jr, Collins WF Jr, Freeman DH Jr: A scalefor evaluation of spinal cord injury. J Neurosurg 54:310–315, 1981.

16. Cheshire DJ: A classification of the functional end-results of injuryto the cervical spinal cord. Paraplegia 8:70–73, 1970.

17. Cohen ME, Ditunno JF Jr, Donovan WH, Maynard FW Jr: A test ofthe 1992 International Standards for Neurological and FunctionalClassification of Spinal Cord Injury. Spinal Cord 36:554–560, 1998.

18. Cohen ME, Sheehan TP, Herbison GJ: Content validity and reli-ability of the International Standards for Neurological Classifica-tion of Spinal Cord Injury. Top Spinal Cord Inj Rehabil 4:15–31,1996.

19. Davis LA, Warren SA, Reid DC, Oberle K, Saboe LA, Grace MG:Incomplete neural deficits in thoracolumbar and lumbar spinefractures: Reliability of Frankel and Sunnybrook scales. Spine18:257–263, 1993.

20. Ditunno JF Jr: Functional assessment measures in CNS trauma.J Neurotrauma 9[Suppl 1]:S301–S305, 1992.

21. Ditunno JF Jr: New spinal cord injury standards, 1992. Paraplegia30:90–91, 1992.

22. Ditunno JF Jr: American spinal injury standards for neurologicaland functional classification of spinal cord injury: Past, presentand future—1992 Heiner Sell Lecture of the American SpinalInjury Association. J Am Paraplegia Soc 17:7–11, 1994.

23. Ditunno JF Jr, Ditunno PL, Graziani V, Scivoletto G, Bernardi M,Castellano V, Marchetti M, Barbeau H, Frankel HL, D’AndreaGreve JM, Ko H-Y, Marshall R, Nance P: Walking Index for SpinalCord Injury (WISCI): An international multicenter validity andreliability study. Spinal Cord 38:234–243, 2000.

24. Ditunno JF Jr, Young W, Donovan WH, Creasey G: The Interna-tional Standards Booklet for Neurological and Functional Classi-fication of Spinal Cord Injury: American Spinal Injury Associa-tion. Paraplegia 32:70–80, 1994.



25. Dodds TA, Martin DP, Stolov WC, Deyo RA: A validation of thefunctional independence measurement and its performanceamong rehabilitation inpatients. Arch Phys Med Rehabil 74:531–536, 1993.

26. El Masry WS, Tsubo M, Katoh S, El Miligui YH, Khan A: Valida-tion of the American Spinal Injury Association (ASIA) motorscore and the National Acute Spinal Cord Injury Study (NASCIS)motor score. Spine 21:614–619, 1996.

27. Field-Fote EC, Fluet GG, Schafer SD, Schneider EM, Smith R,Downey, Ruhl CD: The Spinal Cord Injury Functional Ambula-tion Inventory (SCI-FAI). J Rehabil Med 33:177–181, 2001.

28. Frankel HL, Hancock DO, Hyslop G, Melzak J, Michaelis LS,Ungar GH, Vernon JD, Walsh JJ: The value of postural reductionin the initial management of closed injuries of the spine withparaplegia and tetraplegia. Paraplegia 7:179–192, 1969.

29. Gresham GE, Labi ML, Dittmar SS, Hicks JT, Joyce SZ, StehlikMA: The Quadriplegia Index of Function (QIF): Sensitivity andreliability demonstrated in a study of thirty quadriplegic patients.Paraplegia 24:38–44, 1986.

30. Hamilton BB, Granger CV, Sherwin FS, Zielezny M, Tashman JS:A Uniform National Data System for Medical Rehabilitation, inFuhrer MJ (ed): Rehabilitation Outcomes: Analysis and Measurement.Baltimore, Paul H. Brookes Publishing Co., 1987, pp 137–147.

31. Hamilton BB, Laughlin JA, Fiedler RC, Granger CV: Interraterreliability of the 7-level Functional Independence Measure (FIM).Scand J Rehabil Med 26:115–119, 1994.

32. Hamilton BB, Laughlin JA, Granger CV, Kayton RM: Interrateragreement of the seven-level Functional Independence Measure(FIM). Arch Phys Med Rehabil 72:790, 1991.

33. Jonsson M, Tollback A, Gonzalez H, Borg J: Inter-rater reliabilityof the 1992 international standards for neurological and func-tional classification of incomplete spinal cord injury. Spinal Cord38:675–679, 2000.

34. Klose KJ, Green BA, Smith RS, Adkins RH, MacDonald AM: Universityof Miami Neuro-Spinal Index (UMNI): A quantitative method for de-termining spinal cord function. Paraplegia 18:331–336, 1980.

35. Kucukdeveci AA, Yavuzer G, Tennant A, Suldur N, Sonel B,Arasil T: Adaptation of the modified Barthel Index for use inphysical medicine and rehabilitation in Turkey. Scand J RehabilMed 32:87–92, 2000.

36. Lazar RB, Yarkony GM, Ortolano D, Heinemann AW, Perlow E,Lovell L, Meyer PR: Prediction of functional outcome by motorcapability after spinal cord injury. Arch Phys Med Rehabil 70:819–822, 1989.

37. Lucas JT, Ducker TB: Motor classification of spinal cord injurieswith mobility, morbidity and recovery indices. Am Surg 45:151–158, 1979.

38. Mahoney FI, Barthel DW: Functional evaluation: The BarthelIndex. Md State Med J 14:61–65, 1965.

39. Marino RJ, Huang M, Knight P, Herbison GJ, Ditunno JF Jr, SegalM: Assessing selfcare status in quadriplegia: Comparison of theQuadriplegia Index of Function (QIF) and the Functional Inde-pendence Measure (FIM). Paraplegia 31:225–233, 1993.

40. Marino RJ, Rider-Foster D, Maissel G, Ditunno JF: Superiority ofmotor level over single neurological level in categorizing tetra-plegia. Paraplegia 33:510–513, 1995.

41. Maynard FM Jr, Bracken MB, Creasey G, Ditunno JF Jr, DonovanWH, Ducker TB, Garber SL, Marino RJ, Stover SL, Tator CH, WatersRL, Wilberger JE, Young W: International Standards for Neurologi-cal and Functional Classification of Spinal Cord Injury: AmericanSpinal Injury Association. Spinal Cord 35:266–274, 1997.

42. Maynard FM Jr, Reynolds GG, Fountain S, Wilmot C, Hamilton R:Neurological prognosis after traumatic quadriplegia: Three-yearexperience of California Regional Spinal Cord Injury Care Sys-tem. J Neurosurg 50:611–616, 1979.

43. Ota T, Akaboshi K, Nagata M, Sonoda S, Domen K, Seki M, ChinoN: Functional assessment of patients with spinal cord injury:Measured by the motor score and the Functional IndependenceMeasure. Spinal Cord 34:531–535, 1995.

44. Priebe MM, Waring WP: The interobserver reliability of the re-vised American Spinal Injury Association standards for neurolog-ical classification of spinal injury patients. Am J Phys MedRehabil 70:268–270, 1991.

45. Shah S, Vanclay F, Cooper B: Improving the sensitivity of the BarthelIndex for stroke rehabilitation. J Clin Epidemiol 42:703–709, 1989.

46. Stineman MG, Marino RJ, Deutsch A, Granger CV, Maislin G: Afunctional strategy for classifying patients after traumatic spinalcord injury. Spinal Cord 37:717–725, 1999.

47. Tator CH, Rowed DW, Schwartz ML: Sunnybrook Cord InjuryScales for assessing neurological injury and neurological recov-ery, in Tator CH (ed): Early Management of Acute Spinal CordInjury. New York, Raven Press, 1982, pp 17–24.

48. Toh E, Arima T, Mochida J, Omata M, Matsui S: Functionalevaluation using motor scores after cervical spinal cord injuries.Spinal Cord 36:491–496, 1998.

49. Waters RL, Adkins RH, Yakura JS: Definition of complete spinalcord injury. Paraplegia 9:573–581, 1991.

50. Waters RL, Adkins RH, Yakura JS, Vigil D: Prediction of ambu-latory performance based on motor scores derived from stan-dards of the American Spinal Injury Association. Arch Phys MedRehabil 75:756–760, 1994.

51. Wells JD, Nicosia S: Scoring acute spinal cord injury: A study ofthe utility and limitations of five different grading systems. J Spi-nal Cord Med 18:33–41, 1995.

52. Yarkony GM, Roth E, Lovell L, Heinemann A, Katz RT, Wu Y:Rehabilitation outcomes in complete C5 quadriplegia. Am J PhysMed Rehabil 69:73–76, 1988.

53. Yavuz N, Tezyurek M, Akyuz M: A comparison of two functional testsin quadriplegia: The Quadriplegia Index of Function and the FunctionalIndependence Measure. Spinal Cord 36:832–837, 1998.


CHAPTER 4

Radiographic Assessment of the Cervical Spine inAsymptomatic Trauma Patients

RECOMMENDATIONSSTANDARDS: Radiographic assessment of the cervical spine is not recommended in trauma patients who are

awake, alert, and not intoxicated, who are without neck pain or tenderness, and who do not havesignificant associated injuries that detract from their general evaluation.

RATIONALE

Spinal cord injury (SCI) is a potentially devastating con-sequence of acute trauma and can occur with improperimmobilization of an unstable cervical spine fracture.

Immobilization of an injured patient’s cervical spine aftertrauma is now standard care in most emergency medicalservices (EMS) systems. Immobilization of the cervical spineis maintained until spinal cord or spinal column injury isruled out by clinical assessment and/or radiographic survey.Radiographic study of the cervical spine of every traumapatient is costly and results in significant radiation exposureto a large number of patients, few of whom will have a spinalcolumn injury. The purpose of this review is to define whichradiographic studies are necessary in the assessment of thecervical spine in asymptomatic patients after trauma.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasperformed. MEDLINE subject headings and keywords “spi-nal cord injury,” “spinal fractures,” or “spinal injuries” re-sulted in 7994 matches. Combination with the keyword “cer-vical” resulted in 1844 matches. The search was limited tohuman studies and the English language, resulting in 1268articles. Combination with the keywords “clearance,” “diag-nosis,” or “radiography” yielded 184 matches. The titles andabstracts of these 184 articles were reviewed. All articles fo-cusing on clinical decision-making in the diagnosis of cervicalspine injury in adult patients with trauma were included.Additional references were culled from the reference lists ofthe remaining articles. Finally, members of the author groupwere asked to contribute articles known to them on the subjectmatter that were not found by other search means. The bib-liography developed by the EAST (Eastern Association for theSurgery of Trauma) practice parameter workgroup for cervicalspine clearance was reviewed (15), as was the reference listdeveloped by the NEXUS (National Emergency X-radiographyUtilization Study) group (5, 7).

Nine large, prospective cohort studies were identified;these studies provide Class I evidence. No randomized con-trolled trials in the literature addressed this issue. Manysmaller studies, case series, and retrospective cohort studieswere identified, which provide corroborating Class II andClass III evidence. This guideline was generated from thesearticles. The 13 articles most germane to this issue are sum-marized in Table 4.1.


A missed cervical spine injury can result in devastatingneurological injury. For this reason, radiographic assessmentof the cervical spine is liberally used in patients after acutetrauma. Cervical spine radiographs are relatively inexpensiveand are easy to obtain. Computed tomography (CT) andmagnetic resonance imaging may also be used to evaluate theselected spine in certain circumstances. These studies aremore expensive but remain widely available. Because theoverall incidence of cervical spinal column injury in the gen-eral population of trauma patients is low, many patients areexposed to unnecessary ionizing radiation and may be immo-bilized unnecessarily, sometimes for long periods. For theseconcerns and others (e.g., financial, resource allocation, anduse), the issue of radiographic assessment of asymptomaticpatients after trauma has been raised. A number of investiga-tors have proposed that asymptomatic patients do not requireradiographic assessment of the cervical spine after trauma (2,5–7, 9, 14, 16). Asymptomatic patients after trauma are de-fined as those patients who meet all of the following criteria:

1. Are neurologically normal. These patients must have aGlasgow Coma Scale score of 15 and must not have any ofthe following: a) disorientation to person, place, or time; b)inability to remember three objects at 5 minutes; c) delayedor inappropriate response to external stimuli; or d) anyfocal motor or sensory deficit.

2. Are not intoxicated. Patients should be considered intoxi-cated if they have: a) a recent history of intoxication orintoxicating ingestion; b) evidence of intoxication on clini-cal examination; or c) laboratory evidence for the presence


of drugs that alter the level of alertness, including bloodalcohol levels higher than 0.08 mg/dl.

3. Do not have neck pain or midline tenderness. Midlinetenderness is present if the patient complains of pain onpalpation of the posterior midline neck from the nuchalridge to the first thoracic vertebra.

4. Do not have an associated injury that is distracting to thepatient. Significant distracting injuries have been definedas: a) long bone fractures; b) visceral injuries requiringsurgical consultation; c) large lacerations, degloving, orcrush injuries; d) large burns; and e) any other injury thatmight impair the patient’s ability to participate in a generalphysical, mental, and neurological examination (5).

On the basis of these criteria, approximately one-third of traumapatients evaluated in emergency rooms or trauma centers areasymptomatic (range, 14–58%) (2, 5, 7, 9, 14, 16–18). Avoidance ofradiographic assessment in this patient population will result in adecrease in unnecessary radiation exposure, less patient time im-mobilized and confined in an uncomfortable position, and a signif-icant savings in both cost and resources (13).

Establishing a treatment standard for a therapeutic inter-vention requires the existence of at least one randomizedcontrolled study. However, a treatment standard for the use-fulness of a diagnostic test can be established with evidencederived from well-designed clinical studies that include a“diverse population using a ‘gold standard’ reference test in ablinded evaluation appropriate for the diagnostic applicationsand enabling the assessment of sensitivity, specificity, positiveand negative predictive values, and where applicable, likeli-hood ratios” (19). In evaluating the role of the radiographicassessment of asymptomatic trauma patients, we may con-sider the clinical examination to be a diagnostic test. X-rayimaging studies of the cervical spine may be considered thegold standard in this circumstance, because we are attemptingto ascertain whether the clinical examination can accuratelypredict the results of the radiographic assessment in a givenpopulation of patients. The population in question should berepresentative of the trauma population evaluated at anygiven emergency room or trauma center.

The literature reviewed included nine large cohort studiesthat included a representative trauma population, definedsymptomatic and asymptomatic patients by the criteria listedabove, and reported the incidence of spinal injury in thesegroups of patients as detected by subsequent radiographicassessment alone or by imaging of the cervical spine supple-mented by clinical follow-up (2, 5, 6, 9, 14, 16–18). All ninestudies were judged to provide Class I evidence, allowing theestablishment of a treatment standard. Many case series andretrospective cohort studies exist and provide corroboratingClass II and Class III evidence. These investigations are sum-marized in Table 4.1 and will be briefly discussed below.

The largest study addressing this issue encompassed 34,069patients evaluated at 21 emergency rooms across the United

States (5). All patients were studied with standard three-viewcervical radiography supplemented by CT, magnetic reso-nance imaging, or other studies as needed. Of 1818 patientsfound to have spinal injuries, 576 were considered to beclinically significant. Two patients of the 576 were prospec-tively assigned to the “asymptomatic” group. One patient hada probable injury at C2 that was not treated because thepatient refused treatment. Clinical follow-up of this patientrevealed no sequelae. The second patient had a laminar frac-ture of C6. He subsequently developed paresthesias in thearm and underwent surgery. Taking the worst-case scenario,and assuming that both of these patients were truly asymp-tomatic (the second patient later developed paresthesias), andthe injuries were truly significant (the first patient’s injurywas probably not significant given his subsequent clinicalcourse), the negative predictive value of an asymptomaticexamination was 99.9%. In contrast, the positive predictivevalue of a “symptomatic” examination was 1.9% (5).

Bayless and Ray (2), in 1989, studied a consecutive series of228 patients who had received “significant blunt head injury.”Patients were classified as symptomatic or asymptomatic atadmission to the hospital. All patients were observed for atleast 24 hours in the hospital and were assessed with at leasta three-view cervical spine x-ray series. A chart review wasperformed 2 years after admission, and any subsequent hos-pital visits were noted. Of the 228 patients, 211 were judged tohave adequate three-view cervical spine series. Of these, 122were judged asymptomatic and none had a significant spineinjury (2). Hoffman et al. (6) performed a prospective study of974 consecutive patients with blunt trauma evaluated at auniversity emergency room. All patients underwent at least athree-view cervical spine x-ray series supplemented with CT,oblique views, or flexion/extension views on the basis ofphysician judgment. Quality assurance logs, risk managementrecords, and hospital charts from subsequent admissionswere also reviewed. Of the 974 patients included in the study,353 were judged asymptomatic and none were identified tohave had a cervical spine injury (6).

Kreipke et al. (9) performed a prospective study involving 860consecutive acute trauma patients who arrived at a Level Itrauma center. All patients underwent five-view cervical radiog-raphy supplemented with CT and/or flexion/extension viewswhen required. Of these, 324 patients were judged asymptom-atic and none had a cervical spine injury detected on radio-graphic assessment (9). Neifeld et al. (14) prospectively studied886 trauma patients admitted at an urban emergency room. Allpatients were studied with a five-view cervical spine series. Of241 patients who were asymptomatic, none had a spine injurydetected radiographically.

Roberge et al. (16), in 1988, reported the results of a prospec-tive study involving all patients who received a five-view cervi-cal spine series while in an urban emergency room. Of 467patients studied, 155 were judged to be asymptomatic, and nonewere found to have a spinal injury. Ross et al. (17) prospectively

Radiographic Spinal Assessment in Asymptomatic Trauma Patients


TAB

LE4.

1.Su

mm

ary

ofR

epor

tson

Rad

iogr

aphi

cSp

inal

Ass

essm

ent

ofA

sym

ptom

atic

Trau

ma

Pati

ents

a

Seri

es

(Ref

.N

o.)

Des

crip

tion

ofSt

udy

Evid

ence

Cla

ssC

oncl

usio

ns

Hof

fman

etal

.,20

00(5

)Pr

ospe

ctiv

est

udy

of34

,069

patie

nts.

4309

asym

ptom

atic

.

2ha

d“c

linic

ally

sign

ifica

ntin

juri

es”.

All

patie

nts

radi

ogra

phed

.

Neg

ativ

epr

edic

tive

valu

e:99

.9%

.

Posi

tive

pred

ictiv

eva

lue:

1.9%

.

Not

e:1

of2

“mis

sed

inju

ries

”di

dno

tre

ally

have

a“s

igni

fican

tin

jury

,”as

hew

as

untr

eate

dan

dha

dno

sequ

elae

with

clin

ical

follo

w-u

p.Th

eot

her

patie

ntde

velo

ped

pare

sthe

sias

inhi

sar

man

dw

asfo

und

toha

vea

lam

inar

frac

ture

ofC

6.

IX

-ray

sno

tne

cess

ary

inas

ympt

omat

icpa

tient

s.

Gon

zale

set

al.,

1999

(4)

2176

patie

nts

pros

pect

ivel

yst

udie

dw

ithsc

reen

ing

exam

inat

ion

and

x-ra

ys.

1in

jury

was

dete

cted

bypl

ain

x-ra

ysin

anot

herw

ise

asym

ptom

atic

patie

nt;

how

ever

,pl

ain

x-ra

ys

mis

sed

13in

juri

esov

eral

l.

IPl

ain

film

radi

ogra

phy

does

not

impr

ove

sens

itivi

ty

(com

pare

dw

ithth

eph

ysic

alex

amin

atio

n)fo

rth

ede

tect

ion

ofce

rvic

alsp

ine

inju

ryin

asym

ptom

atic

patie

nts.

Rot

het

al.,

1994

(18)

Pros

pect

ive

stud

yof

682

patie

nts

adm

itted

toem

erge

ncy

depa

rtm

ent

with

trau

ma.

96w

ere

asym

ptom

atic

,no

neha

din

jury

.

Ove

rall

inci

denc

eof

inju

ryw

as2%

.

All

patie

nts

radi

ogra

phed

.

Follo

w-u

pcl

inic

alvi

sit

betw

een

30an

d15

0d

post

inju

ry,

achi

eved

in43

%.

Neg

ativ

epr

edic

tive

valu

eof

asym

ptom

atic

exam

inat

ion:

100%

.

Posi

tive

pred

ictiv

eva

lue

ofsy

mpt

omat

icex

amin

atio

n:2.

7%.

IX

-ray

slik

ely

not

nece

ssar

yin

asym

ptom

atic

patie

nts.

Lind

sey

etal

.,19

93(1

0)16

86pa

tient

sst

udie

dre

tros

pect

ivel

y,59

7pa

tient

sst

udie

dpr

ospe

ctiv

ely.

Ato

tal

of49

patie

nts

with

cerv

ical

spin

ein

juri

esw

ere

iden

tifie

d(o

vera

llin

cide

nce

2.1%

).

No

patie

ntw

ithan

inju

ryw

asas

ympt

omat

ic.

III

The

tota

lnu

mbe

rsof

sym

ptom

atic

and

asym

ptom

atic

patie

nts

are

not

repo

rted

,pr

eclu

ding

the

calc

ulat

ion

ofne

gativ

eor

posi

tive

pred

ictiv

eva

lues

.

Asy

mpt

omat

icpa

tient

sdo

not

requ

ire

radi

ogra

phic

imag

es.

Hof

fman

etal

.,19

92(6

)97

4bl

unt

trau

ma

patie

nts

pros

pect

ivel

yst

udie

d.

Ove

rall

inci

denc

eof

cerv

ical

spin

ein

jury

was

2.8%

.

Of

353

aler

t,as

ympt

omat

icpa

tient

s,no

neha

da

sign

ifica

ntsp

ine

inju

ry.

Follo

w-u

p:ra

diog

raph

sne

gativ

ein

all

353.

Cha

rts,

qual

ityas

sura

nce

logs

,an

dri

skm

anag

emen

tre

cord

sre

view

edw

ith3-

mo

follo

w-u

p.

Neg

ativ

epr

edic

tive

valu

eof

asym

ptom

atic

exam

inat

ion:

100%

.

Posi

tive

pred

ictiv

eva

lue

ofsy

mpt

omat

icex

amin

atio

n:4.

5%.

IA

sym

ptom

atic

patie

nts

dono

tre

quir

ece

rvic

alsp

ine

x-ra

ys.

Ros

set

al.,

1992

(17)

Pros

pect

ive

stud

yof

410

patie

nts

seen

attr

aum

ace

nter

.

196

patie

nts

had

asym

ptom

atic

exam

inat

ion,

none

had

inju

ry.

All

patie

nts

stud

ied

with

plai

nx-

rays

,C

Tus

edas

nece

ssar

y.

Neg

ativ

epr

edic

tive

valu

e:10

0%.

Posi

tive

pred

ictiv

eva

lue:

6.1%

.

IR

adio

grap

hyno

tm

anda

tory

for

asym

ptom

atic

patie

nts.

Mai

npo

int

ofpa

per

was

that

mec

hani

smof

inju

ryis

not

a

valu

able

pred

icto

rof

inju

ry.

McN

amar

aet

al.,

1990

(12)

Ret

rosp

ectiv

ere

view

of28

6pa

tient

sju

dged

tobe

“hig

hri

sk”

bym

echa

nism

ofin

jury

.

178

wer

eas

ympt

omat

ic,

none

had

cerv

ical

spin

ein

jury

.

108

wer

esy

mpt

omat

ic,

5ha

dce

rvic

alsp

ine

inju

ry.

Cha

rtfo

llow

-up

perf

orm

edto

dete

rmin

ein

cide

nce

ofin

jury

.

Neg

ativ

epr

edic

tive

valu

efo

ras

ympt

omat

icex

amin

atio

n:10

0%.

Posi

tive

pred

ictiv

eva

lue

for

sym

ptom

atic

exam

inat

ion:

4.9%

.

III

Man

ypa

tient

sex

clud

edgi

ving

topo

or

docu

men

tatio

n,se

lect

popu

latio

nfo

llow

-up

inad

equa

te(fi

lms

not

done

onev

eryo

ne,

node

laye

d

char

tre

view

).

Cer

vica

lsp

ine

x-ra

ysno

tne

cess

ary

inas

ympt

omat

ic

patie

nts.



TAB

LE4.

1.C

onti

nued

Seri

es

(Ref

.N

o.)

Des

crip

tion

ofSt

udy

Evid

ence

Cla

ssC

oncl

usio

ns

Bay

less

and

Ray

,19

89(2

)Se

ries

of22

8pa

tient

s,21

1w

ithco

mpl

ete

stud

ies.

Ove

rall

inci

denc

eof

sign

ifica

ntsp

inal

inju

ryw

as1.

7%.

Of

122

aler

t,as

ympt

omat

icpa

tient

s,no

neha

da

sign

ifica

ntin

jury

.

Follo

w-u

p:x-

rays

nega

tive

inal

l12

2.

Cha

rts

revi

ewed

for

any

subs

eque

ntre

fera

ble

visi

tsw

ithin

2yr

.

Neg

ativ

epr

edic

tive

valu

eof

asym

ptom

atic

exam

inat

ion:

100%

.

Posi

tive

pred

ictiv

eva

lue

ofsy

mpt

omat

icex

amin

atio

n:3%

.

IA

sym

ptom

atic

patie

nts

dono

tre

quir

ece

rvic

alsp

ine

x-ra

ys.

Kre

ipke

etal

.,19

89(9

)Pr

ospe

ctiv

est

udy

of86

0pa

tient

spr

esen

ting

totr

aum

ace

nter

.

324

asym

ptom

atic

,no

neha

din

jury

.

All

patie

nts

radi

ogra

phed

.

Neg

ativ

epr

edic

tive

valu

eof

asym

ptom

atic

exam

inat

ion:

100%

.

Posi

tive

pred

ictiv

eva

lue

ofsy

mpt

omat

icex

amin

atio

n:4%

.

IX

-ray

sno

tne

cess

ary

inas

ympt

omat

icpa

tient

s.

Mir

vis

etal

.,19

89(1

3)40

8pa

tient

sst

udie

dw

ithst

anda

rdx-

rays

and

CT.

Tota

lpo

pula

tion

seen

was

4135

patie

nts.

241

patie

nts

unde

rwen

tC

Tbe

caus

eof

“sus

pici

ous”

x-ra

ys,

failu

reto

visu

aliz

eex

trem

esof

C-s

pine

,or

for

conf

irm

atio

nof

know

nfr

actu

re.

Of

thes

e24

1,13

8pa

tient

sw

ere

clin

ical

ly

asym

ptom

atic

.

CT

serv

edas

“gol

dst

anda

rd.”

Non

eof

138

patie

nts

had

acl

inic

ally

rele

vant

inju

ry(a

lthou

gh1

had

ano

ndis

plac

edC

7

tran

sver

sepr

oces

sfr

actu

rew

hich

was

trea

ted

with

aco

llar)

.

Neg

ativ

epr

edic

tive

valu

eof

asym

ptom

atic

exam

inat

ion:

99.3

–100

%.

Posi

tive

pred

ictiv

eva

lue

ofsy

mpt

omat

icex

amin

atio

n:12

.6%

.

II

Sele

ctpo

pula

tion

gold

stan

dard

may

befa

lse

end

poin

t

Clin

ical

lyre

leva

ntce

rvic

alsp

ine

inju

ryis

extr

emel

y

unco

mm

onin

asym

ptom

atic

patie

nts.

X-r

ays

may

be

unne

cess

ary.

Nei

feld

etal

.,19

88(1

4)Pr

ospe

ctiv

est

udy

of88

6pa

tient

s24

4as

ympt

omat

icpa

tient

s,no

neha

din

jury

.

All

patie

nts

radi

ogra

phed

.

Neg

ativ

epr

edic

tive

valu

e:10

0%.

Posi

tive

pred

ictiv

eva

lue:

6.2%

.

IA

sym

ptom

atic

patie

nts

dono

tre

quir

ex-

rays

.

Rob

erge

etal

.,19

88(1

6)Pr

ospe

ctiv

est

udy

invo

lvin

g46

7tr

aum

apa

tient

s.

155

asym

ptom

atic

patie

nts

wer

eas

ympt

omat

ic,

none

had

asp

ine

inju

ry.

312

wer

esy

mpt

omat

ic,

8ha

dsp

ine

inju

ries

.

All

patie

nts

“sch

edul

edto

follo

w-u

p”in

surg

ery

clin

ic,

auth

ors

stat

eth

atno

mis

sed

inju

ries

have

been

iden

tifie

d.

Neg

ativ

epr

edic

tive

valu

eof

asym

ptom

atic

exam

inat

ion:

100%

.

Posi

tive

pred

ictiv

eva

lue

ofsy

mpt

omat

icex

amin

atio

n:2.

5%.

IA

sym

ptom

atic

patie

nts

dono

tre

quir

ex-

rays

.

Bac

hulis

etal

.,19

87(1

)18

23of

4941

trau

ma

patie

nts

stud

ied

with

plai

nx-

rays

.

94pa

tient

sfo

und

toha

vein

juri

es.

All

wer

esy

mpt

omat

ic.

No

asym

ptom

atic

patie

ntha

da

radi

ogra

phic

ally

dete

ctab

lein

jury

.

IIIA

sym

ptom

atic

patie

nts

dono

tre

quir

ex-

rays

.

aC

T,co

mpu

ted

tom

ogra

phy;

C-s

pine

,ce

rvic

alsp

ine.

Radiographic Spinal Assessment in Asymptomatic Trauma Patients S33


studied 410 trauma patients admitted to a trauma center in 1992.All patients underwent a three-view cervical spine series sup-plemented as needed with CT, flexion/extension views, fluoros-copy, and radionucleotide bone scans. Of 196 patients judged tobe asymptomatic, none had a cervical spine injury diagnosedwith these imaging modalities. Roth et al. (18), in 1994, prospec-tively studied 682 patients evaluated at a military hospital afterblunt trauma. All patients underwent a three-view cervical spinex-ray assessment. A subsequent chart review revealed no missedinjuries (the hospital was the only military hospital within aradius of 2500 miles), and 45% of patients were successfullycontacted 30 to 150 days after the initial evaluation for additionalclinical follow-up. Of 96 asymptomatic patients, none had acervical spine injury. Recently, Gonzales et al. (4) studied a seriesof 2176 patients evaluated in an emergency room after trauma;1768 were judged to be asymptomatic, but the criteria used wereslightly different from those described above. Three injurieswere later detected in this group of 1768 patients, but two ofthese patients were not truly asymptomatic (one had a sternalfracture, multiple rib fractures, and a splenic hematoma; theother had significant facial fractures), and the third patient’sinjuries were radiographically occult. The third patient’s injurieswere detected by CT and were treated with a collar.

In addition to these studies that provide Class I evidence,other studies have been reported that provide corroboratingClass II and Class III evidence germane to this issue (1, 10, 12,13). For example, Mirvis et al. (13) studied 241 patients with ahistory of blunt trauma who were assessed with cervical spinex-rays supplemented by CT. Aside from a single nondis-placed transverse process fracture of C7 (which was not seenon conventional x-rays), none of the 138 patients deemed to beasymptomatic had a significant spinal injury. McNamara et al.(12) performed a retrospective review of 286 trauma patientsevaluated in an urban emergency room. Of 178 patients char-acterized as asymptomatic, none had a spinal injury detectedwith subsequent radiographic assessment. Bachulis et al. (1)surveyed a prospectively acquired database of 4941 consecu-tive patients evaluated after blunt traumatic injury. Of 1823patients who underwent radiographic assessment of the cer-vical spine, 94 were found to have a spinal injury. All patientswith spinal injuries were symptomatic. No asymptomatic pa-tient had a spinal injury. Lindsey et al. (10) reviewed 2283consecutive trauma patients and determined that no patientwith a spinal injury was asymptomatic.

Case reports exist describing asymptomatic patients whohave harbored potentially unstable spinal injuries (11, 20). Forexample, Woodring and Lee (20) reviewed 216 patients whohad cervical spine injuries diagnosed with CT. They reportedthat 11 of these 216 patients were not reported to be symp-tomatic in the medical record. It is unclear why these 11patients were subjected to computed tomographic evaluationof the cervical spine if they were asymptomatic. These authorsalso reported a 61% false-negative rate with the use of plainx-rays in this population. Woodring and Lee encouraged theliberal use of CT based on the mechanism of injury. Noobjective definitive conclusion can be drawn from this retro-spective case series of a very select patient population. One

must question the usefulness of radiographic assessment ofany patient given a 61% false-negative rate.

Other authors have refuted the contention that “mechanismof injury” is a reliable predictor of cervical spine injury (8, 17).Mace (11) reported the case of a 51-year-old man who wasawake and alert, was neurologically intact, and had no com-plaints of neck pain or other associated distracting injury, butwas found to have an unstable fracture of C2. It is importantto note, however, that the patient had no history of trauma butwas being evaluated for a sore throat. Cervical spine x-rayswere obtained to rule out a peritonsillar abscess. From thesereports, it is clear that potentially unstable spinal injuries mayexist in asymptomatic patients (even those presenting withsore throats). However, these injuries are so rare that they donot appear in even the largest population-based studies. Theexperience of Davis et al. (3) is illustrative. They described theetiology of 34 missed cervical spine injuries in a series of32,117 trauma patients evaluated at a group of emergencyrooms servicing San Diego County. No missed injury oc-curred in an asymptomatic patient in their study.

SUMMARY

Clinical investigations that provide Class I evidence involv-ing nearly 40,000 patients, plus Class II and III evidencestudies involving more than 5000 patients, convincingly dem-onstrate that asymptomatic patients do not require radio-graphic assessment of the cervical spine after trauma. Thecombined negative predictive value of cervical spine x-rayassessment of asymptomatic patients for a significant cervicalspine injury is virtually 100% (2, 4–6, 9, 14, 16–18).

In contrast, the reported incidence of cervical spine injuriesin the symptomatic patient ranged from 1.9 to 6.2% in theseClass I evidence studies. Symptomatic patients require radio-graphic study to rule out the presence of a traumatic cervicalspine injury before the cervical spine immobilization is dis-continued (2, 4–6, 9, 14, 16–18). The type and extent of radio-graphic assessment of symptomatic patients after trauma isreviewed in Chapter 5.


None.


REFERENCES

1. Bachulis BL, Long WB, Hynes GD, Johnson MC: Clinical indica-tions for cervical spine radiographs in the traumatized patient.Am J Surg 153:473–478, 1987.

2. Bayless P, Ray VG: Incidence of cervical spine injuries in associ-ation with blunt head trauma. Am J Emerg Med 7:139–142, 1989.

3. Davis JW, Phreaner DL, Hoyt DB, Mackersie RC: The etiology ofmissed cervical spine injuries. J Trauma 34:342–346, 1993.

4. Gonzales RP, Fried PO, Bukhalo M, Holevar MR, Falimirski ME:Role of clinical examination in screening for blunt cervical spineinjury. J Am Coll Surg 189:152–157, 1999.



5. Hoffman JR, Mower WR, Wolfson AB, Todd KH, Zucker MI:Validity of a set of clinical criteria to rule out injury to the cervicalspine in patients with blunt trauma: National EmergencyX-Radiography Utilization Study Group. N Engl J Med 343:94–99, 2000.

6. Hoffman JR, Schriger DL, Mower W, Luo JS, Zucker M: Low-riskcriteria for cervical spine radiography in blunt trauma: A pro-spective study. Ann Emerg Med 21:1454–1460, 1992.

7. Hoffman JR, Wolfson AB, Todd K, Mower WR: Selective cervicalspine radiography in blunt trauma: Methodology of the NationalEmergency X-Radiography Utilization Study (NEXUS). AnnEmerg Med 32:461–469, 1998.

8. Jacobs LM, Schwartz R: Prospective analysis of acute cervicalspine injury: A methodology to predict injury. Ann Emerg Med15:44–49, 1986.

9. Kreipke DL, Gillespie KR, McCarthy MC, Mail JT, Lappas JC,Broadie TA: Reliability of indications for cervical spine films intrauma patients. J Trauma 29:1438–1439, 1989.

10. Lindsey RW, Diliberti TC, Doherty BJ, Watson AB: Efficacy ofradiographic evaluation of the cervical spine in emergency situ-ations. South Med J 86:1253–1255, 1993.

11. Mace SE: Unstable occult cervical spine fracture. Ann Emerg Med20:1373–1375, 1991.

12. McNamara RM, Heine E, Esposito B: Cervical spine injury andradiography in alert, high risk patients. J Emerg Med 8:177–182, 1990.

13. Mirvis SE, Diaconis JN, Chirico PA, Reiner BI, Joslyn JN, MilitelloP: Protocol driven radiologic evaluation of suspected cervicalspine injury: Efficacy study. Radiology 170:831–834, 1989.

14. Neifeld GL, Keene JG, Hevesy G, Leikin J, Proust A, Thisted RA:Cervical injury in head trauma. J Emerg Med 6:203–207, 1988.

15. Pasquale M, Fabian TC: Practice Management Guidelines forTrauma from the Eastern Association for the Surgery of Trauma.J Trauma 44:945–957, 1998.

16. Roberge RJ, Wears RC, Kelly M, Evans TC, Kenny MA, DaffnerRD, Kremen R, Murray K, Cottington EC: Selective application ofcervical spine radiography in alert victims of blunt trauma: Aprospective study. J Trauma 28:784–788, 1988.

17. Ross SE, O’Malley KF, DeLong WG, Born CT, Schwab CW: Clin-ical predictors of unstable cervical spinal injury in multiply in-jured patients. Injury 23:317–319, 1992.

18. Roth BJ, Martin RR, Foley K, Barcia PJ, Kennedy P: Roentgeno-graphic evaluation of the cervical spine: A selective approach.Arch Surg 129:643–645, 1994.

19. Walters BC: Clinical practice parameter development in neuro-surgery, in Bean JR (ed): Neurosurgery in Transition: The Socioeco-nomic Transformation of Neurological Surgery. Baltimore, Williams& Wilkins, 1998, pp 99–111.

20. Woodring JH, Lee C: Limitations of cervical radiography in theevaluation of acute cervical trauma. J Trauma 34:32–39, 1993.

Frontispiece and plate from La Anatomia del corpo umano composta da M. Giovanni Valverde, Nuouamente Ristampata.Venice, Stamperia de Givnti, 1586. Courtesy, Rare Book Room, Norris Medical Library, Keck School of Medicine, Universityof Southern California, Los Angeles, California.

Radiographic Spinal Assessment in Asymptomatic Trauma Patients S35

CHAPTER 5

Radiographic Assessment of the Cervical Spine in SymptomaticTrauma Patients

RECOMMENDATIONSSTANDARDS: A three-view cervical spine series (anteroposterior, lateral, and odontoid views) is recom-mended for radiographic evaluation of the cervical spine in patients who are symptomatic after traumaticinjury. This should be supplemented with computed tomography (CT) to further define areas that aresuspicious or not well visualized on the plain cervical x-rays.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS:• It is recommended that cervical spine immobilization in awake patients with neck pain or tenderness and

normal cervical spine x-rays (including supplemental CT as necessary) be discontinued after either a) normaland adequate dynamic flexion/extension radiographs, or b) a normal magnetic resonance imaging study isobtained within 48 hours of injury.

• Cervical spine immobilization in obtunded patients with normal cervical spine x-rays (including supple-mental CT as necessary) may be discontinued a) after dynamic flexion/extension studies performed underfluoroscopic guidance, or b) after a normal magnetic resonance imaging study is obtained within 48 hoursof injury, or c) at the discretion of the treating physician.

RATIONALE

Trauma patients who are symptomatic, that is, complain ofneck pain, have cervical spine tenderness, or have symp-toms or signs of a neurological deficit associated with the

cervical spine, and trauma patients who cannot be assessed forsymptoms or signs (those who are unconscious, uncooperative,incoherent, or intoxicated, or who have associated traumaticinjuries that distract from their assessment) require radiographicstudy of the cervical spine before cervical spine immobilizationis discontinued. Many authors have proposed strategies andimaging techniques to accomplish x-ray clearance of the cervicalspine after trauma, particularly in the symptomatic or obtundedpatient. One-, three-, and five-view static cervical spine x-rays,computed tomography (CT), magnetic resonance imaging(MRI), bone scans, flexion/extension radiographs, dynamic flu-oroscopy with or without somatosensory evoked potential mon-itoring, and other studies have all been described as useful fordetermining spinal injury and potential spinal instability aftertraumatic injury (1–9, 11–17, 19–24, 26–28, 30–39, 41–43, 45–54,56, 57, 59–73). The purpose of this review is to determine theoptimal radiographic assessment strategy necessary and suffi-cient to exclude a significant cervical spine injury in the symp-tomatic trauma patient.

SEARCH CRITERIA

A computerized search of the database of the National Libraryof Medicine of the literature published from 1966 to 2001 wasperformed. MEDLINE medical subject headings and keywords

“spinal cord injury,” “spinal fractures,” or “spinal injuries” re-sulted in 7994 matches. Combination with the keyword “ cervi-cal” resulted in 1844 matches. These references were limited tohuman studies and the English language, resulting in 1268 arti-cles. Combination with the keywords “clearance,” “diagnosis,”or “radiography” yielded 184 matches. The titles and abstracts ofthese 184 articles were reviewed. All articles focusing on clinicaldecision-making in diagnosing cervical spine injuries in adultpatients with trauma injuries were included. Additional refer-ences were culled from the reference lists of the remaining arti-cles. The members of the author group were asked to contributearticles known to them on the subject matter that were not foundby other search means. The practice parameters and referencelist developed by the EAST (Eastern Association for the Surgeryof Trauma) (56) practice parameter workgroup for cervical spineclearance was reviewed, as was the reference list developed bythe NEXUS (National Emergency X-radiography UtilizationStudy) group (31, 33). A total of 73 references form the basis forthis guideline.

Twenty-one manuscripts were identified that provided ev-idence germane to the topic of this guideline. Four studiesprovided Class I evidence, six provided Class II evidence, and11 were individual case series and provided Class III evi-dence. These 21 manuscripts are summarized in Table 5.1.


Patients who are asymptomatic with respect to a potentialcervical spinal injury after acute trauma do not require radio-


graphic assessment to rule out a significant injury to thecervical spine (see Chapter 4). Radiographic studies do notincrease the sensitivity or specificity of the clinical examina-tion in this specifically defined population of patients (31, 33).There is, however, a 2 to 6% incidence of significant cervicalspine injury in the symptomatic patient population after acutetrauma (4, 31–33, 42, 53, 61, 62). These patients require radio-graphic assessment to exclude cervical spinal injury before thediscontinuation of cervical spine immobilization. The mostsignificant consequence of premature discontinuation of cer-vical spine immobilization is neurological injury. Prolongedimmobilization, however, is associated with morbidity aswell. Decubitus ulcers, increased cerebrospinal fluid pressure,pain, and pulmonary complications have all been describedwith prolonged immobilization of the cervical spine (18, 44,58). For these reasons, a diagnostic algorithm that is highlysensitive and specific for the occurrence of a significant cer-vical spine injury and that can be applied in an expeditiousfashion is desired.

The single most common cause of missed cervical spineinjury seems to be failure to adequately visualize the region ofinjury. This can be caused by failure to obtain radiographs, orby making judgments on technically suboptimal films. Thisoccurs most commonly at the extremes of the cervical spine,the occiput to C2 and at the C7–T1 levels (17, 25, 59). Davis etal. (17) described 32,117 acute trauma patients. Cervical spineinjuries were missed in 34 symptomatic patients; 23 of these34 symptomatic patients either did not have radiographs orhad inadequate radiographs that did not include the region ofinjury. Eight patients had adequate x-ray studies that weremisread by the treating physician. Only one patient had amissed injury that was undetectable on technically adequatefilms, even after retrospective review. The error in two pa-tients with missed injuries was not described. The reviews byDavis et al. (17) and other investigators (1, 6, 9, 16, 24, 43, 47)confirm that it is uncommon to miss cervical spine injurieswith adequate plain radiographic assessment of the occiputthrough T1.

The most prevalent initial x-ray assessment of the symp-tomatic or obtunded patient is the three-view cervical spineseries. When adequate visualization of the entire cervicalspine is achieved from occiput to T1, the negative predictivevalue of a normal three-view cervical spine series has beenreported to range from 93 to 98% in several Class I studies (1,6, 47), and from 85 to 100% in Class II and III studies (9, 16, 24,43). Although the negative predictive value of the three-viewcervical spine x-ray series is quite high, the sensitivity of thethree-view series is less impressive. The same Class I seriesreferenced above report sensitivity rates for the three-viewcervical spine series of 84, 62.5, and 83%, respectively (1, 6,47). In the best-case clinical scenario, assuming the highestvalues for negative predictive value and sensitivity, approxi-mately 98% of patients with a normal three-view cervicalspine x-ray series will have a truly normal cervical spine.

These same data suggest that the three-view cervical spineseries will also be normal in 15 to 17% of patients who havecervical spine injuries. If we assume a 6% incidence of spinalinjury in a high-risk population (the head-injured multi-trauma patient, for example), then an adequate three-viewcervical spine series alone would be expected to correctlyidentify 5 of 6 spinal injuries in a group of 100 patients, andcorrectly identify 94 of 94 patients without a spinal injury.One patient of the 100 with an injured spine would havecervical radiographs interpreted as normal. The addition ofoblique views (for a five-view series) does not seem to in-crease the overall sensitivity of the examination (24). Obliqueviews may be useful in lieu of a swimmer’s view to visualizeC7–T1 (36). Holliman et al. (34) have questioned the useful-ness of the anteroposterior cervical view, and they argue thatit is not an important addition to the assessment of the acutetrauma patient. The data presented by these authors are ClassIII evidence and have not been verified by others. Severalreports confirm that the lateral x-ray view alone will miss asubstantial portion of cervical spine injuries depicted in athree-view series (14, 26, 65).

To increase the sensitivity of the radiographic assessment ofthe cervical spine in trauma patients, many authors havedescribed experiences with CT and MRI in the acute setting.Several have reported greater sensitivity by using CT to viewareas not well visualized on plain films, typically the cranio-cervical and cervicothoracic junctions, or areas identified assuspicious on plain cervical spine x-rays (6, 9, 24, 48, 67, 68).In a small Class I study of 58 patients, Berne et al. (6) reportedthat helical CT of the entire cervical spine identified all clini-cally significant injuries in a series of patients assessed withplain films, CT, and MRI who were followed clinically forsubsequent events. Two injuries were missed; however, nei-ther required any treatment. The authors report a negativepredictive value of 95% for CT for all spinal injuries and anegative predictive value of 100% for unstable injuries. Otherauthors report 100% sensitivity for the detection of injurieswith CT limited to areas poorly visualized or identified assuspicious on plain films (24, 48, 67, 68). However, all studiescited provide Class II and III evidence, and most were im-paired by a common flaw: they treat CT as the “gold stan-dard” for the detection of injury. Although they suggest thatthe addition of CT increases diagnostic sensitivity, the use ofCT data as the gold standard represents a false end point forthe true variable of clinically relevant spinal injury.

Although the incidence of significant spinal injury with anormal cervical spine series supplemented with CT is ex-tremely low, missed injuries have been reported. Brohi andWilson-Macdonald (11) reported a missed C6–C7 facet dislo-cation in a patient with persistent neck pain who was studiedwith plain films and a CT occiput through C7–T1. Sweeney etal. (66) reported an autopsy series of three patients who diedof traumatic injuries and were found to have spinal injuriesundetected by plain films supplemented with CT through the

Radiographic Spinal Assessment in Symptomatic Trauma Patients


TAB

LE5.

1.Su

mm

ary

ofR

epor

tson

Rad

iogr

aphi

cSp

inal

Ass

essm

ent

ofSy

mpt

omat

icTr

aum

aPa

tien

tsa

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ceC

lass

Con

clus

ions

Ban

itet

al.,

2000

(3)

Com

bine

dre

tros

pect

ive/

pros

pect

ive

stud

y.

4460

patie

nts

eval

uate

d.

2217

thou

ght

tore

quir

ex-

rays

.

6-m

ocl

inic

alfo

llow

-up

and

subs

eque

ntC

T/M

RI

used

as“g

old

stan

dard

”

for

plai

nx-

rays

(aut

hors

clai

mno

mis

sed

inju

ries

,cr

edib

lecl

aim

).

5-vi

ewse

ries

used

inal

lpa

tient

s.

IIIIn

sym

ptom

atic

patie

nts,

sens

itivi

tyof

plai

nx-

rays

was

84%

to88

%.

Inst

itutio

nof

acl

inic

ally

base

dim

agin

gpr

otoc

olre

sulte

din

ade

crea

sein

the

rate

ofm

isse

din

juri

esfr

om4%

to0%

.

Prot

ocol

had

sens

itivi

tyof

100%

and

incl

uded

use

ofde

laye

dex

amin

atio

nof

patie

nts

with

tend

erne

ss/p

ain

with

flexi

on/e

xten

sion

x-ra

ys(fa

lse-

posi

tives

not

give

n).

Ber

neet

al.,

1999

(6)

Pros

pect

ive

stud

yof

sele

ctpo

pula

tion

ofpa

tient

s(u

neva

luab

le,

mul

titra

uma,

havi

ngC

Tdo

nefo

ran

othe

rre

ason

).

58pa

tient

s,al

lun

derw

ent

3-vi

ewx-

ray

seri

esfo

llow

edby

helic

alC

Tof

entir

esp

ine.

“Sus

pici

ous

but

not

diag

nost

ic”

exam

inat

ions

wer

eev

alua

ted

with

MR

I,

flexi

on/e

xten

sion

view

s,or

repe

ated

clin

ical

exam

inat

ion.

I20

/58

(34%

)ha

din

juri

esde

tect

ed.

Plai

nx-

rays

iden

tifie

d12

for

ase

nsiti

vity

of60

%,

posi

tive

pred

ictiv

eva

lue

of10

0%,

nega

tive

pred

ictiv

eva

lue

of85

%.

CT

mis

sed

2in

juri

es(b

oth

“sta

ble”

).

Sens

itivi

ty:

90%

.

Spec

ifici

ty:

100%

.

Posi

tive

pred

ictiv

eva

lue:

100%

.

Neg

ativ

epr

edic

tive

valu

e:95

%.

D’A

lise

etal

.,19

99(1

5)12

1ob

tund

edpa

tient

sw

ithno

rmal

x-ra

ysst

udie

dw

ithM

RI.

CT

used

tost

udy

area

sof

MR

Iab

norm

ality

.

All

patie

nts

with

nega

tive

MR

Iun

derw

ent

flexi

on/e

xten

sion

imag

ing

imm

edia

tely

upon

“cle

aran

ce.”

III31

/121

(26%

)ha

din

juri

esde

tect

edon

MR

I.

90/1

21(7

4.4%

)ha

dno

inju

ryan

dw

ere

clea

red

(ver

ified

with

flexi

on/e

xten

sion

).

8pa

tient

sde

term

ined

toha

vesp

inal

inst

abili

ty(c

linic

al,

CT,

etc.

).

No

flexi

on/e

xten

sion

perf

orm

edon

patie

nts

with

abno

rmal

MR

I.

Can

not

dete

rmin

esi

gnifi

canc

eof

MR

Ifin

ding

sin

23/3

1pa

tient

s.

Aut

hors

indi

cate

that

nega

tive

MR

Ieq

uiva

lent

tone

gativ

efle

xion

/ext

ensi

on.

Kat

zber

get

al.,

1999

(39)

Pros

pect

ive

stud

yof

199

patie

nts

who

unde

rwen

tM

RI

inad

ditio

nto

stan

dard

radi

ogra

phic

stud

y.H

alf

ofpa

tient

sw

ere

sele

cted

beca

use

of

susp

ecte

dhi

ghpr

obab

ility

ofin

jury

.

IIIM

RI

dete

cted

inju

ries

ina

high

erfr

actio

nof

thes

epa

tient

sth

andi

dco

nven

tiona

lx-

rays

and

CT.

Sign

ifica

nce

ofth

ese

inju

ries

?

Gol

dst

anda

rd?

Kle

inet

al.,

1999

(41)

Ret

rosp

ectiv

ere

view

of32

patie

nts

with

75kn

own

spin

efr

actu

res.

Blin

ded

revi

ewof

MR

Isc

ans

byra

diol

ogis

ts.

II

Sele

ctpo

pula

tion

Post

erio

r/an

teri

orel

emen

tin

juri

es:

Sens

itivi

ty:

11.5

%/3

6.7%

.

Spec

ifici

ty:

97.0

%/9

8%.

Posi

tive

pred

ictiv

eva

lue:

83%

/91.

2%.

Neg

ativ

epr

edic

tive

valu

e:46

%/6

4%.

MR

Ino

tgo

odfo

rev

alua

ting

bony

path

olog

y.

Tan

etal

.,19

99(6

7)R

etro

spec

tive

revi

ewof

360

patie

nts

trea

ted

for

blun

tin

jury

who

unde

rwen

t3-

view

C-s

pine

film

ssu

pple

men

ted

with

CT

beca

use

of

nonv

isua

lizat

ion

ofC

7–T1

.

CT

findi

ngs

cons

ider

edgo

ldst

anda

rdfo

rde

tect

ion

offr

actu

re.

III11

inju

ries

dete

cted

byC

Tw

hich

wer

eno

tvi

sibl

eon

plai

nx-

rays

.

Sens

itivi

tyof

inad

equa

tepl

ain

x-ra

ysre

lativ

eto

CT

for

this

purp

ose:

97%

.

Whi

teet

al.,

1999

(72)

31pa

tient

sw

ithkn

own

orsu

spec

ted

spin

ein

jury

eval

uate

dw

ithM

RI.

IIIPr

ever

tebr

alhe

mat

oma

pick

edup

mor

eof

ten

byM

RI

than

bypl

ain

x-ra

ys(2

4/31

vers

us14

/30)

.

Sugg

ests

that

sens

itivi

tyof

plai

nx-

rays

for

prev

erte

bral

hem

atom

ais

66%

.

Aja

niet

al.,

1998

(1)

100

cons

ecut

ive

patie

nts

stud

ied

pros

pect

ivel

y.

All

radi

ogra

phed

(3vi

ews)

.

Follo

w-u

pcl

inic

alex

amin

atio

n,C

T,M

RI,

and

flexi

on/e

xten

sion

view

s

perf

orm

ed.

I1/

6in

juri

esm

isse

dby

x-ra

y(s

ensi

tivity

,84

%),

7/12

x-ra

yab

norm

aliti

esfo

und

tobe

insi

gnifi

cant

.

Posi

tive

pred

ictiv

eva

lue:

45%

.

Neg

ativ

epr

edic

tive

valu

e:98

.9%

.

1m

isse

din

jury

dete

cted

byfle

xion

/ext

ensi

onvi

ews.

Sees

etal

.,19

98(6

4)20

patie

nts

unde

rwen

tbe

dsid

efle

xion

/ext

ensi

onun

der

fluor

osco

pyaf

ter

3-vi

ewC

-spi

nex-

rays

norm

al.

III(fo

rflu

oros

copy

),II

for

3-vi

ewC

-spi

new

ith

fluor

osco

pyas

gold

stan

dard

One

patie

ntfo

und

toha

vesu

blux

atio

n.

No

gold

stan

dard

for

flexi

on/e

xten

sion

orflu

oros

copy

.

Sens

itivi

tyof

plai

nx-

rays

with

flexi

on/e

xten

sion

asgo

ldst

anda

rd:

95%

.



TAB

LE5.

1.C

onti

nued

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ceC

lass

Con

clus

ions

Ben

zel

etal

.,19

96(5

)17

4pa

tient

ssu

spec

ted

ofha

ving

cerv

ical

spin

ein

jury

(equ

ivoc

alpl

ain

x-ra

ys/C

Tor

posi

tive

sym

ptom

s).

Und

erw

ent

MR

I.

CT

scan

sob

tain

edth

roug

har

eaof

inju

ryde

fined

byM

RI.

III36

%(6

2/17

4)ha

dM

RI

evid

ence

ofin

jury

.

61/6

2m

anag

edw

ithim

mob

iliza

tion

for

1–2

mo.

All

patie

nts

with

nega

tive

MR

Isc

ans

wer

ecl

eare

d,no

inst

ance

sof

late

inst

abili

ty.

Neg

ativ

epr

edic

tive

valu

eof

MR

I:10

0%.

Posi

tive

pred

ictiv

eva

lue?

Spec

ifici

ty?

Dav

iset

al.,

1995

(16)

116

patie

nts

with

GC

S�

13an

dno

rmal

x-ra

ysev

alua

ted

with

flexi

on/

exte

nsio

nvi

ews

unde

rflu

oros

copy

.

Ifo

rpl

ain

x-ra

ysve

rsus

flexi

on/e

xten

sion

asgo

ld

stan

dard

,III

(follo

w-u

p

ques

tiona

ble)

for

flexi

on/

exte

nsio

nru

ling

out

inju

ry.

113

patie

nts

had

noab

norm

ality

dete

cted

.

2pa

tient

sha

d“s

tabl

e”fa

cet

frac

ture

s.

1pa

tient

had

2m

mof

subl

uxat

ion

and

was

trea

ted

ina

colla

r(n

ofo

llow

-up

onth

ispa

tient

).

No

patie

ntha

dre

fera

ble

neur

olog

ical

inju

ryw

ithcl

inic

alfo

llow

-up.

Dec

ubitu

sul

cers

foun

dun

der

colla

rsin

44%

ofpa

tient

sw

ithm

ean

colla

rtim

eof

6d.

Neg

ativ

epr

edic

tive

valu

eof

flexi

on/e

xten

sion

fluor

osco

py:

100%

.

Hol

liman

,19

91(3

4)R

etro

spec

tive

seri

esof

148

patie

nts

with

know

nsp

ine

inju

ries

.

Late

ral

and

odon

toid

x-ra

ysre

tros

pect

ivel

yre

view

edse

para

tely

from

ante

ropo

ster

ior

film

.60

sets

offil

mav

aila

ble

for

revi

ew.

IIIIn

thes

e60

patie

nts,

all

inju

ries

note

don

ante

ropo

ster

ior

film

sw

ere

also

dete

ctab

leon

late

ral

orod

onto

idfil

ms.

Tehr

anza

deh

etal

.,19

94

(68)

Ret

rosp

ectiv

ere

view

of10

0pa

tient

sw

ithbl

unt

inju

ryan

dno

nvis

ualiz

ed

C7–

T1on

plai

nx-

rays

.

CT

findi

ngs

cons

ider

edgo

ldst

anda

rd.

III3

patie

nts

foun

dto

have

inju

ries

onC

Tno

tvi

sual

ized

bypl

ain

x-ra

ys.

Sens

itivi

tyof

inad

equa

tepl

ain

x-ra

ys:

97%

.

Bor

ock

etal

.,19

91(9

)U

sed

CT

toev

alua

tece

rvic

alsp

ine

in17

9pa

tient

sw

how

ere

sym

ptom

atic

with

norm

alx-

rays

(2),

who

seen

tire

cerv

ical

spin

eco

uld

not

bevi

sual

ized

(123

),or

who

had

equi

voca

l(1

3)or

abno

rmal

(41)

x-ra

ys.

Plai

nfil

mse

nsiti

vity

calc

ulat

edus

ing

CT

asgo

ldst

anda

rd;

auth

ors

clai

m

nom

isse

din

juri

es.

II

Poss

ible

fals

een

dpo

int

39/5

4x-

ray

abno

rmal

ities

wer

eve

rifie

dw

ithC

T(p

ositi

vepr

edic

tive

valu

e:72

%).

X-r

ays

mis

sed

both

inju

ries

insy

mpt

omat

icpa

tient

san

d1

C7

tran

sver

sepr

oces

sfr

actu

re

(neg

ativ

epr

edic

tive

valu

e:97

.6%

).

Coh

net

al.,

1991

(14)

60pa

tient

spr

ospe

ctiv

ely

stud

ied

with

late

ral

x-ra

ysin

emer

genc

y

depa

rtm

ent.

Full

radi

ogra

phic

wor

k-up

(3-

or5-

view

)fo

llow

ed.

Res

ults

ofla

tera

lvi

ewto

full

seri

esco

mpa

red.

II

Poss

ible

fals

een

dpo

int

(Iif

used

only

asco

mpa

riso

n)

Late

ral

view

mis

sed

3/7

tota

lin

juri

es.

Late

ral

view

posi

tive

pred

ictiv

eva

lue:

100%

.

Neg

ativ

epr

edic

tive

valu

e:94

%.

Sens

itivi

ty:

57%

.

Lew

iset

al.,

1991

(43)

Ret

rosp

ectiv

ere

view

of14

1pa

tient

sw

ithac

tive

flexi

on/e

xten

sion

x-ra

ys

perf

orm

edaf

ter

3-vi

ewse

ries

was

norm

al.

IIfo

rpl

ain

x-ra

ysve

rsus

flexi

on/e

xten

sion

asgo

ld

stan

dard

,II

for

flexi

on/

exte

nsio

nvi

ews

11/1

41fle

xion

/ext

ensi

onse

tsre

adas

unst

able

,4

ofw

hom

had

norm

al3-

view

seri

es(r

emai

nder

wer

equ

estio

nabl

e).

All

patie

nts

with

inst

abili

tyha

dpa

inor

wer

ein

toxi

cate

d.

1fa

lse-

nega

tive

flexi

on/e

xten

sion

stud

y.

For

plai

nfil

ms

vers

usfle

xion

/ext

ensi

onan

dpl

ain

film

s:

Sens

itivi

ty:

71%

and

99%

.

Spec

ifici

ty:

89%

and

89%

.

Neg

ativ

epr

edic

tive

valu

e:93

%an

d93

%.

Posi

tive

pred

ictiv

eva

lue:

67%

and

99%

.

Mac

dona

ldet

al.,

1990

(47)

775

patie

nts.

3vi

ews

com

pare

dag

ains

tgo

ldst

anda

rdof

all

othe

rst

udie

spe

rfor

med

and

clin

ical

outc

ome.

I3-

view

seri

es:

Sens

itivi

ty:

83%

.

Spec

ifici

ty:

97%

.

Posi

tive

pred

ictiv

eva

lue:

81%

.

Neg

ativ

epr

edic

tive

valu

e:98

%.

Radiographic Spinal Assessment in Symptomatic Trauma Patients S39


region of injury. Thin-cut computed tomographic imagesthrough the entire spine may increase sensitivity somewhat(6, 55), but no direct comparison between the two imagingstrategies in an appropriate patient population has been per-formed to date.

MRI has been used to evaluate patients at risk for acutespinal injury. Results have been mixed. Benzel et al. (5) stud-ied 174 symptomatic patients with low-field MRI within 48hours of injury. Soft tissue abnormalities were visualized onMRI scans in 62 patients. Two of these 62 were considered tohave unstable injuries. Both had plain film and CT abnormal-ities that revealed the injuries. The 60 patients with MRIabnormalities not thought to be significant were immobilizedfor 1 to 3 months and then studied with flexion/extensionradiographs. Not one was found to have an unstable injury.Patients with “negative” MRI studies were cleared of spinalprecautions, and no adverse events were reported. D’Alise etal. (15) reported their results of a Class III evidence study ofMRI in 121 obtunded patients. Ninety patients had normalstudies and were cleared. Follow-up flexion/extension radio-graphs did not reveal a single abnormality in this group.Thirty-one patients had injuries to soft tissues of the cervicalspine identified by MRI not detected by plain radiographs.Eight of these patients ultimately required surgery. Katzberget al. (39) and White et al. (72) have also described increasedsensitivity of MRI for the detection of soft tissue injuries of thecervical spine after trauma.

These studies demonstrate that MRI abnormalities are vi-sualized in a substantial number of cervical spine studiesperformed on patients after trauma. It is impossible to deter-mine the true incidence of clinically significant ligamentousinjury in this group examined with MRI because all patientswith MRI abnormalities were treated with immobilization.The incidence of significant cervical spine injury in previousstudies looking at similar patient populations is between 2and 6%, but the incidence of MRI abnormalities is reported tobe between 25 and 40%. MRI seems to “overcall” significantinjury. It should be noted that the optimal time frame for MRIassessment of the cervical spine is limited. MRI studies arepreferred within the first 48 hours after injury (5, 15, 21, 39,72). Even then, some injuries are poorly visualized. Emery etal. (21) used MRI to study 37 patients with known cervicalspine injury and found that MRI missed ligamentous injury in2 of 19 patients known to have ligamentous injury (abnormalflexion/extension films or surgical confirmation). These im-ages were obtained an average of 10.8 days after injury. Kleinet al. (41), comparing computed tomographic and MRI scansobtained from the same patients, demonstrated that MRI wasnot as effective as CT for recognizing bony abnormalities.MRI, when used early after trauma in conjunction with plainradiographs and CT, is exquisitely sensitive for detecting softtissue abnormalities of the cervical spine. The importance ofthese findings for most patients is, however, unknown.

Flexion/extension radiographs have been used to rule outligamentous injury of the cervical spine. In the awake patient,this maneuver is generally considered safe and effective. Nu-merous series have used flexion/extension films as the goldstandard for the exclusion of ligamentous injury in this pop-TA

BLE

5.1.

Con

tinu

ed

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ceC

lass

Con

clus

ions

Emer

yet

al.,

1989

(21)

MR

Ius

edto

stud

y37

patie

nts

with

know

nsp

ine

inju

ries

.

All

patie

nts

also

stud

ied

with

som

eco

mbi

natio

nof

plai

nx-

rays

/CT/

tom

ogra

ms/

surg

ical

expl

orat

ion/

clin

ical

follo

w-u

p.

II

Sele

ctpo

pula

tion

19pa

tient

sfo

und

toha

velig

amen

tous

inju

ry.

MR

Ide

tect

ed17

/19.

MR

Iw

asne

gativ

ein

18/1

8pa

tient

sw

ithin

tact

ligam

ents

.

Posi

tive

pred

ictiv

eva

lue:

100%

.

Neg

ativ

epr

edic

tive

valu

e:90

%.

Sens

itivi

ty:

89.5

%.

Free

mye

ret

al.,

1989

(24)

Pros

pect

ive

stud

yof

“hig

h-ri

sk”

sym

ptom

atic

patie

nts

imag

edw

ith5-

view

seri

es.

Rad

iogr

aphi

cre

view

of3

view

sco

mpa

red

with

5vi

ews.

CT

used

asgo

ldst

anda

rd.

II

Sele

ctpo

pula

tion

58pa

tient

sst

udie

d.

68in

juri

esde

tect

edin

33pa

tient

s.

Sens

itivi

tyof

3-vi

ewse

ries

was

83%

com

pare

dw

ithC

T.

Add

ition

ofob

lique

view

sha

dno

effe

cton

over

all

sens

itivi

ty.

Rei

det

al.,

1987

(59)

253

patie

nts

with

274

spin

alin

juri

esev

alua

ted.

38ha

dde

lay

indi

agno

sis.

Eval

uate

dca

use

for

dela

yin

diag

nosi

s.

III20

inju

ries

mis

sed

desp

itead

equa

tefil

ms.

Sens

itivi

ty:

92.3

%.

Shaf

fer

and

Dor

is,

1981

(65)

Ret

rosp

ectiv

ean

alys

isof

all

C-s

pine

inju

ries

dete

cted

inco

mm

unity

emer

genc

yro

oms.

Eval

uate

dre

adin

gof

late

ral

film

tore

adin

gof

3-vi

ewse

ries

(3-v

iew

seri

es

was

gold

stan

dard

).

Aut

hors

clai

mno

mis

sed

inju

ries

with

3-vi

ewse

ries

.

III35

inju

ries

dete

cted

.

9ca

ses

mis

sed

with

late

ral

view

alon

e.

Sens

itivi

tyof

late

ral

view

com

pare

dw

ith3

view

s:74

%.

aC

T,co

mpu

ted

tom

ogra

phy;

MR

I,m

agne

tic

reso

nanc

eim

agin

g;C

-spi

ne,

cerv

ical

spin

e;G

CS,

Gla

sgow

Com

aSc

ale

scor

e.



ulation, and no serious adverse events have been reported (1,3, 5, 10, 11, 15). Brady et al. (10) used dynamic flexion/extension spine films to study 451 awake patients with blunttrauma evaluated in an urban emergency room. Flexion/extension views detected abnormalities in 5 of 372 patients inwhom static plain cervical spine films were thought to benormal. None of these patients required “invasive stabiliza-tion,” indicating that the abnormal examinations may havebeen false-positives. It should be noted, however, that false-negative examinations also occur, although infrequently.Lewis et al. (43) reported one false-negative examination in aseries of 141 patients studied with dynamic flexion/extensionfilms. These authors report the negative predictive value forthe combination of plain films and flexion/extension films tobe more than 99%.

The obtunded patient is not able to actively flex or extendthe neck for dynamic radiographic evaluation. Dynamic flu-oroscopy has been used to clear the cervical spine in thesepatients, and results of several series are available (16, 64).Ajani et al. (1) reported an unstable cervical spine injurydetected by flexion/extension radiographs in a patient withnormal plain films and CT (one of 100 patients studied). Daviset al. (16) used dynamic fluoroscopy to study 116 obtundedpatients who had normal cervical radiographs. Only one pa-tient was found to have an injury not visualized on plain filmsor CT. The significance of this injury, a 2-mm subluxation ina patient who was treated in a collar and subsequently lost tofollow-up, is questionable. Sees et al. (64) studied 20 obtundedpatients with normal three-view cervical spine series. Theyperformed bedside flexion/extension under fluoroscopy andfound one patient with C4–C5 subluxation caused by a facetinjury not appreciated on plain films but later confirmed withCT. It should be noted that 30% of the patients in the Sees etal. (64) series could not be cleared because of difficulty visu-alizing the lower cervical spine, whereas Davis et al. (16), byusing radiology staff in the fluoroscopy suite, were able tovisualize the entire spine in virtually all patients.

Because of the high negative predictive value of plain filmsand supplemental CT, application of MRI or flexion/extensionfluoroscopy for clearance of the cervical spine is probably notindicated for every obtunded patient. Use of these modalitiesshould be guided by clinical judgment based on patient historyand physical examination. Subgroups of obtunded trauma pa-tients exist with a low likelihood of cervical spine injury, andexhaustive study is not indicated for these patients. Hanson et al.(29) found that the incidence of cervical spine injury in a series of3684 patients without high-risk criteria was 0.2%, and that all ofthese injuries were detected by plain radiographs supplementedwith CT for poorly visualized or suspicious areas. The high-riskcriteria cited were: a high-speed motor vehicle accident (�35mph); an automobile crash with a death at the scene; a fall frommore than 10 feet; a significant traumatic closed-head injury ortraumatic intracranial hemorrhage; neurological signs or symp-toms referable to the cervical spine; or pelvis or multiple extrem-ity fractures. In support of this issue, Kaups and Davis (40) didnot identify a single cervical spine injury in a group of 215patients with gunshot wounds to the head. Similarly, Patton etal. (57) used MRI and flexion/extension fluoroscopy as a sup-

plement to x-rays to assess the cervical spines of a series ofpatients with isolated head injuries sustained as a result ofassault. They found no undiagnosed injuries.

SUMMARY

In summary, no single radiographic study can adequatelyrule out cervical spine injury in all symptomatic patients. Athree-view cervical spine series supplemented by CT throughareas difficult to visualize and “suspicious” areas will detectmost spinal injuries. This combination of studies representsthe minimum required for clearance of the cervical spine inthe symptomatic patient. The negative predictive value of thiscombination of studies is reported to be between 99 and 100%in several Class II and III evidence studies (9, 11, 24, 48, 67, 68).

In the awake patient, dynamic flexion/extension views(with at least 30-degree excursion in each direction) are safeand effective for detecting most “occult” cervical spine inju-ries not identified on plain x-rays. The negative predictivevalue of a normal three-view series and flexion/extensionviews exceeds 99% (43). Patients who are unable to cooperatewith active flexion/extension radiographs because of pain ormuscle spasm may be maintained in a cervical collar untilthey are able to cooperate, or they may be studied with MRI.A negative MRI study within the first 48 hours of injury, inaddition to normal radiographs and supplemental CT, seemsto be sufficient for clearing the cervical spine. The significanceof a positive MRI study is currently unclear. It is suggestedthat cervical immobilization be continued in these patientsuntil delayed flexion/extension views can be obtained.

In the obtunded patient with a normal three-view x-rayseries and appropriate CT of the cervical spine, the incidenceof significant spine injury is less than 1%. On the basis ofmechanism of injury and clinical judgment, the cervical spinein selected patients may be considered cleared without fur-ther study. In the remainder of cases, flexion/extension per-formed under fluoroscopic visualization seems to be safe andeffective for ruling out significant ligamentous injury, with areported negative predictive value of more than 99% (16).Because the incidence of occult injury diagnosed with dy-namic flexion/extension fluoroscopy in the setting of normalplain cervical spine x-rays and CT images is low, it is probablymost efficient for these procedures to be performed by staff inthe department of radiology, although variances in local ex-perience should be respected. MRI represents another optionfor clearance of the spine in this patient population, and anegative MRI within 48 hours of injury seems to effectivelyeliminate the likelihood of a significant ligamentous injury.However, MRI evaluation will result in a large number offalse-positive examinations, and the consequences of pro-longed unnecessary immobilization in the obtunded patientare not insignificant (18, 44, 58).


The significance of positive MRI findings after cervicaltrauma should be evaluated by using flexion/extension ra-diographs and clinical follow-up as the gold standard. Theincidence of abnormal findings on flexion/extension fluoro-



scopic studies in obtunded patients should be evaluated in aprospective fashion with appropriate clinical follow-up. Aprospective comparison should be made between the three-view cervical spine series supplemented with selective CTthrough poorly visualized or suspicious areas and CT of theentire cervical spine.


REFERENCES

1. Ajani AE, Cooper DJ, Scheinkestel CD, Laidlaw J, Tuxen DV:Optimal assessment of cervical spine trauma in critically ill pa-tients: A prospective evaluation. Anaesth Intensive Care 26:487–491, 1998.

2. Bachulis BL, Long WB, Hynes GD, Johnson MC: Clinical indica-tions for cervical spine radiographs in the traumatized patient.Am J Surg 153:473–478, 1987.

3. Banit DM, Grau G, Fisher JR: Evaluation of the acute cervicalspine: A management algorithm. J Trauma 49:450–456, 2000.

4. Bayless P, Ray VG: Incidence of cervical spine injuries in associ-ation with blunt head trauma. Am J Emerg Med 7:139–142, 1989.

5. Benzel EC, Hart BL, Ball PA, Baldwin NG, Orrison WW, EspinosaMC: Magnetic resonance imaging for the evaluation of patientswith occult cervical spine injury. J Neurosurg 85:824–829, 1996.

6. Berne JD, Velmahos GC, El-Tawil Q, Demetriades D, Asensio JA,Murray JA, Cornwell EE, Belzberg H, Berne TV: Value of com-plete cervical helical computed tomographic scanning in identi-fying cervical spine injury in the unevaluable blunt trauma pa-tient with multiple injuries: A prospective study. J Trauma 47:896–903, 1999.

7. Blackmore CC, Emerson SS, Mann FA, Koepsell TD: Cervicalspine imaging in patients with trauma: Determination of fracturerisk to optimize use. Radiology 211:759–765, 1999.

8. Blackmore CC, Ramsey SD, Mann FA, Deyo RA: Cervical spinescreening with CT in trauma patients: A cost effective analysis.Radiology 212:117–125, 1999.

9. Borock EC, Gabram SG, Jacobs LM, Murphy MA: A prospectiveanalysis of a two-year experience using computed tomography asan adjunct for cervical spine clearance. J Trauma 31:1001–1006,1991.

10. Brady WJ, Moghtader J, Cutcher D, Exline C, Young J: ED use offlexion-extension cervical spine radiography in the evaluation ofblunt trauma. Am J Emerg Med 17:504–508, 1999.

11. Brohi K, Wilson-Macdonald J: Evaluation of unstable cervicalspine injury: A 6-year experience. J Trauma 49:76–80, 2000.

12. Chee SG: Review of the role of magnetic resonance imaging inacute cervical spine injuries. Ann Acad Med Singapore 22:757–761, 1993.

13. Clancy MJ: Clearing the cervical spine of adult victims of trauma.J Accid Emerg Med 16:208–214, 1999.

14. Cohn SM, Lyle WG, Linden CH, Lancey RA: Exclusion of cervicalspine injury: A prospective study. J Trauma 31:570–574, 1991.

15. D’Alise MD, Benzel EC, Hart BL: Magnetic resonance imagingevaluation of the cervical spine in the comatose or obtundedtrauma patient. J Neurosurg 91[Suppl 1]:54–59, 1999.

16. Davis JW, Parks SN, Detlefs CL, Williams GG, Williams JL, SmithRW: Clearing the cervical spine in obtunded patients: The use ofdynamic fluoroscopy. J Trauma 39:435–438, 1995.

17. Davis JW, Phreaner DL, Hoyt DB, Mackersie RC: The etiology ofmissed cervical spine injuries. J Trauma 34:342–346, 1993.

18. Dodd FM, Simon E, McKeown D, Patrick MR: The effect of acervical collar on the tidal volume of anaesthetised adult patients.Anaesthesia 50:961–963, 1995.

19. Domeier RM, Evans RW, Swor RA, Rivera-Rivera EJ, FrederiksenSM: High-risk criteria for performing pre-hospital spinal immo-bilization in trauma. Ann Emerg Med 25:141–142, 1995.

20. Domeier RM, Evans RW, Swor RA, Rivera-Rivera EJ, FrederiksenSM: Prospective validation of out-of-hospital spinal clearancecriteria: A preliminary report. Acad Emerg Med 4:643–646, 1997.

21. Emery SE, Pathria MN, Wilber RG, Masaryk T, Bohlman HH:Magnetic resonance imaging of posttraumatic spinal ligamentinjury. J Spinal Disord 2:229–233, 1989.

22. Fehlings MG, Rao SC, Tator CH, Skaf G, Arnold P, Benzel EC,Dickman C, Cuddy B, Green B, Hitchon P, Northrup B, SonntagVKH, Wagner F, Wilberger JE: The optimal method for assessingspinal canal compromise and cord compression in patients withcervical spinal cord injury. Spine 24:605–613, 1999.

23. Flanders AE, Schaefer DM, Doan HT, Mishkin MM, Gonzales CF,Northrup BE: Acute cervical spine trauma: Correlation of MRimaging findings with degree of neurologic deficit. Radiology177:25–33, 1990.

24. Freemyer B, Knopp R, Piche J, Wales L, Williams J: Comparison offive-view and three-vies cervical spine series in the evaluation ofpatients with cervical trauma. Ann Emerg Med 18:818–821, 1989.

25. Gerrelts BD, Petersen EU, Mabry J, Petersen SR: Delayed diagno-sis of cervical spine injuries. J Trauma 31:1622–1626, 1991.

26. Gonzales RP, Fried PO, Bukhalo M, Holevar MR, Falimirski ME:Role of clinical examination in screening for blunt injury. J AmColl Surg 189:152–157, 1999.

27. Grossman MD, Reilly PM, Gillett T, Gillett D: National survey ofthe incidence of cervical spine injury and approach to cervicalspine clearance in US trauma centers. J Trauma 47:684–690, 1999.

28. Hall AJ, Wagle VG, Raycroft J, Goldman RL, Butler AR: Magneticresonance imaging in cervical spine trauma. J Trauma 34:21–26,1993.

29. Hanson JA, Blackmore CC, Mann FA, Wilson AJ: Cervical spineinjury: A clinical decision rule to identify high risk patients forhelical CT screening. Am J Roentgenol 174:713–717, 2000.

30. Herr CH, Ball PA, Sargent SK, Quinton HB: Sensitivity ofprevertebral soft tissue measurement at C3 for detection of cer-vical spine fractures and dislocations. Am J Emerg Med 16:346–349, 1998.

31. Hoffman JR, Mower WR, Wolfson AB, Todd KH, Zucker MI:Validity of a set of clinical criteria to rule out injury to the cervicalspine in patients with blunt trauma. N Engl J Med 343:94–99,2000.

32. Hoffman JR, Schriger DL, Mower W, Luo JS, Zucker M: Low-riskcriteria for cervical spine radiography in blunt trauma: A pro-spective study. Ann Emerg Med 21:1454–1460, 1992.

33. Hoffman JR, Wolfson AB, Todd K, Mower WR: Selective cervicalspine radiography in blunt trauma: Methodology of the NationalEmergency X-Radiography Utilization Study (NEXUS). AnnEmerg Med 32:461–469, 1998.

34. Holliman CJ, Mayer JS, Cook RT Jr, Smith JS Jr: Is the anteropos-terior cervical spine radiograph necessary in initial traumascreening? Am J Emerg Med 9:421–425, 1991.

35. Iida H, Tachibana S, Kitahara T, Horiike S, Ohwada T, Fujii K:Association of head trauma with cervical spine injury, spinal cordinjury, or both. J Trauma 46:450–452, 1999.

36. Ireland AJ, Britton I, Forrester AW: Do supine oblique viewsprovide better imaging of the cervicothoracic junction than swim-mer’s views? J Accid Emerg Med 15:151–154, 1998.




38. Kaneriya PP, Schweitzer ME, Spettell C, Cohen MJ, Karasick D:The cost effectiveness of oblique radiography in the exclusion ofC7–T1 injury in trauma patients. Am J Radiol 171:959–962, 1998.

39. Katzberg RW, Benedetti PF, Drake CM, Ivanovic M, Levine RA,Beatty CS, Nemzek WR, McFall RA, Ontell FK, Bishop DM,Poirier VC, Chong BW: Acute cervical spine injuries: ProspectiveMR imaging assessment at a level one trauma center. Radiology213:203–212, 1999.

40. Kaups K, Davis J: Patients with gunshot wounds to the head donot require cervical spine immobilization and evaluation.J Trauma 44:865–867, 1998.

41. Klein GR, Vaccaro AR, Albert TJ, Schweitzer M, Deely D, KarasickD, Cotler JM: Efficacy of magnetic resonance imaging in theevaluation of posterior cervical spine fractures. Spine 24:771–774,1999.


43. Lewis LM, Docherty M, Ruoff BE, Fortney JP, Keltner RA Jr,Britton P: Flexion-extension views in the evaluation of cervicalspine injuries. Ann Emerg Med 20:117–121, 1991.

44. Liew SC, Hill DA: Complications of hard cervical collars in multi-trauma patients. Aust N Z J Surg 64:139–140, 1994.

45. Lindsey RW, Diliberti TC, Doherty BJ, Watson AB: Efficacy ofradiographic evaluation of the cervical spine in emergency situ-ations. South Med J 86:1253–1255, 1993.

46. Lockey AS, Handley R, Willett K: “Clearance” of cervical spineinjury in the obtunded patient. Injury 29:493–497, 1998.

47. Macdonald RL, Schwartz ML, Mirich D, Sharkey PW, Nelson WR:Diagnosis of cervical spine injury in motor vehicle crash victims:How many x-rays are enough? J Trauma 30:392–397, 1990.

48. Mace SE: Emergency evaluation of cervical spine injuries: CTversus plain radiographs. Ann Emerg Med 14:973–975, 1985.

49. Mace SE: Unstable occult cervical spine fracture. Ann Emerg Med20:1373–1375, 1991.

50. Mahadevan S, Mower WR, Hoffman JR, Peeples N, Goldberg W,Sonner R: Interrater reliability of cervical spine injury criteria inpatients with blunt trauma. Ann Emerg Med 31:197–201, 1998.

51. McNamara RM, Heine E, Esposito B: Cervical spine injury andradiography in alert, high risk patients. J Emerg Med 8:177–182,1990.

52. Mirvis SE, Diaconis JN, Chirico PA, Reiner BI, Joslyn JN, MilitelloP: Protocol driven radiologic evaluation of suspected cervicalspine injury: Efficacy study. Radiology 170:831–834, 1989.

53. Neifeld GL Keene JG, Hevesy G, Leikin J, Proust A, Thisted RA:Cervical injury in head trauma. J Emerg Med 6:203–207, 1988.

54. Nunez DB Jr, Quencer RM: The role of helical CT in the assess-ment of cervical spine injuries. AAJR Am J Radiol 171:951–957,1998.

55. Nunez DB Jr, Zuluaga A, Fuentes-Bernardo DA, Rivas LA,Beccera JL: Cervical spine trauma: How much more do we learnby routinely using helical CT? Radiographics 16:1318–1321, 1996.

56. Pasquale M, Fabian TC: Practice management guidelines fortrauma from the Eastern Association for the Surgery of Trauma.J Trauma 44:945–946, 1998.

57. Patton JH, Kralovich KA, Cuschieri J, Gasparri M: Clearing thecervical spine in victims of blunt assault to the head and neck:What is necessary? Am Surg 66:326–330, 2000.

58. Raphael JH, Chotai R: Effects of the cervical collar of cerebrospi-nal fluid pressure. Anaesthesia 49:437–439, 1994.

59. Reid DC, Henderson R, Saboe L, Miller JD: Etiology and clinicalcourse of missed spine fractures. J Trauma 27:980–986, 1987.

60. Roberge RJ, Wears RC, Kelly M, Evans TC, Kenny MA, DaffnerRD, Kremen R, Murray K, Cottington EC: Selective application ofcervical spine radiography in alert victims of blunt trauma: Aprospective study. J Trauma 28:784–788, 1988.

61. Ross SE, O’Malley KF, DeLong WG, Born CT, Schwab CW: Clin-ical predictors of unstable cervical spine injury in multiply injuredpatients. Injury 23:317–319, 1992.

62. Roth BJ, Martin RR, Foley K, Barcia PJ, Kennedy P: Roentgeno-graphic evaluation of the cervical spine. Arch Surg 129:643–645,1994.

63. Scarrow AM, Levy EI, Resnick DK, Adelson PD, Sclabassi RJ:Cervical spine evaluation in obtunded or comatose pediatrictrauma patients: A pilot study. Pediatr Neurosurg 30:169–175,1999.

64. Sees DW, Cruz LR, Flaherty SF, Ciceri DP: The use of bedsidefluoroscopy to evaluate the cervical spine in obtunded patients.J Trauma 45:768–771, 1998.

65. Shaffer MA, Doris PE: Limitation of the cross table lateral view indetecting cervical spine injuries: A prospective analysis. AnnEmerg Med 10:508–513, 1981.

66. Sweeney JF, Rosemurgy AS, Gill S, Albrink MH: Is the cervicalspine clear? Undetected cervical fractures diagnosed only at au-topsy. Ann Emerg Med 21:1288–1290, 1992.

67. Tan E, Schweitzer ME, Vaccaro L, Spetell AC: Is computed to-mography of nonvisualized C7–T1 cost effective? J Spinal Disord12:472–476, 1999.

68. Tehranzadeh J, Bonk RT, Ansari A, Mesgarzadeh M: Efficacy oflimited CT for nonvisualized lower cervical spine in patients withblunt trauma. Skeletal Radiol 23:349–352, 1994.

69. Tracy PT, Wright RM, Hanigan WC: Magnetic resonance imagingof spinal injury. Spine 14:292–301, 1989.

70. Vaccaro AR, An HS, Lin S, Sun S, Balderston RA, Cotler JM:Noncontiguous injuries of the spine. J Spinal Disord 5:320–329,1992.

71. Walters BC: Clinical practice parameter development in neuro-surgery, in Bean JR (ed): Neurosurgery in Transition: The Socioeco-nomic Transformation of Neurological Surgery. Baltimore, Williams& Wilkins, 1998, pp 99–111.

72. White P, Seymour R, Powell N: MRI assessment of the pre-vertebral soft tissues in acute cervical spine trauma. Br J Radiol72:818–823, 1999.




CHAPTER 6

Initial Closed Reduction of Cervical SpineFracture-Dislocation Injuries

RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS:• Early closed reduction of cervical spine fracture-dislocation injuries with craniocervical traction is recom-

mended to restore anatomic alignment of the cervical spine in awake patients.• Closed reduction in patients with an additional rostral injury is not recommended.• Patients with cervical spine fracture-dislocation injuries who cannot be examined during attempted closed

reduction, or before open posterior reduction, should undergo magnetic resonance imaging (MRI) beforeattempted reduction. The presence of a significant disc herniation in this setting is a relative indication fora ventral decompression before reduction.

• MRI study of patients who fail attempts at closed reduction is recommended.• Prereduction MRI performed in patients with cervical fracture dislocation injury will demonstrate disrupted

or herniated intervertebral discs in one-third to one-half of patients with facet subluxation. These findingsdo not seem to significantly influence outcome after closed reduction in awake patients; therefore, theusefulness of prereduction MRI in this circumstance is uncertain.

RATIONALE

Spinal cord injury (SCI) is frequently associated withtraumatic cervical spine fractures and cervical facet dis-location injuries because the spinal canal can be nar-

rowed by displacement of fracture fragments or subluxationof one vertebra over another. Reduction of the deformityhelps to restore the diameter of the bony canal and eliminatesbony compression of the spinal cord caused by the vertebralfracture and/or subluxation. Theoretically, early decompres-sion of the spinal cord after injury may lead to improvedneurological outcome. Several large case series of patientsdescribe excellent results with closed reduction of cervicalfractures and facet subluxations. However, descriptive seriesusing prereduction magnetic resonance imaging (MRI) havereported a high incidence of cervical disc herniation in pa-tients with facet dislocation. Furthermore, there are case re-ports and small case series of patients who worsened neuro-logically after closed cervical spine reduction. Several of thesereports implicate ventral compression of the spinal cord bydisplaced disc material. The purpose of this qualitative reviewis to address the following issues:

1. Is closed reduction safe and effective for reducing cervicalspine deformity in patients with cervical fractures or uni-lateral or bilateral facet dislocation injuries?

2. Is a prereduction MRI study essential for managing thesepatients?

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasperformed. The following subject headings were combinedwith “spinal cord injury”: “spinal fracture,” “spinal injury,”and “human.” Approximately 12,300 citations were acquired.Non-English language citations were deleted. Searching thisset of publications with the keyword “cervical” resulted in2,154 matches. Further refining the search with the terms“reduction” or “fracture” yielded 606 articles; titles and ab-stracts of the articles were reviewed. Clinical series dealingwith adult patients in the acute setting were selected. Casereports and case collections were included in the review butare not listed in the table. Additional references were culledfrom the reference lists of the remaining articles. The membersof the author group were asked to contribute articles knownto them on the subject matter that were not found by othersearch means. Finally, the tables of contents of the journalSpine from 1993 through May 2000 were hand-searched. Atotal of 42 articles relevant to this topic were identified andform the basis of this guideline. The larger clinical series aresummarized in Table 6.1.


Among the articles reviewed, there were no randomizedclinical trials, no prospective cohort studies, and no case-control studies. One historical cohort study comparedtraction-reduction in awake patients with manipulation underanesthesia (MUA). The remainder of the articles consisted ofcase series of patients with acute or subacute unilateral orbilateral cervical facet dislocation injuries treated with open orclosed reduction. Several case reports and case series of pa-tients who deteriorated after closed reduction were identifiedand are included. Several studies included pre- and pos-treduction MRI findings.


Walton (39), in 1893, first described closed reduction bymanipulation of cervical spine deformity caused by facet dis-location. Crutchfield (7) introduced tongs for in-line traction-reduction in 1933, and similar techniques have been success-fully used for traction-reduction of cervical deformity by alarge number of authors (1–4, 6, 11, 12, 16, 21, 23, 27, 29–31, 34,37, 38). Observations by Evans (10) and Kleyn (19) popular-ized reduction under anesthesia, although other authors con-demned the procedure as potentially dangerous comparedwith craniocervical traction-reduction. MUA is still a fre-quently practiced technique, usually used after failure oftraction-reduction but occasionally used as a primary meansof achieving reduction (9, 16, 41). Only one cohort study hasbeen performed comparing the two modalities. Lee et al. (20)found a higher rate of success and a lower complication ratewith traction-reduction as opposed to MUA. The significanceof their results is questionable because of the historical cohortdesign of the study. Lee et al. attributed the higher complica-tion rate in the MUA group to the effects of anesthesia onperfusion of the injured spinal cord. It is possible, however,that advances in the pharmacological and medical manage-ment of SCI patients during the 10-year period of data accrualaccounted for the improved results the authors noted in thetraction-reduction group. For this reason, the evidence pro-vided by this study is considered to be Class III medicalevidence.

Recent reports of neurological deterioration after closed oropen posterior reduction of cervical fracture-dislocation inju-ries has led some authors to recommend the use of prereduc-tion MRI to assess for ventral cord compromise caused bytraumatic disc disruption. Several investigators believe thatdisc disruption in association with facet fracture-dislocationincreases the risk of spinal cord injury by disc material afterreduction (7, 8, 20, 25). Other authors, however, have foundno relationship between findings on prereduction MRI, neu-rological outcome, or findings on postreduction MRI (33). Thenature of the injury predisposes a large percentage of patientswith cervical facet dislocation injuries to have MRI evidenceof disc material ventral to the spinal cord. Rizzolo et al. (28)found evidence of disc disruption/herniation in 42% of pa-

tients studied with prereduction MRI. The clinical relevanceof these findings has not been proven. The use of prereductionMRI may delay reduction of the spinal deformity and there-fore may delay decompression of the compromised spinalcord. Prereduction MRI assessment requires the transport of apatient with a highly unstable cervical spine fracture to theMRI suite. Recent laboratory work, as well as Class III evi-dence, suggests that early reduction of fracture-dislocationinjuries may improve neurological outcome (2, 10, 16, 20, 26).If stabilization of the unstable cervical spine protects againstadditional injury to the cervical spinal cord, then the informa-tion gained by prereduction MRI must be of sufficient value towarrant the delay in treatment and the associated potentialmorbidity of transport.

Most of the clinical series reviewed were based on patientdata accrued before the introduction of MRI. These combinedseries encompass more than 1200 patients treated with closedreduction in the acute or subacute period after injury. Thesuccess rate for restoring anatomic alignment by closed re-duction in these studies was approximately 80%. The reportedpermanent neurological complication rate was less than 1.0%(3, 6, 9, 11, 12, 15, 16, 20–21, 23, 27–31, 34, 36, 38, 42). Of the 11patients reported to develop new permanent neurologicaldeficits with attempted closed reduction, two had root inju-ries (11, 12), and two had ascending spinal cord deficits notedat the time of reduction (3, 30). Seven patients were noted tohave decreased American Spinal Injury Association (ASIA)scores postreduction, but neither the nature nor the cause ofthe new deficits in these patients was described (16).

Transient neurological deterioration after closed reductionhas been reported. In addition to the permanent deficits notedabove, temporary deficits have been described in 20 addi-tional patients of these 1200. These deficits reversed sponta-neously or improved after reduction of weight or after openreduction (3, 11, 12, 16, 20, 31). The causes of neurologicaldeterioration associated with closed reduction in these andother series included overdistraction (3, 21, 30), failure torecognize a more rostral noncontiguous lesion (30, 32), discherniation (11), epidural hematoma (17), and spinal cordedema (19, 20).

Several authors have provided general suggestions on howcraniocervical traction for closed reduction of cervical spinefracture-dislocation injuries is best accomplished (14, 15, 18,40). No study has been undertaken to determine the superi-ority of one method or technique over another. Tongs(Crutchfield or Gardner-Wells) or a halo ring are applied tothe patient’s head using sterile technique and local anesthesiaat the tong or pin insertion sites. Most contemporary descrip-tions incorporate the use of an MRI-compatible halo ring asthe cranial fixation device for four-point fixation of the cra-nium. This device gives better control of the head and neck ifpositioning and directional traction are needed (e.g., passivedirectional traction in positions of flexion or extension of theneck depending on the injury type) and facilitates halo-ring

Initial Closed Reduction of Fracture-Dislocation Injuries


TAB

LE6.

1.Su

mm

ary

ofR

epor

tson

Clo

sed

Red

ucti

onof

Cer

vica

lSp

inal

Frac

ture

-Dis

loca

tion

Inju

ries

a

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yR

esul

tsEv

iden

ce

Cla

ssC

oncl

usio

ns

Gra

ntet

al.,

1999

(13)

82pa

tient

s.

Ret

rosp

ectiv

ese

ries

.

All

clos

edC

-spi

nein

juri

esw

ithm

alal

ignm

ent

incl

uded

.

Uni

late

ral

and

bila

tera

llo

cked

face

ts.

Earl

yra

pid

clos

edre

duct

ion

atte

mpt

edin

all

patie

nts.

MR

Isc

ans

obta

ined

afte

rre

duct

ion.

ASI

Aan

dFr

anke

lgr

ades

dete

rmin

edon

adm

issi

onan

dat

6an

d24

h.

Wei

ght

upto

80%

ofpa

tient

’sbo

dyw

eigh

t.

Succ

essf

ulre

duct

ion

in97

.6%

.

Ave

rage

time

tore

duct

ion

2.1

�0.

24h.

Ove

rall

ASI

Asc

ores

impr

oved

by24

haf

ter

redu

ctio

n.

1pa

tient

dete

rior

ated

6h

post

redu

ctio

n(p

roba

ble

root

lesi

on).

46%

had

disc

inju

ryon

MR

I,22

%ha

dhe

rnia

tion.

Dis

cin

jury

onM

RI

did

corr

elat

ew

ithco

rded

ema

onM

RI.

IIIC

lose

dre

duct

ion

isef

fect

ive

and

safe

,de

spite

high

inci

denc

eof

MR

I-de

mon

stra

ble

disc

inju

ries

/her

niat

ions

.

Vita

let

al.,

1998

(38)

168

patie

nts.

Ret

rosp

ectiv

ese

ries

.

Uni

late

ral

and

bila

tera

lfa

cet

inju

ries

.

Empl

oyed

man

ipul

atio

nun

der

gene

ral

anes

thes

iain

min

ority

ofca

ses.

Use

dre

lativ

ely

light

wei

ghts

(max

imum

8.8

lbpl

us2.

2pe

rle

vel

for

max

imum

of40

lb).

All

patie

nts

oper

ated

onim

med

iate

lypo

stre

duct

ion

orpo

st-f

ailu

reof

redu

ctio

n.

MR

Isc

ans

not

done

prer

educ

tion

(alth

ough

disc

sno

ted

in7

patie

nts?

).

43%

redu

ced

bytr

actio

nw

ithou

tan

esth

esia

(tim

e�

2h)

.

30%

redu

ced

byM

UA

.

27%

redu

ced

intr

aope

rativ

ely.

5pa

tient

sdi

dno

tre

duce

(del

ayed

refe

rral

,su

rgic

aler

ror)

.

Aut

hors

obse

rved

noca

ses

ofne

urol

ogic

alde

teri

orat

ion.

IIIA

utho

rspr

omot

eth

eir

prot

ocol

asa

safe

and

effe

ctiv

em

eans

for

redu

ctio

nan

dst

abili

zatio

nof

frac

ture

s.

Lee

etal

.,19

94(2

0)21

0pa

tient

s.

RT

in11

9,M

UA

in91

.

Ret

rosp

ectiv

ehi

stor

ical

coho

rtst

udy.

Gro

ups

sim

ilar

exce

pttr

actio

ngr

oup

had

long

erde

lay

totr

eatm

ent.

Wei

ghts

upto

150

lbus

ed.

No

MR

Ido

nepr

ered

uctio

n.

Red

uctio

nsu

cces

sful

:

MU

A:

66/9

1(7

3%).

RT:

105/

119

(88%

).

All

failu

res

inR

Tgr

oup

wer

edu

eto

asso

ciat

edfr

actu

res

or

dela

yed

refe

rral

.

Tim

eto

redu

ctio

n:

RT

21m

in.

MU

A:

not

repo

rted

.

No

loss

ofFr

anke

lgr

ade

inei

ther

grou

p.

6M

UA

and

1R

Tha

dde

teri

orat

ion

onA

SIA

scor

e.

IIIR

Tsu

peri

orto

MU

A,

both

proc

edur

essa

fean

d

effe

ctiv

e,M

RI

not

done

.

Cot

ler

etal

.,19

93(5

)24

patie

nts

(all

awak

e).

Pros

pect

ive

stud

y.

No

frac

ture

dfa

cets

.

All

acut

ein

juri

es(1

5-d

patie

nt).

Wei

ghts

upto

140

lbus

ed.

No

CT

orM

RI

done

.

All

24re

duce

d.

No

inci

denc

eof

neur

olog

ical

dete

rior

atio

n.

Man

ipul

atio

nus

edin

addi

tion

tow

eigh

tsin

9pa

tient

s(w

hen

face

tspe

rche

d).

Tim

ere

quir

edra

nged

from

8to

187

min

.

IIIR

educ

tion

with

wei

ghts

upto

140

lbis

safe

and

effe

ctiv

ein

mon

itore

dse

tting

with

expe

rien

ced

phys

icia

ns.

Mah

ale

etal

.,19

93(2

4)34

1pa

tient

str

eate

dfo

rtr

aum

atic

disl

ocat

ions

ofce

rvic

alsp

ine.

15ha

dne

urol

ogic

alde

teri

orat

ion.

Var

iety

oftr

eatm

ents

used

tore

duce

defo

rmity

,in

clud

ing

oper

ativ

e

redu

ctio

n.

Com

plet

ein

juri

es:

6af

ter

OR

,1

afte

rm

anip

ulat

ion.

Inco

mpl

ete

inju

ries

:1

afte

rO

R,

3af

ter

man

ipul

atio

n,2

afte

r

trac

tion,

1du

ring

appl

icat

ion

ofca

st.

Rad

icul

opat

hy:

1(o

ccur

red

whe

nto

ngs

slip

ped

duri

ng

trac

tion)

.

Det

erio

ratio

nde

laye

din

11pa

tient

s.

IIIN

umbe

rsof

patie

nts

subj

ect

toea

chtr

eatm

ent

arm

not

give

n.Pu

rely

ade

scri

ptiv

epa

per.

Onl

y

conc

lusi

onis

that

neur

olog

ical

dete

rior

atio

nca

n

happ

en.

Had

ley

etal

.,19

92(1

5)68

patie

nts.

Ret

rosp

ectiv

ese

ries

.

Face

tfr

actu

redi

sloc

atio

nson

ly.

Uni

late

ral

and

bila

tera

llo

cked

face

ts.

66tr

eate

dw

ithea

rly

atte

mpt

edcl

osed

redu

ctio

n(2

late

refe

rral

s).

Ave

rage

wei

ghts

used

for

succ

essf

ulre

duct

ion

wer

ebe

twee

n9

and

10

lb/c

rani

alle

vel.

58%

ofpa

tient

sha

dsu

cces

sful

redu

ctio

n.

Ove

rall,

mos

tpa

tient

s(7

8%)

dem

onst

rate

dne

urol

ogic

al

reco

very

byla

stfo

llow

-up

(not

quan

tifie

d).

7pa

tient

sde

teri

orat

eddu

ring

“tre

atm

ent”

(6im

prov

edaf

ter

oper

ativ

ere

duct

ion,

1pe

rman

ent

root

defic

itaf

ter

trac

tion)

.

No

MR

Ida

tare

port

ed.

IIIEa

rly

deco

mpr

essi

onby

redu

ctio

nle

dto

impr

oved

outc

omes

,ba

sed

onfa

ctth

atpa

tient

s

who

did

best

wer

ere

duce

dea

rly

(�5–

8h)

.

No

com

pari

son

poss

ible

betw

een

CR

and

oper

ativ

ere

duct

ion

due

tosm

all

num

bers

.

1.2%

perm

anen

tde

ficit

(roo

t)re

late

dto

trac

tion.



TAB

LE6.

1.C

onti

nued

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yR

esul

tsEv

iden

ce

Cla

ssC

oncl

usio

ns

Star

etal

.,19

90(3

5)57

patie

nts.

Ret

rosp

ectiv

ese

ries

.

Uni

late

ral

and

bila

tera

lin

juri

es.

Earl

yra

pid

redu

ctio

nat

tem

pted

inal

lpa

tient

s.

No

MR

Ido

nepr

ered

uctio

n.

1pa

tient

was

ade

laye

dtr

ansf

er.

Wei

ghts

upto

160

lb(b

egan

at10

lb).

Fran

kel

grad

esre

cord

edat

adm

issi

onan

ddi

scha

rge.

53/5

7(9

3%)

redu

ced

mea

ntim

eto

redu

ctio

nw

as8

h.

No

patie

ntde

teri

orat

eda

Fran

kel

grad

e.

2pa

tient

slo

stro

otfu

nctio

n,1

tran

sien

tly.

45%

impr

oved

one

Fran

kel

grad

eby

time

ofdi

scha

rge,

23%

impr

oved

less

subs

tant

ially

.

75%

ofpa

tient

sre

quir

ed�

50lb

wei

ght.

IIIC

lose

dre

duct

ion

issa

fean

def

fect

ive

for

deco

mpr

essi

ngco

rdan

des

tabl

ishi

ngsp

inal

alig

nmen

t.

Sabi

ston

etal

.,19

88(3

1)39

patie

nts.

Ret

rosp

ectiv

ese

ries

.

Uni

late

ral

and

bila

tera

lin

juri

es.

Up

to70

%of

body

wei

ght

used

.

All

acut

ein

juri

es.

No

MR

I.

35/3

9(9

0%)

ofin

juri

essu

cces

sful

lyre

duce

d,av

erag

e

wei

ght

used

62.5

lb.

No

neur

olog

ical

dete

rior

atio

n.

Failu

res

due

tosu

rgeo

nun

will

ingn

ess

tous

em

ore

wei

ght.

IIIC

lose

dre

duct

ion

with

upto

70%

ofbo

dyw

eigh

tis

safe

and

effe

ctiv

efo

rre

duci

nglo

cked

face

ts.

Aut

hors

stat

eth

atpa

tient

sse

enin

dela

yed

fash

ion

(�10

d)ar

eun

likel

yto

redu

ce(n

oev

iden

cepr

esen

ted

here

).

Mai

man

etal

.,19

86(2

5)28

patie

nts.

Var

iety

oftr

eatm

ents

offe

red.

No

MR

I.

18pa

tient

sha

dat

tem

pted

clos

edre

duct

ion

(max

imum

wei

ght

50lb

).

10/1

8re

duce

dw

ithtr

actio

n.

No

patie

nttr

eate

dby

auth

ors

dete

rior

ated

.

One

refe

rred

patie

ntha

dan

over

dist

ract

ion

inju

ry.

IIIM

ixed

grou

pof

patie

nts

and

trea

tmen

ts.

Inge

nera

l,

trac

tion

seem

edto

besa

fe.

Kle

yn,

1984

(19)

101

patie

nts.

Uni

late

ral

and

bila

tera

lin

juri

es,

all

with

neur

olog

ical

invo

lvem

ent.

All

trea

ted

with

trac

tion.

Ifin

jury

�24

h,M

UA

atte

mpt

edin

itial

ly;

iftr

actio

nre

duct

ion

fails

with

max

imum

18kg

wei

ght,

MU

Ape

rfor

med

.

Pre-

MR

Iut

ilize

d.

82/1

01su

cces

sful

lyre

duce

d(4

open

redu

ctio

n,6

part

ial

redu

ctio

nac

cept

ed,

9no

furt

her

atte

mpt

due

topo

or

cond

ition

ofpa

tient

).

37/4

5in

com

plet

ele

sion

sim

prov

ed.

7/56

com

plet

ele

sion

sim

prov

ed.

No

neur

olog

ical

dete

rior

atio

n.

IIITr

actio

nfo

llow

edby

MU

Ais

safe

,us

ually

(80%

)

effe

ctiv

e,an

dm

ayre

sult

inim

prov

edne

urol

ogic

al

func

tion.

Sonn

tag,

1981

(34)

15pa

tient

s.

Ret

rosp

ectiv

ean

alys

is.

All

bila

tera

llo

cked

face

ts.

All

acut

ein

juri

es.

Man

ual

trac

tion,

tong

trac

tion,

and

open

redu

ctio

nus

ed.

Wei

ght

used

rang

edfr

om30

to75

lb.

No

MR

Ido

ne.

Red

uctio

nw

ithtr

actio

nsu

cces

sful

in10

patie

nts.

5fa

iled:

1w

ithC

1fr

actu

rew

hich

did

not

allo

wtr

actio

n,

2w

ithfr

actu

red

face

ts,

1w

ithra

dicu

lar

sym

ptom

s

wor

sene

dby

trac

tion

(tran

sien

t),1

with

anas

cend

ing

spin

alco

rdin

jury

(pat

ient

died

ofpu

lmon

ary

com

plic

atio

ns2

wk

late

r).

IIISt

epw

ise

algo

rith

m(tr

actio

n,m

anua

lm

anip

ulat

ion,

oper

ativ

ere

duct

ion)

isin

dica

ted.

Clo

sed

redu

ctio

nby

wei

ght

appl

icat

ion

isth

e

pref

erre

dm

etho

dfo

rre

duct

ion

ofde

form

ity.

Shro

sbre

e,19

79(3

3)21

6pa

tient

sid

entif

ied

with

lock

edfa

cets

(86

died

with

in3

mo,

excl

uded

from

seri

es).

Use

dtr

actio

n(n

ow

eigh

tsp

ecifi

ed)

follo

wed

bym

anip

ulat

ion

unde

r

anes

thes

iaif

trac

tion

faile

d.

Pre-

MR

I.

70/9

5un

ilate

ral

ijuri

esre

duce

d(7

4%).

77/1

21bi

late

ral

inju

ries

redu

ced

(64%

).

No

neur

olog

ical

mor

bidi

tyre

port

ed.

Patie

nts

who

wer

esu

cces

sful

lyre

duce

dim

prov

edm

ore

ofte

nth

anpa

tient

sw

how

ere

not

succ

essf

ully

redu

ced

(41%

vers

us32

%un

ilate

ral,

16%

vers

us0%

bila

tera

l).

IIID

isca

rded

patie

nts

and

lack

ofst

atis

tical

anal

ysis

prec

lude

firm

stat

emen

ts.

Hig

hly

sugg

estiv

epa

per.

Con

clus

ions

:tr

actio

nfo

llow

edby

man

ipul

atio

nis

safe

and

usua

llyef

fect

ive,

and

redu

ctio

nse

ems

toim

prov

e

outc

ome

(or

patie

nts

who

are

redu

cibl

edo

bette

r).

Bur

kean

dB

erry

man

,19

71

(3)

41pa

tient

str

eate

dby

MU

A.

Ligh

ttr

actio

nan

dth

enin

duct

ion

ofan

esth

esia

and

intu

batio

n,th

en

MU

Aif

nece

ssar

y(s

ame

asEv

ans

[10]

).

32pa

tient

str

eate

dw

ithtr

actio

nal

one.

3tr

eate

dby

trac

tion

afte

rM

UA

faile

d.

C7–

T1no

tat

tem

pted

.Pr

e-M

RI.

37/4

1su

cces

sful

lyre

duce

dby

MU

A.

21/2

5re

duce

dw

ithtr

actio

nbe

fore

anes

thet

ic.

7pa

tient

sw

ere

judg

edto

osi

ckfo

ran

esth

esia

and

unde

rwen

ttr

actio

nfo

rst

abili

zatio

n,no

tre

duce

d.

2ca

ses

ofne

urol

ogic

alde

teri

orat

ion:

1ov

erdi

stra

ctio

n,1

unre

cogn

ized

inju

ry.

IIIM

UA

and

trac

tion

both

safe

ifpr

oper

diag

nosi

san

d

care

ful

atte

ntio

npa

idto

radi

ogra

phs.

Evan

s,19

61(1

0)17

patie

nts.

Trea

ted

byin

duct

ion

ofan

esth

esia

and

intu

batio

n,

som

etim

esw

ithm

anip

ulat

ion

unde

ran

esth

esia

.Pr

e-M

RI.

No

neur

olog

ical

dete

rior

atio

nno

ted.

All

succ

essf

ully

redu

ced.

2un

chan

ged,

2di

ed,

13

impr

oved

.

IIIR

educ

tion

unde

ran

esth

esia

safe

and

effe

ctiv

e.

Smal

lse

ries

.

aM

RI,

mag

neti

cre

sona

nce

imag

ing;

ASI

A,A

mer

ican

Spin

alIn

jury

Ass

ocia

tion

;MU

A,m

anip

ulat

ion

unde

ran

esth

esia

;RT,

rapi

dtr

acti

on-r

educ

tion

;CT,

com

pute

dto

mog

raph

y;O

R,

open

redu

ctio

n;O

RIF

,op

enre

duct

ion

wit

hin

tern

alfix

atio

n;C

R,

clos

edre

duct

ion;

C-s

pine

,ce

rvic

alsp

ine.

Initial Closed Reduction of Fracture-Dislocation Injuries S47


vest application after closed reduction has been accomplished(14, 15, 18, 40).

Hadley et al. (14) suggest that closed reduction of acutecervical spine fracture-dislocation injuries is best accom-plished as part of the early overall medical management of thepotential SCI patient in the intensive care unit setting usingbedside fluoroscopy, with close monitoring of each patient’sclinical and neurological status, as well as cardiac, respiratory,and hemodynamic parameters. Pain control and modestsedation-relaxation is provided using short-acting intrave-nous agents that do not impair the patient’s level of conscious-ness or alter hemodynamic performance parameters (14, 18).

Craniocervical traction is typically arbitrarily initiated withthe application of weight beginning with 3 pounds per supe-rior injury level. By this scheme, a patient with an isolatedC5–C6 fracture-dislocation injury would begin treatment withan initial weight of 15 pounds (3 lb � 5 rostral vertebrallevels). Caveats to the use of this initial weight suggestioninclude patients with fracture-dislocations involving C2 andabove and patients with ankylosing spondylitis in whom verylittle weight, if any, may be needed to accomplish reduction.Judgment and experience must be used in this setting becausesome more-proximal cervical spine injuries may be distractioninjuries and will not require traction. These injuries are per-haps best managed with realignment and compression.

Weight may be added as often as every 10 to 15 minutes aslong as close clinical, neurological, and radiographic monitor-ing is reassessed throughout the process. No upper limit ofweight has been described in the literature. In general, in-creasing weight is applied until closed reduction and realign-ment occurs or until patient complaints are great (increasingintractable pain or a subjective worsening of neurologicalstatus), the patient’s neurological examination worsens,overdistraction occurs as noted on fluoroscopy or lateral cer-vical spine x-rays, it is impractical to add further weight (noadditional weight can be applied to the distraction device,no additional weight available, patient sliding upward inbed), or the clinician judges that closed reduction has failed.

After closed reduction has been accomplished, or once thedetermination has been made that closed reduction has failed,it is recommended that the patient be immobilized to providestabilization of the cervical spine injury before transport toradiology for further assessment or to the operating room forsurgical management (14, 18).

Herniated disc fragments causing compression of the cer-vical spinal cord at the level of facet dislocation have beendescribed by several authors (3, 8, 11, 16, 20, 22, 25, 28, 37).Eismont et al. (9) reported a series of 63 patients managedwith closed traction-reduction and then by open reduction ifclosed reduction was unsuccessful. One of these patientsworsened after posterior open reduction and fusion. A herni-ated disc was found ventral to the cord on postproceduremyelography. Herniated discs were found in three other pa-tients in whom closed reduction failed and in two patientswith static neurological deficits after fracture-dislocation re-duction (one open, one closed). One of these patients deteri-orated after subsequent anterior cervical discectomy and fu-sion. Eismont et al. (9) did not report their overall results with

closed reduction. However, it is clear from their case descrip-tions that no awake patient experienced neurological deteri-oration as a result of attempted closed reduction. Olerud andJohnson (26) described two patients found to have disc her-niations on postreduction MRI or CT myelography. Both pa-tients deteriorated after open reduction after failure of at-tempted closed reduction. Robertson and Ryan (29) reportedthree patients who deteriorated during management of cervi-cal subluxation injuries. One of their patients worsened dur-ing transport to the hospital. That patient’s vertebral injurywas found to have spontaneously partially reduced. MRIrevealed a disc fragment compressing the cord. A secondpatient deteriorated after posterior open reduction. MRI re-vealed disc fragments compressing the cord. The patient un-derwent subsequent anterior decompression. The third pa-tient deteriorated 3 days after successful closed reduction. Asubsequent MRI study demonstrated disc material compress-ing the ventral cervical spinal cord. Five days after deteriora-tion, the patient underwent anterior decompression. All threepatients recovered to at least their pre-deterioration neurolog-ical examination. Grant et al. (13) reported a single case ofneurological deterioration in their series associated withclosed reduction that also occurred in a delayed fashion (6 hafter reduction). This occurred in a patient subsequentlyfound to have a herniated disc at the level of injury.

Mahale et al. (24) reviewed 16 cases of neurological deteri-oration in patients with cervical spinal cord injuries afterreduction of facet dislocations. Seven of the 16 patients devel-oped complete cord injuries, 6 after open reduction, and 1after manipulation under anesthesia. Five patients developedpartial injuries, three after manipulation under anesthesia andtwo after closed traction-reduction. Of the two patients whodeteriorated after closed reduction, one was found to beoverdistracted. Minor injuries were sustained by the remain-ing four patients, including one who deteriorated when thecranial traction pins slipped, one who deteriorated in a plasterbrace, one who lost reduction and had neurological worsen-ing, and one patient who underwent open reduction. Sixpatients were investigated with myelography after deteriora-tion, two with MRI, and one with CT. A disc protrusion wasnoted in one patient, and a “disc prolapse with hematoma”was noted in another. Both of these patients were treatedconservatively. The most common imaging finding in thesenine patients was cord edema (20).

The prevalence of MRI-documented disc herniation in as-sociation with cervical facet injury with subluxation hascaused a number of authors to recommend prereduction MRIin patients with these injuries. Harrington et al. (16) reporteda series of 37 patients managed with closed reduction, inwhom a 97% rate of successful reduction was achieved withno neurological morbidity. Postreduction imaging revealeddisc herniations in nine patients, four of whom underwentlater anterior decompression. Doran et al. (8) reported a seriesof 13 patients drawn from four institutions during an unspec-ified time period. Nine patients were treated with attemptedearly closed reduction. Subluxations in three patients werereduced without incident; failure to reduce was noted inthree patients. Closed reduction was abandoned in an-



other three patients because of worsening pain (one patient)or arm weakness (two patients). All patients underwent MRIevaluation (four prereduction). Herniated discs were visual-ized in 10 patients, bulging discs were imaged in 3. No treatedpatient developed a permanent neurological deficit as a resultof attempted closed reduction. No patient who underwentsuccessful closed reduction deteriorated. All three who hadinjuries that were successfully reduced showed significantneurological improvement despite the MRI appearance of adisc herniation in two and a disc bulge in the third. Theauthors concluded that prereduction MRI studies should beobtained before closed reduction in patients with cervicalspine facet dislocation injuries. Because there was no correla-tion between the presence of disc herniation and neurologicaldeterioration, the recommendation for a prereduction MRI inpatients with cervical facet dislocation injuries cannot besupported.

Vaccaro et al. (37) studied 11 consecutive patients with MRIbefore and after reduction. The authors found herniated discsin two patients in the prereduction group and in five of ninepatients who underwent successful closed reduction. Thepresence of a herniated disc on prereduction MRI or pos-treduction MRI did not predict neurological deterioration. Nocase of deterioration after successful reduction occurred (37).Grant et al. (13) obtained postreduction MRI studies on 80patients treated with closed reduction and found herniated orbulging discs in 46%. They found no correlation between MRIresults and neurological outcome. Rizzolo et al. (28) per-formed MRI prereduction on 55 patients with cervical frac-tures and dislocation injuries. They found evidence of discherniation in 54% of these patients. Awake and alert patientsunderwent closed traction-reduction. There were no instancesof neurological deterioration in this group. The authors didnot attempt closed reduction in patients who were not awake.

In summary, a review of the literature reveals only twodocumented cases (11, 25) of neurological deterioration asso-ciated with attempted closed reduction of cervical spinefracture-dislocation injuries caused by cord compression fromdisc herniation. Both of these cases were characterized bydeterioration hours to days after closed reduction. A numberof large clinical series have failed to establish a relationshipbetween the presence of a prereduction herniated disc andsubsequent neurological deterioration with attempted closedtraction-reduction in awake patients.

SUMMARY

Closed reduction of fracture-dislocation injuries of the cer-vical spine by traction-reduction seems to be safe and effectivefor the reduction of spinal deformity in awake patients. Ap-proximately 80% of patients will have their injuries reducedwith this technique. The overall permanent neurological com-plication rate of closed reduction is approximately 1%. Theassociated risk of a transient injury with closed reductionseems to be 2 to 4%. Closed traction-reduction seems to besafer than manipulation under anesthesia.

There are numerous causes of neurological deterioration inpatients with unstable cervical spine injuries. These include

inadequate immobilization, unrecognized rostral injuries,overdistraction, loss of reduction, and cardiac, respiratory,and hemodynamic instability. Therefore, the treatment of cer-vical spine fracture-dislocation injuries must be supervised byan appropriately trained specialist.

Although prereduction MRI will demonstrate disc herniationin up to half of patients with facet subluxation injuries, theclinical importance of these findings is questionable. Only twocase reports exist that document neurological deteriorationcaused by disc herniation after successful closed traction-reduction in awake patients. Both occurred in delayed fashionafter closed reduction. In addition, several investigators havedemonstrated the lack of correlation between the MRI finding ofdisc herniation and neurological deterioration in this patientpopulation. The use of prereduction MRI has therefore not beenshown to improve the safety or efficacy of closed traction-reduction in awake patients. MRI before fracture-dislocationreduction may result in unnecessary delays in accomplishingfracture realignment and decompression of the spinal cord. Be-cause Class III evidence exists in support of early closed reduc-tion of cervical fracture-dislocation injuries with respect to neu-rological function, prereduction MRI in this setting is notnecessary. The ideal timing of reduction is unknown, but manyinvestigators favor reduction as rapidly as possible after injury tomaximize the potential for neurological recovery.

Patients in whom attempted closed reduction of cervical frac-ture injuries fails have a higher incidence of anatomic obstaclesto reduction, including facet fractures and disc herniation. Pa-tients in whom closed reduction fails should undergo moredetailed radiographic study before attempts at open reduction.The presence of a significant disc herniation in this setting is arelative indication for an anterior decompression procedure, ei-ther in lieu of or preceding a posterior procedure.

Patients with cervical fracture-dislocation who cannot be ex-amined, because of head injury or intoxication, cannot be as-sessed for neurological deterioration during attempted closedtraction-reduction. For this reason, an MRI study before at-tempted reduction is recommended as a treatment option.


A prospective cohort study of patients with cervical spinalcord injuries caused by facet fracture-subluxation injuries treatedwith or without prereduction MRI would provide Class II med-ical evidence in support of a treatment recommendation on thisissue. This could address issues of timing. A randomized con-trolled trial may provide Class I medical evidence.

No prospective comparative study has been performed ofclosed reduction versus anterior decompression and stabili-zation for patients with MRI-documented herniated discs inassociation with unreduced cervical fracture-dislocation inju-ries. A prospective study of this issue would provide Class IImedical evidence in support of a treatment recommendation.


Initial Closed Reduction of Fracture-Dislocation Injuries S49


REFERENCES

1. Alexander E Jr, Davis CH Jr, Forsyth HF: Reduction and fusion offracture dislocation of the cervical spine. J Neurosurg 27:588–591,1967.

2. Brunette DD, Rockswold GL: Neurologic recovery followingrapid spinal realignment for complete cervical spinal cord injury.J Trauma 27:445–447, 1987.

3. Burke DC, Berryman D: The place of closed manipulation in themanagement of flexion-rotation dislocations of the cervical spine.J Bone Joint Surg Br 53B:165–182, 1971.

4. Cloward RB: Reduction of traumatic dislocation of the cervicalspine with locked facets: Technical note. J Neurosurg 38:527–531,1973.

5. Cotler JM, Herbison GJ, Nasuti JF, Ditunno JF Jr, An H, Wolff BE:Closed reduction of traumatic cervical spine dislocation usingtraction weights up to 140 pounds. Spine 18:386–390, 1993.

6. Cotler HB, Miller LS, DeLucia FA, Cotler JM, Dayne SH: Closedreduction of cervical spine dislocations. Clin Orthop 214:185–199,1987.

7. Crutchfield W: Skeletal traction in treatment of injuries to thecervical spine. JAMA 155:29–32, 1954.

8. Doran SE, Papadopoulos SM, Ducker TB, Lillehei KO: Magneticresonance imaging documentation of coexistent traumatic lockedfacets of the cervical spine and disc herniation. J Neurosurg79:341–345, 1993.

9. Eismont FJ, Arena MJ, Green BA: Extrusion of an intervertebraldisc associated with traumatic subluxation or dislocation of cer-vical facets. J Bone Joint Surg Am 73A:1555–1560, 1991.

10. Evans D: Reduction of cervical dislocations. J Bone Joint Surg Br43B:552–555, 1961.

11. Farmer J, Vaccaro A, Albert TJ, Malone S, Balderston RA, CotlerJM: Neurologic deterioration after spinal cord injury. J SpinalDisord 11:192–196, 1998.

12. Fehlings MG, Tator CH: An evidence based review of decompres-sive surgery in acute spinal cord injury: Rationale, indications,and timing based on experimental and clinical studies.J Neurosurg 91[Suppl 1]:1–11, 1999.

13. Grant GA, Mirza SK, Chapman JR, Winn HR, Newell DW, JonesDT, Grady MS: Risk of early closed reduction in cervical spinesubluxation injuries. J Neurosurg 90[Suppl 1]:13–18, 1999.

14. Hadley MN, Argires PJ: The acute/emergent management ofvertebral column fracture dislocation injuries, in NeurologicalEmergencies. Park Ridge, AANS, 1994, vol 2, pp 249–262.

15. Hadley MN, Fitzpatrick BC, Sonntag VKH, Browner CM: Facetfracture-dislocation injuries of the cervical spine. Neurosurgery30:661–666, 1992.

16. Harrington JF, Likavec MJ, Smith AS: Disc herniation in cervicalfracture subluxation. Neurosurgery 29:374–379, 1991.

17. Key A: Cervical spine dislocations with unilateral facet interlock-ing. Paraplegia 13:208–215, 1975.

18. Kilburn MP, Smith DP, Hadley MN: The initial evaluation andtreatment of the patient with spinal trauma, in Batjer HH, LoftusCM (eds): Textbook of Neurological Surgery: Principles and Practice.Philadelphia, Lippincott Williams & Wilkins (in press).

19. Kleyn PJ: Dislocations of the cervical spine: Closed reductionunder anaesthesia. Paraplegia 22:271–281, 1984.

20. Lee AS, MacLean JC, Newton DA: Rapid traction for reduction ofcervical spine dislocations. J Bone Joint Surg Br 76B:352–356, 1994.

21. Lu K, Lee TC, Chen HJ: Closed reduction of bilateral locked facetsof the cervical spine under general anaesthesia. Acta Neurochir(Wien) 40:1055–1061, 1998.

22. Ludwig SC, Vaccaro AR, Balderston RA, Cotler JM: Immediatequadriparesis after manipulation for bilateral cervical facet sub-luxation. J Bone Joint Surg Am 79A:587–590, 1997.

23. Mahale YJ, Silver JR: Progressive paralysis after bilateral facetdislocation of the cervical spine. J Bone Joint Surg Br 74B:219–223, 1992.

24. Mahale YJ, Silver JR, Henderson NJ: Neurological complicationsof the reduction of cervical spine dislocations. J Bone Joint SurgBr 75B:403–409, 1993.

25. Maiman DJ, Barolat G, Larson SJ: Management of bilateral lockedfacets of the cervical spine. Neurosurgery 18:542–547, 1986.

26. Olerud C, Jonsson H Jr: Compression of the cervical spinal cordafter reduction of fracture dislocations: Report of 2 cases. ActaOrthop Scand 62:599–601, 1991.

27. Osti OL, Fraser RD, Griffiths ER: Reduction and stabilisation ofcervical dislocations: An analysis of 167 cases. J Bone Joint SurgBr 71B:275–282, 1989.

28. Rizzolo SJ, Piazza MR, Cotler JM, Balderston RA, Schaefer D,Flanders A: Intervertebral disc injury complicating cervical spinetrauma. Spine 16(6 Suppl):S187–S189, 1991.

29. Robertson PA, Ryan MD: Neurological deterioration after reduc-tion of cervical subluxation: Mechanical compression by disctissue. J Bone Joint Surg Br 74B:224–227, 1992.

30. Rosenfeld JF, Vaccaro AR, Albert TJ, Klein GR, Cotler JM: Thebenefits of early decompression in cervical spinal cord injury.Am J Orthop 27:23–28, 1998.

31. Sabiston CP, Wing PC, Schweigel JF, Van Peteghem PK, Yu W:Closed reduction of dislocations of the lower cervical spine.J Trauma 28:832–835, 1988.

32. Schaefer DM, Flanders A, Northrup BE, Doan HT, Osterholm JL:Magnetic resonance imaging of acute cervical spine trauma.Spine 14:1090–1095, 1989.

33. Shrosbree RD: Neurological sequelae of reduction of fracturedislocations of the cervical spine. Paraplegia 17:212–221, 1979.

34. Sonntag VKH: Management of bilateral locked facets of the cer-vical spine. Neurosurgery 8:150–152, 1981.

35. Star AM, Jones AA, Cotler JM, Balderston RA, Sinha R: Immedi-ate closed reduction of cervical spine dislocations using traction.Spine 15:1068–1072, 1990.


37. Vaccaro AR, Falatyn SP, Flanders AE, Balderston RA, NorthrupBE, Cotler JM: Magnetic resonance evaluation of the interverte-bral disc, spinal ligaments, and spinal cord before and after closedtraction-reduction of cervical spine dislocations. Spine 24:1210–1217, 1999.

38. Vital JM, Gille O, Senegas J, Pointillart V: Reduction technique foruni- and biarticular dislocations of the cervical spine. Spine 23:949–955, 1998.

39. Walton G: A new method of reducing dislocation of cervicalvertebrae. J Nerv Ment Dis 20:609, 1893.

40. Wilberger JE: Immobilization and traction, in Tator CH, BenzelEC (eds): Contemporary Management of Spinal Cord Injury: FromImpact to Rehabilitation. Park Ridge, AANS, 2000, pp 91–98.

41. Xiong XH, Bean A, Anthony A, Inglis G, Walton D: Manipulationfor cervical spinal dislocation under general anaesthesia: Serialreview for 4 years. Spinal Cord 36:21–24, 1998.

42. Yashon D, Tyson G, Vise WM: Rapid closed reduction of cervicalfracture dislocations. Surg Neurol 4:513–514, 1975.



CHAPTER 7

Management of Acute Spinal Cord Injuries in an IntensiveCare Unit or Other Monitored Setting

RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS:• Management of patients with acute spinal cord injury, particularly patients with severe cervical level

injuries, in an intensive care unit or similar monitored setting is recommended.• Use of cardiac, hemodynamic, and respiratory monitoring devices to detect cardiovascular dysfunction and

respiratory insufficiency in patients after acute cervical spinal cord injury is recommended.

RATIONALE

The intensive care unit (ICU) setting has traditionally beenreserved for critically ill patients who require aggressivemedical care and exceptional medical attention. Most con-

temporary medical centers have multiple critical care units; eachunit is designed to provide discipline-specific observation andintensive care to patients in need. Select institutions have createdacute spinal cord injury centers and offer multidisciplinary care,including ICU care, to patients who have sustained acute spinalcord injuries (ASCIs) (2, 11, 12, 16, 21, 22, 25, 30–32, 34). Severalreports describe improved patient management and lower mor-bidity and mortality after ASCIs with ICU monitoring and ag-gressive medical management (11, 12, 16, 22, 25, 31, 32, 34).Despite this interest in and commitment to more comprehensivecare for the patient with an ASCI over the last 30 years by someindividuals and centers, many patients who sustain an ASCI arenot managed in an ICU setting, nor are they routinely monitoredfor cardiac or respiratory dysfunction. There exist divergentmanagement strategies for ASCI patients within regions, com-munities, even within institutions, depending on the trainingand experiences of the clinicians providing care. Recently com-pleted randomized clinical trials to investigate pharmacologicalagents in the treatment of ASCI patients did not suggest aspecific, common medical management paradigm to guide pa-tient care provided by participating investigators, other than thetiming and dosage of the pharmacological agents being tested(4–7, 9, 10). These studies included large numbers of seriouslyinjured ASCI patients managed outside the ICU setting, mostwithout continuous cardiac or respiratory monitoring.

QUESTIONS

1. Do patients with acute spinal cord injuries benefit fromcare in the ICU setting?

2. Is monitoring of cardiac, hemodynamic, and pulmonaryperformance of benefit to patients who have sustained anASCI?

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The following medical subject headings wereused in combination with “spinal cord injury”: medical man-agement, nonoperative management, surgical management,hypotension, respiratory insufficiency, pulmonary complica-tions, and ICU. Approximately 3400 citations were acquired.Non-English language citations were deleted. Titles and ab-stracts of the remaining publications were reviewed, and rel-evant articles were selected to develop this guideline. Wefocused on four specific topics concerning human patientswith ASCIs: management in an ICU (18 articles reviewed),cardiac instability (8 articles reviewed), hypotension (22 arti-cles reviewed), and respiratory/pulmonary dysfunction (20articles reviewed). Additional references were culled from thereference lists of the remaining papers. Finally, members ofthe author group were asked to contribute articles known tothem on the subject matter that were not found by othersearch means. Articles describing nonhuman laboratory in-vestigations germane to the topic and related general reviewarticles referenced in the Scientific Foundation are includedamong the 34 references in the reference list. Articles describ-ing economics, epidemiology, anesthesia, monitoring tech-niques, penetrating cord injuries, nursing care, infectious orurological complications, chronic complications, or patientswith remote spinal cord injuries (SCIs) were not included.These efforts resulted in 17 manuscripts, all of which arereports of case series (Class III medical evidence), which form


the foundation for this guideline. These articles are summa-rized in Table 7.1.

SCIENTIFIC FOUNDATIONThe pathophysiology of ASCI is complex and multifaceted.

It involves a primary mechanical injury by way of compres-sion, penetration, laceration, shear, and/or distraction. Theprimary injury seems to initiate a host of secondary injury

mechanisms, including: 1) vascular compromise leading toreduced blood flow, loss of autoregulation, loss of microcir-culation, vasospasm, thrombosis, and hemorrhage; 2) electro-lyte shifts, permeability changes, loss of cellular membraneintegrity, edema, and loss of energy metabolism; and 3) bio-chemical changes including neurotransmitter accumulation,arachidonic acid release, free radical and prostaglandin pro-duction, and lipid peroxidation (1, 13, 26–29). These mecha-

TABLE 7.1. Summary of Reports on Management of Patients with Acute Spinal Cord Injuries in an Intensive CareUnit Settinga


Lu et al., 2000 (18) Retrospective review of apnea in 36 ASCI patients. III Delayed apnea most likely in ASCI patients with severe,diffuse ASCI. Apnea most likely within first 7–10 d.

Botel et al., 1997 (3) 225 ASCI patients treated in ICU. Only 87 admittedwithin 24 h of injury.

III Significant numbers of multiply injured and head-injuredpatients. No complete injury recorded. Improved outcomewhen admitted to ICU early after injury.

Vale et al., 1997 (32) Prospective assessment of 77 ASCI patients treated inICU, aggressive hemodynamic support, MAP �85.

III Improved outcome with aggressive medical care, distinctfrom potential benefit from surgery at 1-yr follow-up.

Levi et al., 1993 (16) 50 patients treated in ICU, aggressive medicaltreatment, MAP �90.

III Improved outcome with aggressive hemodynamic support at6 wk postinjury.

Tator et al., 1993 (30) 201 ASCI patients, ICU care, hemodynamic supportcompared with 351 earlier patients.

III Less severe cord injuries due to immobilization,resuscitation, and early transfer to ICU setting.

Levi et al., 1991 (17) 103 ASCI patients: 50 incomplete (Group A), 53complete (Group B). ICU care, hemodynamicsupport, MAP �85.

III Improved neurological outcome, no significant differencebetween early and late surgery in either group.

Wolf et al., 1991 (33) 52 patients with locked facets reduced within 4 h,ICU care, MAP �85. 49 operated on, 23 Day 1, 26delayed (8.7 d mean).

III Closed reduction, 61%.Closed (a), 15%.52% follow-up at 1 y; in general, improved neurologicaloutcome.

Lehmann et al., 1987 (15) 71 consecutive ASCI patients, ICU care, monitoringof cardiac/hemodynamic parameters.

III Bradycardia, 100%; hypotension (�90 systolic); 68% life-threatening bradyarrhythmias; 16% incidence related toseverity of SCI.

Reines and Harris, 1987 (22) 123 cases. ASCI patients in ICU, aggressivepulmonary treatment.

III Respiratory insufficiency major cause of morbidity andmortality after ASCI. Aggressive ICU care, pulmonarytreatment reduces incidence.

Piepmeier et al., 1985 (21) 45 ASCI patients, all managed in ICU setting withcardiac, hemodynamic monitoring.

III Cardiac dysrhythmia, hypotension, and hypoxia common infirst 2 wk after ASCI. Incidence related to severity of injury.

Bose et al., 1984 (2) 28 ASCI patients, 22 managed in ICU setting.Group I: medical treatment.Group II: medical/surgical treatment.

III Improved neurological outcome at discharge for Group IIbut better scores initially. Group I with intrinsic cord injuryversus Group II compression on myelography and/orinstability.

Tator et al., 1984 (31) 144 ASCI patients. ICU care, hemodynamic support,compared with prior series.

III Improved neurological outcome, less mortality with earlytransfer, avoidance of hypotension, and ICU care.

Ledsome and Sharp, 1981 (14) Reassessment of pulmonary function in ASCI patients,comparison over time.

III Reduced vital capacity, flow rates, and hypoxia after ASCI.Incidence related to severity of SCI. Marked improvement inpulmonary functions 3 mo postinjury.

McMichan et al., 1980 (20) Prospective study of pulmonary complications in 22ASCI patients compared with 22 earlier patientsmanaged with aggressive ICU care.

III No deaths in series versus 9 of 22 deaths in earlier group.ICU care and vigorous pulmonary therapy improvessurvival, reduces complications.

Gschaedler et al., 1979 (11) 51 ASCI patients managed in ICU, aggressive medicaltreatment, avoidance of hypotension.

III Improved morbidity and mortality with early transfer,avoidance of hypotension, respiratory insufficiency.

Hachen, 1977 (12) 188 ASCI patients managed in ICU, aggressivetreatment of hypotension, respiratory insufficiency.

III Reduced morbidity and mortality with early transfer,attentive ICU care and monitoring, and aggressive treatmentof hypotension and respiratory failure.

Zach et al., 1976 (34) 117 ASCI patients at Swiss Center, ICU setting,aggressive blood pressure and volume therapy.Rheomacrodex � 5 d.Dexamethasone � 10 d.

III Improved neurological outcome with aggressive medicaltreatment. Better outcome for early referrals.

a ASCI, acute spinal cord injury; ICU, intensive care unit; MAP, mean arterial pressure; closed (a), closed reduction under anesthesia.



nisms, if unchecked, result in axonal disruption and cellulardeath. A number of contemporary reviews describe thesetheories and provide experimental evidence in their support(1, 26–28).

Animal models of SCI suggest that ischemia of the spinalcord underlies much of the mechanism of posttraumatic SCIand is the important common denominator resulting in neu-rological deficit after primary injury (1, 26, 27). Ischemiaseems to be related to both local and systemic vascular alter-ations after severe injury. Local vascular alterations are due tothe direct SCI and focal, postinjury vasospasm, both of whichlead to loss of autoregulation of spinal cord blood flow (1, 8,13, 23, 24, 26, 27, 29). Systemic vascular alterations of bloodflow to the spinal cord after ASCI observed in both animalstudies and in human SCI patients include reduced heart rate,cardiac rhythm irregularities, reduced mean arterial bloodpressure, reduced peripheral vascular resistance, and com-promised cardiac output (1, 8, 15, 17, 18, 21, 24, 26, 27, 29, 32).Any of these untoward hemodynamic occurrences can con-tribute to systemic hypotension after severe injury (15–17, 21,25–27, 32). Systemic hypotension in the setting of ASCI, withcoincident loss of spinal cord autoregulatory function, com-pounds local spinal cord ischemia by further reducing spinalcord blood flow and perfusion (1, 26, 27, 29).

Respiratory insufficiency and pulmonary dysfunction iscommon after traumatic SCI, particularly when the injuryoccurs at cervical spinal cord levels (11, 12, 14, 18–20, 22, 25).Severely injured patients demonstrate marked reductions inexpected vital capacity and inspiratory capacity and mayexperience relative hypoxemia, all of which contribute toglobal hypoxemia and can exacerbate spinal cord ischemiaafter acute injury (14, 18–20, 22). It seems that the earliercardiac and/or ventilatory/pulmonary dysfunction is de-tected, the more likely effective, often life-saving treatmentcan be initiated. It is for these reasons that the issues of earlyICU care and cardiac and pulmonary monitoring for humanpatients after ASCI have been raised.

Several clinical series have been reported in which humanpatients with ASCIs have been managed in ICU environmentswith attention to heart rate, cardiac function, pulmonary per-formance, and mean arterial blood pressure (2, 3, 11, 12,14–22, 25, 31–34). Zach et al. (34), in 1976, provided a prelim-inary report on their prospective medical management para-digm in the treatment of 117 consecutive ASCI patients in theSwiss Paraplegic Center of Basel, Switzerland. All patientswere treated in the ICU with central venous pressure moni-toring and were given dexamethasone, 0.5 mg/kg for 4 days,with a tapering dose through 10 days, and volume expansionwith Rheomacrodex 40 (Medisan, Parsippany, NJ), 500 ml/dfor 7 days. Patients were stratified by injury level, degree ofdeficit (Frankel grade), and time of admission after injury. Theauthors reported that 62% of cervical level SCI patients theymanaged in this way improved at last follow-up, including 8of 18 Frankel Grade A patients, two patients by two gradesand one patient by three grades. No patient with a cervicalinjury worsened; 38% were unchanged from admission. Pa-tients with thoracic T1–T10 level SCIs fared less well; 38%improved, none worsened, and 62% were without change,

including 22 of 26 Frankel Grade A patients. Two FrankelGrade A patients experienced a complete recovery. Seventypercent of acute T11–L1 level SCIs improved with this treat-ment paradigm, none worsened, and 30% were unchangedfrom admission. Of patients who arrived within 12 hours ofinjury, 67% improved compared with their admission neuro-logical examination. Of patients admitted between 12 and 48hours after injury, only 59% improved. When admission oc-curred 48 hours after injury, improvement was seen in only50% of patients. The authors concluded that early transfer andimmediate medical specific treatment of the spinal injuryseemed to improve neurological recovery (34).

Hachen (12), in 1977, reported the 10-year experience withacute traumatic tetraplegia from the National Spinal InjuriesCenter in Geneva, Switzerland. He described 188 ASCI pa-tients treated in a 10-year period in the ICU setting afterimmediate transfer from the scene of the injury. The centerreported a marked reduction in mortality rates after acutecervical SCI compared with annual statistics from 1966. Mor-tality for complete tetraplegia was reduced from 32.5 to 6.8%over the 10-year period. Mortality for patients with incom-plete tetraplegia dropped from 9.9% in 1966 to 1.4% in 1976.Most early deaths in the center’s experience were related topulmonary complications. The likelihood of severe respira-tory insufficiency was related to the severity of the cervicalSCI. Seventy percent of patients with complete lesions expe-rienced severe respiratory insufficiency in the center’sexperience, compared with 27% of patients with incompletelesions. The improvement in mortality rates described wasdirectly related to early monitoring and treatment of respira-tory insufficiency in the ICU setting. Hachen stressed thatfacilities for continuous monitoring of central venous pres-sure, arterial pressure, pulse, respiration rate and pattern, andoxygenation-perfusion parameters must be available for allpatients with neurological injuries after ASCI, particularlythose injuries above the C6 level.

In 1979, Gschaedler et al. (11) described the comprehensivemanagement of 51 patients with acute cervical SCI in the ICUsetting in Colmar, France. Forty percent of the patients theymanaged had multiple organ system injuries. The authorsreported a low mortality rate (7.8%) and described severalseverely injured patients who made important neurologicalimprovements, including one patient improving from FrankelGrade A to Grade D and two patients improving fromFrankel Grade B to Grade D. The authors cited early transportafter injury and comprehensive intensive medical care withattention to and avoidance of hypotension and respiratoryinsufficiency as essential to the improved management out-come they experienced.

McMichan et al. (20) reported in 1980 their prospectiveassessment of pulmonary complications identified in 22 pa-tients with cervical level ASCI managed in the ICU setting.They compared their results with 22 retrospective patientswith similar injuries. Use of a new, aggressive pulmonarytreatment paradigm resulted in no deaths and fewer respira-tory complications compared with those experienced by theretrospective group (nine deaths). The authors concluded thatvigorous pulmonary therapy initiated early after ASCI was

Management of Acute Spinal Cord Injuries in an ICU Setting S53


associated with increased survival, a reduced incidence ofpulmonary complications, and a decreased need for ventila-tory support.

Ledsome and Sharp (14) measured pulmonary function in16 cervical level complete ASCI patients and compared initialvalues with those obtained in the same patients at 1, 3, and 5weeks and 3 and 5 months after injury. In their 1981 report,they noted profound reductions in forced vital capacity (FVC)and expiratory flow rates immediately after injury. Patientswith an FVC less than 25% of expected had a high incidenceof respiratory failure requiring ventilator support. This wasespecially true of patients with injuries at C4 or above. Theauthors found a significant increase in FVC at 5 weeks postin-jury and an approximate doubling of FVC at 3 months, irre-spective of the level of cervical cord injury. Importantly, theyidentified hypoxemia (PO2 � 80 mm Hg) in most of theirpatients (74% of those who did not require ventilator sup-port), despite adequate alveolar ventilation (PCO2 normaldespite low FVC). Ledsome and Sharp attributed this to aventilation perfusion imbalance that occurs immediately afterASCI. Systemic hypoxemia was identified by blood gas mea-surements and was effectively treated with the addition ofsupplemental oxygen in most patients.

Piepmeier et al. (21) identified cardiovascular instabilityafter acute cervical SCI in 45 patients they managed in the ICUsetting in New Haven, CT. Twenty-three patients had FrankelGrade A injuries, eight were Grade B, seven Grade C, andseven Grade D. The authors discovered a high incidence ofcardiovascular irregularities in these patients and identified adirect correlation between the severity of the cord injury andthe incidence and severity of cardiovascular problems. Threepatients returned to the ICU setting during the 2-week obser-vation period of the study because of cardiac dysfunction,despite a period of initial stability. Of the 45 patients, 29 hadan average daily pulse rate of less than 55 beats per minute; 32had episodes during which their pulse rate was below 50beats per minute for prolonged periods. Hypotension wascommon after ASCI in their series, but most patients re-sponded well to volume replacement. Nine patients requiredvasopressors to maintain a systolic pressure above 100mm Hg, therapy that ranged from hours to 5 days duration.Cardiac arrest occurred in five (11%) patients. All had FrankelGrade A injuries. Three arrests occurred during endotrachealsuctioning. The authors found that the first week after injurywas the timeframe during which patients were most vulner-able to cardiovascular instability. Patients with the most se-vere neurological injuries were most likely to experience car-diovascular instability after ASCI. These events occurreddespite the absence of complete autonomic disruption. Hyp-oxia and endotracheal suctioning were associated with car-diac arrest in most instances. Piepmeier et al. concluded thatcareful monitoring of severely injured ASCI patients in theICU setting reduces the risk of life-threatening emergencies.

In 1984, Tator et al. (31) described their experience with 144patients with ASCIs managed between 1974 and 1979 at adedicated SCI unit at Sunnybrook Medical Centre in Toronto,Ontario, Canada. The authors compared their results with acohort of 358 SCI patients managed between 1948 and 1973,

before the development of the acute care SCI facility. All 144patients managed from 1974 to 1979 were treated in an ICUsetting with strict attention to the treatment of hypotensionand respiratory failure. Their medical paradigm was devel-oped on the principle that avoiding hypotension is one of themost important aspects of the immediate management ofacute cord injury (31). Hypotension was treated vigorouslywith crystalloid and transfusion of whole blood or plasma forvolume expansion. Patients with respiratory dysfunctionwere treated with ventilatory support as indicated. They re-ported that mean time from injury to admission and treatmentwas 4.9 hours during the period 1974 to 1979 compared withmore than 12 hours during the period 1948 to 1973. Neuro-logical improvement was observed in 41 (43%) of 95 patientsmanaged under the aggressive ICU medical paradigm; 52patients (55%) demonstrated no improvement; only 2 patients(2%) deteriorated. The authors reported lower mortality, re-duced morbidity, shorter length of stay, and lower cost oftreatment with their contemporary comprehensive manage-ment paradigm compared with the 1948 to 1973 experience.They cited improved respiratory management in their ICU asone of the principal factors responsible for reduced mortalityand credited the avoidance of hypotension, sepsis, and uro-logical complications for reduced morbidity after injury.These improved management results were realized despitethe fact that 28% of the ASCI patients they treated had addi-tional injuries that increased their risk of morbidity andmortality.

Lehmann et al. (15), in a follow-up study in 1987, reportedon 71 ASCI patients they managed in the ICU at the Yale/New Haven Medical Center. Patients were admitted within 12hours of injury and were stratified by level and severity ofneurological injury (Frankel scale). No patient had an associ-ated head injury, a history of diabetes mellitus, a preexistingcardiac disorder, or a history of cardiac medication use. Allwere monitored and aggressively treated to avoid hypoten-sion. The authors found that all patients with severe cervicalSCIs, Frankel Grades A and B, had persistent bradycardia,defined as a heart rate below 60 beats per minute for at least1 day. Thirty-five percent of Frankel Grade C and D patientswere identified to have persistent bradycardia. Only 13% ofthoracic and lumbar SCIs had this finding. Similarly, markedbradycardia, below 45 beats per minute, was frequent in thesevere cervical injury group (71%) and less common in themilder cervical injury (12%) and thoracolumbar injury pa-tients (4%). Sinus node slowing was often profound enough toproduce hemodynamic compromise and systemic hypoten-sion. Bolus injections of atropine or placement of a temporarypacemaker was often performed. This therapy was requiredby 29% of the severe cervical injury patients and by none inthe two other injury groups. Episodic hypotension unrelatedto hypovolemia was identified in 68% of the severe cervicalinjury group and in none of the other two injury groups.Thirty-five percent of the severe cervical injury group patientsrequired the use of intravenous pressors to maintain an ac-ceptable blood pressure. Five (16%) of 31 patients in thesevere injury group experienced a primary cardiac arrest,three of which were fatal. All five patients had Frankel Grade



A SCIs. No patient in the study experienced a significantcardiac rate disturbance or spontaneous episode of hypoten-sion beyond 14 days of injury. The authors concluded thatpotentially life-threatening cardiac arrhythmias and episodesof hypotension regularly accompanied acute severe injury tothe cervical spinal cord within the first 14 days of injury.These events were not solely attributable to disruption of theautonomic nervous system. Detection and treatment was bestaccomplished in the ICU setting.

Wolf et al. (33), in 1991, described their experience withbilateral facet dislocation injuries of the cervical spine at theUniversity of Maryland in Baltimore. The authors described52 patients with acute cervical trauma managed with an ag-gressive treatment paradigm that included ICU care, aggres-sive resuscitation, invasive monitoring, and hemodynamicmanipulation to maintain mean blood pressure above 85mm Hg for 5 days. Thirty-four patients had complete neuro-logical injuries, 13 had incomplete injuries, and 5 patientswere intact. The authors attempted closed reduction within 4hours of patient arrival to their center and performed earlyopen reduction on patients who could not be reduced byclosed means, including closed reduction under anesthesia.All but three patients underwent surgery for stabilization andfusion. The authors reported neurological improvement atdischarge in 21% of complete SCI patients and in 62% ofpatients with incomplete cervical SCI at admission. No intactpatient deteriorated. Only 52% 1-year follow-up was pro-vided. The authors concluded that their protocol of aggres-sive, early medical and surgical management of patients withASCIs improved outcome after injury. Treatment in the ICUsetting, hemodynamic monitoring with maintenance of meanarterial pressure, and early decompression of the spinal cordby open or closed means seemed to reduce secondary com-plications after ASCIs in their study.

Levi et al. (16) treated 50 acute cervical SCI patients in theICU at the University of Maryland in Baltimore according toan aggressive management protocol that included invasivehemodynamic monitoring and volume and pressor support tomaintain a hemodynamic profile with adequate cardiac out-put and mean blood pressure above 90 mm Hg. Their 1993report described 31 patients with Frankel Grade A injuries atadmission, 8 patients with Frankel Grade B injuries, and 11patients in Frankel Grades C and D. Eight patients had shockat the time of admission (systolic blood pressure �90mm Hg), and 82% of patients had volume-resistant hypoten-sion requiring pressors within the first 7 days of treatment.This was 5.5 times more common among patients with com-plete motor injuries. The authors reported that the overallmean peripheral vascular resistance index for the 50 patientsthey studied was less than the normal range, and it was lessthan the normal value in 58% of patients. Half of their ASCIpatients had a lower than normal systemic vascular resistanceindex. No patient with a complete motor deficit (FrankelGrades A and B) and marked deficits in indexes of peripheraland/or systemic vascular resistance experienced neurologicalrecovery at 6 weeks. At 6 weeks after injury, 40% of patientsmanaged by protocol improved, including several with com-plete injuries, 42% remained unchanged, and 18% (9 patients)

died. Minimal morbidity was associated with invasive hemo-dynamic monitoring. The authors concluded that hemody-namic monitoring in the ICU allows early identification andprompt treatment of cardiac dysfunction and hemodynamicinstability and can reduce the potential morbidity and mor-tality after ASCI.

Vale et al. (32), in 1997, reported their experience with anonrandomized, prospective pilot study in the assessment ofaggressive medical resuscitation and blood pressure manage-ment in 77 consecutive ASCI patients treated at the Universityof Alabama in Birmingham. There was no control group. Allpatients were managed in the ICU with invasive monitoring(Swan Ganz catheters and arterial lines) and blood pressureaugmentation to maintain mean arterial pressure (MAP)above 85 mm Hg for 7 days postinjury. The authors reported10 patients with complete cervical SCIs (American SpinalCord Injury Association [ASIA] Grade A), 25 patients withincomplete cervical injuries (ASIA Grades B, C, and D), 21patients with complete thoracic SCIs, and 8 patients withincomplete thoracic level SCIs (Grades B, C, and D). Theaverage admission MAP for Grade A cervical SCI patientswas 66 mm Hg. Nine of 10 patients required pressors aftervolume replacement to maintain MAP of 85 mm Hg. Fifty-two percent of incomplete cervical SCI patients required pres-sors to maintain MAP at 85 mm Hg. Only 9 of 29 patients withthoracic level SCIs required the use of pressors. The authorsreported minimal morbidity with the use of invasive moni-toring or with pharmacological therapy to augment MAP. At1-year follow-up (mean, 17 mo), neurological recovery wasvariable and typically incomplete. Three of 10 ASIA Grade Acervical SCI patients regained ambulatory capacity, and tworegained bladder function. Incomplete cervical SCI patientsfared better. Twenty-three of these patients regained ambula-tory function at 12 months follow-up, only four of whom hadinitial examination scores consistent with walking. Twenty-two (88%) of 25 patients regained bladder control. Thirty-oneof 35 cervical SCI patients and 27 of 29 thoracic level SCIpatients were treated surgically. The authors statistically com-pared selection for and timing of surgery with admissionneurological function and compared surgical treatment, earlyand late, with neurological outcome and found no statisticalcorrelation. They concluded that the enhanced neurologicaloutcome identified in their series after ASCI was optimized byearly and aggressive volume resuscitation and blood pressureaugmentation and was in addition to and/or distinct fromany potential benefit provided by surgery.

SUMMARY

Patients with severe ASCIs, particularly cervical level inju-ries, or patients with multisystem traumatic injury, frequentlyexperience hypotension, hypoxemia, and pulmonary dys-function, and many exhibit cardiovascular instability, despiteearly acceptable cardiac and pulmonary function after initialresuscitation. These occurrences are not limited to ASCI pa-tients with complete autonomic disruption. Life-threateningcardiovascular instability and respiratory insufficiency maybe transient and episodic and may occur in patients who seem



to have stable cardiac and respiratory function early in theirpostinjury course. Patients with the most severe neurologicalinjuries after ASCI seem to have the greatest risk of theselife-threatening events. Monitoring allows the early detectionof hemodynamic instability, cardiac rate disturbances, pulmo-nary dysfunction, and hypoxemia. Identification and treat-ment of these events seems to reduce cardiac- and respiratory-related morbidity and mortality. Management in an ICU orsimilar setting with cardiovascular and pulmonary monitor-ing has an effect on neurological outcome after ASCI. Patientswith ASCIs seem to be best managed in the ICU setting for thefirst 7 to 14 days after injury, the timeframe during which theyseem most susceptible to significant fluctuations in cardiacand pulmonary performance. This seems to be particularlytrue for severe cervical SCI patients, specifically acute ASIAGrades A and B.


The length of stay in the ICU setting necessary to provideoptimal management of patients with ASCI is unknown. Theavailable evidence suggests that most untoward and poten-tially life-threatening cardiac and respiratory events occurwithin the first week or two after injury. Patients with lesssevere ASCIs may require less time in a monitored settingthan patients with more severe injuries. These issues could beaddressed in a prospective cohort study or, potentially, aretrospective case-control study.


REFERENCES

1. Amar AP, Levy ML: Pathogenesis and pharmacological strategiesfor mitigating secondary damage in acute spinal cord injury.Neurosurgery 44:1027–1040, 1999.

2. Bose B, Northrup BE, Osterholm JL, Cotler JM, DiTunno JF:Reanalysis of central cervical cord injury management. Neuro-surgery 15:367–372, 1984.

3. Botel U, Gläser E, Niedeggen A: The surgical treatment of acutespinal paralysed patients. Spinal Cord 35:420–428, 1997.

4. Bracken MB, Collins WF, Freeman DF, Shepard MJ, Wagner FW, SiltenRM, Hellenbrand KG, Ransohoff J, Hunt WE, Perot PL Jr, GrossmanRG, Green BA, Eisenberg HM, Rifkinson N, Goodman JH, Meagher JN,Fischer B, Clifton GL, Flamm ES, Rawe SE: Efficacy of methylpred-nisolone in acute spinal cord injury. JAMA 251:45–52, 1984.

5. Bracken MB, Shepard MJ, Collins WF, Holford TR, Baskin DS,Eisenberg HM, Flamm E, Leo-Summers L, Maroon JC, MarshallLF, Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC Jr,Wilberger JL, Winn HR, Young W: Methylprednisolone or nalox-one treatment after acute spinal cord injury: 1-year follow updata—Results of the Second National Acute Spinal Cord InjuryStudy. J Neurosurg 76:23–31, 1992.

6. Bracken MB, Shepard MJ, Collins WF, Holford TR, Young W,Baskin DS, Eisenberg HM, Flamm E, Leo-Summers L, Maroon J,Marshall LF, Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC,Wilberger JE, Winn HR: A randomized trial of methylpred-nisolone or naloxone in the treatment of acute spinal cord injury:Results of the Second National Acute Spinal Cord Injury Study(NASCIS II). N Engl J Med 322:1405–1411, 1990.

7. Bracken MB, Shepard MJ, Holford TR, Leo-Summers L, AldrichEF, Fazl M, Fehlings M, Herr DL, Hitchon PW, Marshall LF,Nockels RP, Pascale V, Perot PL Jr, Piepmeier JM, Sonntag VKH,Wagner F, Wilberger JE, Winn HR, Young W: Administration ofmethylprednisolone for 24 or 48 hours or tirilazad mesylate for 48hours in the treatment of acute spinal cord injury: Results of theThird National Acute Spinal Cord Injury Randomized ControlledTrial—National Acute Spinal Cord Injury Study. JAMA 277:1597–1604, 1997.

8. Dolan EJ, Tator CH: The effect of blood transfusion, dopamine,and gamma hydroxybutyrate on posttraumatic ischemia of thespinal cord. J Neurosurg 56:350–358, 1982.

9. Geisler FH: GM-1 ganglioside and motor recovery following hu-man spinal cord injury. J Emerg Med 11[Suppl 1]:49–55, 1993.

10. Geisler FH, Dorsey FC, Coleman WP: Recovery of motor functionafter spinal-cord injury: A randomized, placebo-controlled trialwith GM-1 ganglioside. N Engl J Med 324:1829–1838, 1991.

11. Gschaedler R, Dollfus P, Molé JP, Molé L, Loeb JP: Reflections onthe intensive care of acute cervical spinal cord injuries in a generaltraumatology centre. Paraplegia 17:58–61, 1979.

12. Hachen HJ: Idealized care of the acutely injured spinal cord inSwitzerland. J Trauma 17:931–936, 1977.

13. Hall ED, Wolf DL: A pharmacological analysis of the pathophys-iological mechanisms of posttraumatic spinal cord ischemia.J Neurosurg 64:951–961, 1986.

14. Ledsome JR, Sharp JM: Pulmonary function in acute cervical cordinjury. Am Rev Respir Dis 124:41–44, 1981.

15. Lehmann KG, Lane JG, Piepmeier JM, Batsford WP: Cardiovas-cular abnormalities accompanying acute spinal cord injury inhumans: Incidence, time course and severity. J Am Coll Cardiol10:46–52, 1987.

16. Levi L, Wolf A, Belzberg H: Hemodynamic parameters in patientswith acute cervical cord trauma: Description, intervention, andprediction of outcome. Neurosurgery 33:1007–1017, 1993.

17. Levi L, Wolf A, Rigamonti D, Ragheb J, Mirvis S, Robinson WL:Anterior decompression in cervical spine trauma: Does the timingof surgery affect the outcome? Neurosurgery 29:216–222, 1991.

18. Lu K, Lee TC, Liang CL, Chen HJ: Delayed apnea in patients withmid- to lower cervical spinal cord injury. Spine 25:1332–1338,2000.

19. Mansel JK, Norman JR: Respiratory complications and manage-ment of spinal cord injuries. Chest 97:1446–1452, 1990.

20. McMichan JC, Michel L, Westbrook PR: Pulmonary dysfunctionfollowing traumatic quadriplegia: Recognition, prevention, andtreatment. JAMA 243:528–531, 1980.

21. Piepmeier JM, Lehmann KB, Lane JG: Cardiovascular instabilityfollowing acute cervical spinal cord trauma. Cent Nerv SystTrauma 2:153–160, 1985.

22. Reines HD, Harris RC: Pulmonary complications of acute spinalcord injuries. Neurosurgery 21:193–196, 1987.

23. Sandler AN, Tator CH: Effect of acute spinal cord compressioninjury on regional spinal cord blood flow in primates.J Neurosurg 45:660–676, 1976.

24. Sandler AN, Tator CH: Review of the effect of spinal cord traumaon the vessels and blood flow in the spinal cord. J Neurosurg45:638–646, 1976.

25. Tator CH: Vascular effects and blood flow in acute spinal cordinjuries. J Neurosurg Sci 28:115–119, 1984.

26. Tator CH: Ischemia as a secondary neural injury, in Salzman SK,Faden AI (eds): Neurobiology of Central Nervous System Trauma.New York, Oxford University Press, 1994, pp 209–215.



27. Tator CH: Experimental and clinical studies of the pathophysiol-ogy and management of acute spinal cord injury. J Spinal CordMed 19:206–214, 1996.

28. Tator CH: Biology of neurological recovery and functionalrestoration after spinal cord injury. Neurosurgery 42:696–708,1998.

29. Tator CH, Fehlings MG: Review of the secondary injury theory ofacute spinal cord trauma with emphasis on vascular mechanisms.J Neurosurg 75:15–26, 1991.

30. Tator CH, Duncan EG, Edmonds VE, Lapczak LI, Andrews DF:Changes in epidemiology of acute spinal cord injury from 1947 to1981. Surg Neurol 40:207–215, 1993.

31. Tator CH, Rowed DW, Schwartz MI, Gertzbein SD, Bharatwal N,Barkin M, Edmonds VE: Management of acute spinal cord inju-ries. Can J Surg 27:289–293, 296, 1984.

32. Vale FL, Burns J, Jackson AB, Hadley MN: Combined medical andsurgical treatment after acute spinal cord injury: Results of a prospectivepilot study to assess the merits of aggressive medical resuscitation andblood pressure management. J Neurosurg 87:239–246, 1997.

33. Wolf A, Levi L, Mirvis S, Ragheb J, Huhn S, Rigamonti D,Robinson WL: Operative management of bilateral facet disloca-tion. J Neurosurg 75:883–890, 1991.

34. Zach GA, Seiler W, Dollfus P: Treatment results of spinal cord injuriesin the Swiss Paraplegic Centre of Basel. Paraplegia 14:58–65, 1976.

Drawings by Leonardo da Vinci of the human spine. His representation of the spinal column was perhaps the first to showthe correct curvatures, articulations, and number of vertebrae. Courtesy, Dr. Edwin Todd, Pasadena, California.


CHAPTER 8

Blood Pressure Management after Acute Spinal Cord Injury

RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS:• Hypotension (systolic blood pressure <90 mm Hg) should be avoided if possible or corrected as soon as

possible after acute spinal cord injury.• Maintenance of mean arterial blood pressure at 85 to 90 mm Hg for the first 7 days after acute spinal cord

injury to improve spinal cord perfusion is recommended.

RATIONALE

Acute traumatic spinal cord injury is frequently associ-ated with systemic hypotension. Hypotension may beattributable to associated traumatic injuries with hypo-

volemia, direct severe spinal cord trauma itself, or a combi-nation. The occurrence of hypotension has been shown to beassociated with worse outcomes after traumatic injury, in-cluding severe head injury (1, 2, 8, 16, 21, 25). Although aprospective controlled assessment of the effects of hypoten-sion on acute spinal cord injury (ASCI) in humans has notbeen performed, laboratory evidence suggests that hypoten-sion contributes to secondary injury after ASCI by furtherreducing spinal cord blood flow and perfusion (1, 3, 4, 8, 16,18–22, 25). Hypotension in animal models of spinal cord in-jury (SCI) results in worse neurological outcome (13, 14, 23,26, 28, 29). Several clinical series of human patients with ASCImanaged in an aggressive fashion with attention to bloodpressure, oxygenation, and hemodynamic performance reportno deleterious effects of therapy and suggest improved neu-rological outcome (13, 14, 23, 26, 28, 29). Despite these obser-vations, most patients with ASCI treated in contemporarypractice are not routinely monitored or treated with bloodpressure augmentation after injury. For these reasons, theissues of routine blood pressure support and threshold levelsof mean arterial pressure maintenance after ASCI have beenraised.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature from 1966 to 2001 was undertaken.The following medical subject headings were used in combi-nation with “spinal cord injury”: medical management, non-operative management, hypotension, and spinal cord bloodflow. Approximately 3000 citations were acquired. Non-English language citations were deleted. Titles and abstractsof the remaining publications were reviewed, and relevantarticles were selected to develop the guidelines. We focused

on two specific topics concerning human patients with ASCI:hypotension (22 articles reviewed) and spinal cord blood flow(no articles identified). Additional references were culledfrom the reference lists of the remaining papers. Finally, mem-bers of the author group were asked to contribute articlesknown to them on the subject matter that were not found byother search means. Articles describing nonhuman laboratoryinvestigations germane to the topic, related general reviewarticles, and relevant studies of hypotension and human trau-matic brain injury referenced in the Scientific Foundation areincluded among the 29 citations in the references. These ef-forts resulted in six articles describing clinical case series(Class III medical evidence), which form the foundation forthis guideline. They are summarized in Table 8.1.


Ischemia of the spinal cord is thought to be one of the mostimportant contributors to neuronal injury and neurologicaldeficit after ASCI. Both local and systemic vascular alterationscan contribute to ischemia after ASCI by further reducingspinal cord blood flow that can exacerbate and extend theprincipal spinal cord insult (1, 6, 8, 16, 21, 25).

In the normal, noninjured spinal cord, arterial blood supplyis diffuse, primarily delivered via a single anterior spinalartery and two posterior spinal arteries. A variable number ofanterior and posterior radicular arteries provide segmentalcontributions over the length of the cord (24, 25). They feedanastomotic arterial channels over the pial surface that supplythe outer half of the cord and penetrating central arteries fromthe anterior spinal artery, which supply the central portion ofthe cord. Terminal branches of the central arteries extendrostral and caudal to overlap with adjacent terminal arteries,but the terminal arterioles that originate from the terminalarteries do not interconnect within the cord. They in turn giverise to an extensive capillary network, which does intercon-nect within the deep gray and white matter of the cord.Capillaries are much more numerous and extensive in the


gray matter than in the white matter, reflecting the increasedmetabolic needs of cell bodies compared with axons (24, 25).Perfusion of the spinal cord under normal physiological cir-cumstances is maintained over a wide range of systemic bloodpressure by autoregulatory mechanisms that seem identical tothose that regulate cerebral blood flow (1, 3–5, 7, 9, 10, 15, 16,18, 20–22, 25).

Local vascular alterations after ASCI are multiple, and theprecise mechanisms of injury-induced ischemia of the cordhave yet to be elucidated. Most investigators cite direct vas-cular injury at the site of the primary trauma as the earliestcomponent of the ischemic injury process (1, 6, 8, 19–21). Theprincipal SCI not only leads to white and gray matter injury atthe insult site but, because of sulcal vessels and collateralterminal arteries that pass through the primary injury site,creates white matter ischemia distal to the direct injury site (8,21, 22, 25). In addition, the primary SCI creates intraluminalthrombosis and vasospasm and initiates a variety of second-ary injury biochemical phenomena that further reduce bloodflow, injure endothelium, increase edema and microvascularcompression, and contribute to microvascular collapse (1, 8,19, 21, 22, 27). Posttraumatic spinal cord ischemia has beenshown to become progressively worse over the first severalhours after injury in animals (1, 4, 6, 7, 16, 21). Laboratory

models of SCI have convincingly demonstrated that autoreg-ulation of spinal cord blood flow is lost after injury, exacer-bating local spinal cord ischemia and rendering the spinalcord vulnerable to systemic hypotension (1, 3–5, 7, 8, 16, 18,21, 27). This is analogous to what often occurs in regionalcerebrovasculature after acute traumatic brain injury (1, 4, 5,7–9, 15, 16, 19, 21, 25, 27).

Systemic hemodynamic alterations after ASCI have beenwell documented and include hypotension, cardiac dysrhyth-mias, reduced peripheral vascular resistance, and reducedcardiac output (1, 12–14, 17, 21, 26). Patients with the mostsevere injuries, particularly those with severe cervical SCIs,are at greatest risk for cardiac, hemodynamic, and respiratorydisturbances in the first week after ASCI (11, 12, 17). Theseuntoward occurrences, which may be episodic in nature, canresult in hypotension and hypoxia. If, as many investigatorssuspect, ASCI with loss of spinal cord autoregulation is anal-ogous to acute traumatic brain injury, hypotension and hyp-oxia can worsen the severity of the original insult and can bedisastrous for potential neurological recovery (1, 8, 20, 21).Although the relationship between systemic hypotension andoutcome after ASCI has not been directly studied in humanpatients, inference from studies of patients with traumaticbrain injury seems appropriate (2, 8, 21). Prospectively col-

TABLE 8.1. Summary of Reports on Blood Pressure Management after Acute Spinal Cord Injurya


Vale et al., 1997 (26) Prospective assessment of 77 ASCI patientstreated in ICU, aggressive hemodynamicsupport, MAP �85.No control group.

III Improved outcome with aggressivemedical care, distinct from potentialbenefit from surgery at 1-yr follow-up.

Levi et al., 1993 (13) 50 patients treated in ICU, aggressivemedical treatment, MAP �90.

III Improved outcome with aggressivehemodynamic support at 6 wk postinjury.

Levi et al., 1991 (14) 103 ACSI patients, 50 incomplete (GroupA), 53 complete (Group B), ICU care,hemodynamic support, MAP �85.

III Improved neurological outcome, nosignificant difference between early andlate surgery in either group.

Wolf et al., 1991 (28) 52 patients with locked facets reducedwithin 4 h, ICU care, MAP �85. 49operated on, 23 day 1, 26 delayed.

III Closed reduction 61%. 52% 1-yr follow-up. In general, improved neurologicaloutcome with hemodynamic therapy.

Tator et al., 1984 (23) 144 ASCI patients managed per protocolof ICU care, hemodynamic support.Compared with earlier cohort.

III Improved neurological outcome, lessmortality with early transfer and ICU care.

Zach et al., 1976 (29) Prospective assessment of 117 ACSIpatients at Swiss center, ICU setting.Aggressive medical therapy and bloodpressure support (Rheomacrodexb � 7 d;dexamethasone � 10 d).No comparison or control group.

III Improved neurological outcome withaggressive medical treatment and bloodpressure management. Better outcome forearly referrals.

a ASCI, acute spinal cord injury; MAP, mean arterial pressure; ICU, intensive care unit.b Rheomacrodex, dextran 40 (Medisan, Parsippany, NJ).

Blood Pressure Management after Acute Cervical Spinal Cord Injury


lected data from the Traumatic Coma Data Bank (Class IIevidence) demonstrate that hypotension (systolic blood pres-sure �90 mm Hg) or hypoxia (paO2 �60 mm Hg) was inde-pendently associated with significant increases of morbidityand mortality after severe traumatic brain injury (2). A singleepisode of hypotension was associated with a 150% increasein mortality. It is in this very setting that therapeutic inter-vention aimed at correcting hypotension and maintainingthreshold levels of mean arterial pressure (MAP) to improvecerebral or spinal cord perfusion has its greatest potential.Several reports of case series suggest that treatment of hypo-tension and resuscitation to maintain MAP at high-normallevels, 85 to 90 mm Hg, may enhance neurological outcomeafter acute traumatic SCI (13, 14, 23, 26, 28, 29).

Zach et al. (29) used a prospective aggressive medical man-agement paradigm in the treatment of 117 consecutive ASCIpatients. All patients were treated in the intensive care unit(ICU) with central venous pressure monitoring and weretreated with volume expansion (Rheomacrodex 40 [dextran40; Medisan, Parsippany, NJ], 500 ml/d) for maintenance ofsystemic blood pressure for 7 days. Patients were stratified byinjury level, degree of deficit (Frankel grade), and time ofadmission after injury. The authors reported that 62% ofcervical level SCI patients they managed in this way im-proved at last follow-up, including 8 of 18 Frankel Grade Apatients, with 2 patients improving by two grades, and 1patient improving by three grades. No patient with a cervicalinjury worsened; 38% were unchanged from admission. Ofpatients who arrived within 12 hours of injury, 67% improvedcompared with their admission neurological examination. Ofpatients admitted between 12 and 48 hours after injury, only59% improved. When admission occurred 48 hours after in-jury, improvement was seen in only 50% of patients. Theauthors concluded that early transfer and immediate medicalspecific treatment of the spinal injury with attention to main-tenance of acceptable blood pressure seemed to improve neu-rological recovery (29).

Tator et al. (23) in 1984 described their experience with 144patients with ASCI managed between 1974 and 1979 at adedicated SCI unit in Toronto, Ontario, Canada. The authorscompared their results with a cohort of 358 SCI patientsmanaged between 1948 and 1973, before the development ofthe acute care SCI facility. All 144 patients managed from 1974to 1979 were treated in an ICU setting with strict attention tothe treatment of hypotension and respiratory failure. Hypo-tension was “treated vigorously” with crystalloid and trans-fusion of whole blood or plasma for volume expansion. Pa-tients with respiratory dysfunction were treated withventilatory support as indicated. Tator et al. reported thatmean time from injury to admission and treatment was 4.9hours, compared with more than 12 hours from 1948 to 1973.Neurological improvement was observed in 41 (43%) of 95patients managed under the aggressive ICU medical para-digm. Fifty-two patients (55%) demonstrated no improve-ment. Only two patients (2%) deteriorated. The authors re-ported lower mortality, reduced morbidity, shorter length ofstay, and lower cost of treatment with their contemporarycomprehensive management paradigm compared with the

1948 to 1973 experience. They cited improved respiratorymanagement in their ICU as one of the principal factorsresponsible for reduced mortality and credited the avoidanceof hypotension, sepsis, and urological complications for re-duced morbidity after injury. These improved managementresults were realized despite the fact that 28% of the ASCIpatients they treated had additional injuries that increasedtheir risk of morbidity and mortality.

Wolf et al. (28), in 1991, reported their experience with 52patients with acute cervical bilateral facet dislocation injuriesmanaged with an aggressive treatment paradigm that in-cluded ICU care, aggressive resuscitation, invasive monitor-ing, and hemodynamic manipulation to maintain mean bloodpressure above 85 mm Hg for 5 days. Thirty-four patients hadcomplete neurological injuries, 13 patients had incompleteinjuries, and 5 patients were intact. The authors attemptedclosed reduction within 4 hours of patient arrival to theircenter and performed early open reduction on patients whocould not be reduced by closed means. The authors describedneurological improvement at discharge in 21% of completeSCI patients and in 62% of patients with incomplete cervicalSCIs at admission. No intact patient deteriorated. The authorsconcluded that their protocol of aggressive, early medical andsurgical management of patients with ASCI improved out-come after injury. Treatment in the ICU setting, hemodynamicmonitoring with maintenance of MAP above 85 mm Hg, andearly decompression of the spinal cord by open or closedmeans seemed to reduce secondary complications after ASCIin their study.

Levi et al. (13) treated 50 acute cervical SCI patients in theICU setting according to an aggressive management protocolthat included invasive hemodynamic monitoring and volumeand pressor support to maintain a hemodynamic profile withadequate cardiac output and mean blood pressure above 90mm Hg. Their 1993 report described 31 patients with FrankelGrade A injuries at admission, 8 patients with Frankel GradeB injuries, and 11 patients with Frankel Grades C and D. Eightpatients had shock at the time of admission (systolic bloodpressure �90 mm Hg), and 82% of patients had volume-resistant hypotension requiring pressors within the first 7days of treatment. Volume-resistant hypotension was 5.5times more common among patients with complete motorinjuries. Forty percent of patients managed by protocol im-proved, including several with complete injuries; 42% re-mained unchanged; and 18% (9 patients) died. There wasminimal morbidity associated with invasive hemodynamicmonitoring. The authors concluded that hemodynamic mon-itoring in the ICU allows early identification and prompttreatment of cardiac dysfunction and hemodynamic instabil-ity and can reduce the potential morbidity and mortality afterASCI.

Vale et al. (26), in 1997, reported their experience with anonrandomized, prospective pilot study in the assessment ofaggressive medical resuscitation and blood pressure manage-ment in 77 consecutive ASCI patients. All patients were man-aged in the ICU with invasive monitoring (Swan Ganz cath-eters and arterial lines) and blood pressure augmentation tomaintain MAP above 85 mm Hg for 7 days after injury. The



authors reported 10 patients with complete cervical SCIs, 25patients with incomplete cervical injuries, 21 patients withcomplete thoracic SCIs, and 8 patients with incomplete tho-racic level SCIs. The average admission MAP for completecervical SCI patients was 66 mm Hg. Nine of 10 completecervical SCI patients required pressors after volume replace-ment to maintain MAP at 85 mm Hg. Fifty-two percent ofincomplete cervical SCI patients required pressors to maintainMAP at 85 mm Hg. Only 9 of 29 patients with thoracic levelSCIs required the use of pressors. The authors reported min-imal morbidity with the use of invasive monitoring or withpharmacological therapy to augment MAP. At 1-yearfollow-up (mean, 17 mo), 3 of 10 complete cervical SCI pa-tients regained ambulatory capacity and 2 patients regainedbladder function. Incomplete cervical SCI patients fared bet-ter. Twenty-three of these patients regained ambulatory func-tion at 12 months follow-up, only four of whom had initialexamination scores consistent with walking. Twenty-two(88%) of 25 patients regained bladder control. Thirty-one of 35cervical SCI patients and 27 of 29 thoracic level SCI patientswere treated surgically. The authors statistically comparedselection for and timing of surgery with admission neurolog-ical function and compared surgical treatment, early and late,with neurological outcome and found no statistical correla-tion. They concluded that the enhanced neurological outcomeidentified in their series after ASCI was optimized by earlyand aggressive volume resuscitation and blood pressure aug-mentation and was in addition to and/or distinct from anypotential benefit provided by surgery.

The collective experience described in these case series (ClassIII evidence) strongly suggests that maintenance of MAP at 85 to90 mm Hg improves spinal cord perfusion or affects neurologi-cal outcome (13, 14, 23, 26, 28, 29). Prompt treatment of hypo-tension and resuscitation to MAP levels of 85 to 90 mm Hg issafe and suggests that elevation of MAP to threshold levels maybe beneficial to patients with ASCIs. The 7-day duration oftreatment and the threshold levels of MAP maintenance seem tohave been chosen arbitrarily by the individual clinical investiga-tors (13, 26, 28). They are thought to be analogous to initialduration and threshold MAP level recommendations for man-agement of patients after acute traumatic brain injury. None ofthe authors provides a specific recipe or an algorithm to guideblood pressure augmentation. All of the articles describe acutelyinjured patients who have arterial lines and central venous orSwan Ganz catheters in place to monitor pressures and volumestatus (13, 14, 23, 26, 28, 29). Initially, crystalloid is given intra-venously in response to MAP levels below 85 mm Hg. Colloid isadministered if the hematocrit is low (blood) or as a volumeexpander (albumin). If the patient’s volume status is optimal butthe MAP remains below threshold, the authors describe the useof pressors, typically (although not exclusively) a �-agonist (do-pamine), before the addition of an �-agonist (Neo-Synephrine[Sanofi Winthrop Pharmaceuticals, New York, NY]), to elevatethe MAP. These agents are titrated to the appropriate dose levelto achieve the threshold MAP using volume, pressure, and car-diac performance data provided by the invasive monitoringdevices.

SUMMARY

Hypotension is common after acute traumatic SCI in hu-mans. Hypotension contributes to spinal cord ischemia afterinjury in animal models and can worsen the initial insult andreduce the potential for neurological recovery. Although un-proven by Class I medical evidence studies, it is likely thatthis occurs in human SCI patients as well. Because the correc-tion of hypotension and maintenance of homeostasis is a basicprinciple of ethical medical practice in the treatment of pa-tients with traumatic neurological injuries, depriving ASCIpatients of this treatment would be untenable. For this reason,Class I evidence about the effects of hypotension on outcomeafter acute human SCI will never be obtained. However,correction of hypotension has been shown to reduce morbid-ity and mortality after acute human traumatic brain injuryand is a guideline level recommendation for the managementof traumatic brain injury. Although a similar treatment guide-line cannot be supported by the existing SCI literature, cor-rection of hypotension in the setting of acute human SCI isoffered as a strong treatment option. Class III evidence fromthe literature suggests that maintenance of MAP at 85 to 90mm Hg after ASCI for 7 days is safe and may improve spinalcord perfusion and, ultimately, neurological outcome.

KEY ISSUES FOR FUTURE RESEARCH

The issue of whether blood pressure augmentation affectsoutcome after human SCI is important and deserves furtherstudy. If augmentation of MAP is determined to be of poten-tial benefit, the threshold levels of MAP most appropriate andthe length of augmentation therapy need definition. Theseissues are best analyzed in a multi-institutional prospectivecohort study or a properly designed multi-institutional retro-spective case-control study.


REFERENCES


2. Chesnut RM, Marshall LF, Klauber MR, Blunt BA, Baldwin N,Eisenberg HM, Jane JA, Marmarou A, Foulkes MA: The role ofsecondary brain injury in determining outcome from severe headinjury. J Trauma 34:216–222, 1993.


4. Ducker TB, Kindt GW, Kempf LG: Pathological findings in acuteexperimental spinal cord trauma. J Neurosurg 35:700–708, 1971.

5. Flohr H, Pöll W, Brock M: Regulation of spinal cord blood flow,in Russell RWR (ed): Brain and Blood Flow: Proceedings of the 4thInternational Symposium on the Regulation of Cerebral Blood Flow.London, Pitman Medical, 1971, pp 406–409.

6. Hall ED, Wolf DL: A pharmacological analysis of the pathophys-iological mechanisms of posttraumatic spinal cord ischemia.J Neurosurg 64:951–961, 1986.

Blood Pressure Management after Acute Cervical Spinal Cord Injury S61


7. Kindt GW, Ducker TB, Huddlestone J: Regulation of spinal cordblood flow, in Russell RWR (ed): Brain and Blood Flow: Proceedingsof the 4th International Symposium on the Regulation of Cerebral BloodFlow. London, Pitman Medical, 1971, pp 401–405.

8. King BS, Gupta R, Narayan RK: The early assessment and inten-sive care unit management of patients with severe traumatic brainand spinal cord injuries. Surg Clin North Am 80:855–870, 2000.

9. Kobrine AI, Doyle TF, Martins AN: Autoregulation of spinal cordblood flow. Clin Neurosurg 22:573–581, 1975.

10. Kobrine AI, Doyle TF, Rizzoli HV: Spinal cord blood flow asaffected by changes in systemic arterial blood pressure.J Neurosurg 44:12–15, 1976.

11. Ledsome JR, Sharp JM: Pulmonary function in acute cervical cordinjury. Am Rev Respir Dis 124:41–44, 1981.

12. Lehmann KG, Lane JG, Piepmeier JM, Batsford WP: Cardiovas-cular abnormalities accompanying acute spinal cord injury inhumans: Incidence, time course and severity. J Am Coll Cardiol10:46–52, 1987.

13. Levi L, Wolf A, Belzberg H: Hemodynamic parameters in patientswith acute cervical cord trauma: Description, intervention, andprediction of outcome. Neurosurgery 33:1007–1017, 1993.

14. Levi L, Wolf A, Rigamonti D, Ragheb J, Mirvis S, Robinson WL:Anterior decompression in cervical spine trauma: Does the timingof surgery affect the outcome? Neurosurgery 29:216–222, 1991.

15. Lewelt W, Jenkins LW, Miller JD: Autoregulation of cerebralblood flow after experimental fluid percussion injury of the brain.J Neurosurg 53:500–511, 1980.

16. Osterholm JL: The pathophysiological response to spinal cordinjury: The current status of related research. J Neurosurg 40:5–33, 1974.

17. Piepmeier JM, Lehmann KB, Lane JG: Cardiovascular instabilityfollowing acute cervical spinal cord trauma. Cent Nerv SystTrauma 2:153–160, 1985.

18. Senter HJ, Venes JL: Loss of autoregulation and posttraumaticischemia following experimental spinal cord trauma. J Neurosurg50:198–206, 1979.


20. Tator CH: Hemodynamic issues and vascular factors in acuteexperimental spinal cord injury. J Neurotrauma 9:139–141, 1992.

21. Tator CH: Ischemia as a secondary neural injury, in Salzman SK,Faden AI (eds): Neurobiology of Central Nervous System Trauma.New York, Oxford University Press, 1994, pp 209–215.


23. Tator CH, Rowed DW, Schwartz MI, Gertzbein SD, Bharatwal N,Barkin M, Edmonds VE: Management of acute spinal cord inju-ries. Can J Surg 27:289–293, 296, 1984.

24. Turnbull IM: Microvasculature of the human spinal cord.J Neurosurg 35:141–147, 1971.

25. Turnbull IM: Blood supply of the spinal cord: Normal and patho-logical considerations. Clin Neurosurg 20:56–84, 1973.

26. Vale FL, Burns J, Jackson AB, Hadley MN: Combined medical andsurgical treatment after acute spinal cord injury: Results of aprospective pilot study to assess the merits of aggressive medicalresuscitation and blood pressure management. J Neurosurg 87:239–246, 1997.

27. Wallace MC, Tator CH: Successful improvement of blood pres-sure, cardiac output, and spinal cord blood flow after experimen-tal spinal cord injury. Neurosurgery 20:710–715, 1987.


29. Zach GA, Seiler W, Dollfus P: Treatment results of spinal cordinjuries in the Swiss Paraplegic Centre of Basel. Paraplegia 14:58–65, 1976.

Images of the human spine. From, Tilney F, Alsop Riley H: The Form and Functions of the Central Nervous System: An Intro-duction to the Study of Nervous Diseases. New York, Harper & Brothers, 1938, 3rd ed.


CHAPTER 9

Pharmacological Therapy after Acute Cervical SpinalCord Injury

RECOMMENDATIONSCORTICOSTEROIDS:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options: Treatment with methylprednisolone for either 24 or 48 hours is recommended as an option in the

treatment of patients with acute spinal cord injuries that should be undertaken only with the knowledgethat the evidence suggesting harmful side effects is more consistent than any suggestion of clinical benefit.

GM-1 GANGLIOSIDE:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options: Treatment of patients with acute spinal cord injuries with GM-1 ganglioside is recommended as an

option without demonstrated clinical benefit.

RATIONALE

The hope that administration of a pharmacological agentdelivered shortly after acute spinal cord injury (ASCI)might improve neurological function and/or assist neu-

rological recovery has long been held. A variety of promisingsubstances have been tested in animal models of ASCI, butfew have had potential application to human spinal cordinjury (SCI) patients. Four pharmacological substances havemet rigorous criteria in laboratory testing and initial humaninvestigations: two corticosteroids (methylprednisolone andtirilazad mesylate), naloxone, and GM-1 ganglioside. All fourpharmacological agents have been evaluated in controlled,randomized, blinded clinical trials of human patients withASCIs. Two of these substances, tirilazad and naloxone, havebeen studied less extensively and as yet have unclear efficacyin the management of acute human SCI. The purpose of thismedical evidence-based review is to define the usefulness ofadministration of methylprednisolone with or without GM-1ganglioside in the contemporary management of ASCI patients.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of literature published from 1966 to 2001 was un-dertaken. The following medical subject headings were usedin combination with “spinal cord injury” and “neurologicaldeficit”: steroids, methylprednisolone, and GM-1 ganglioside.

Approximately 2400 citations were acquired. Non-Englishlanguage citations and nonhuman experimental studies weredeleted. Titles and abstracts of 652 manuscripts were re-viewed, 639 on the topic of corticosteroids and human SCIand 13 on the topic of GM-1 ganglioside and human SCI. Ad-ditional references were culled from the reference lists of theremaining papers. Finally, the members of the author groupwere asked to contribute articles known to them on the subjectmatter that were not found by other search means. Duplica-tions, case reports, pharmacokinetic reports, general reviews,and articles with mention of one agent or another but withoutscientific assessment were eliminated. Several editorials, cri-tiques, and responses to published reports and studies wereincluded. Forty-six published references on the topic of meth-ylprednisolone in the treatment of patients with ASCI andseven published references for GM-1 ganglioside provide thebasis for this guideline. Thirteen studies on methylpred-nisolone and two studies on GM-1 ganglioside are summa-rized in Tables 9.1 and 9.2.


Methylprednisolone

Corticosteroids, particularly methylprednisolone, havebeen studied extensively in animal models of SCI (2, 19, 47, 48,50, 51). Although their precise mechanisms of action are not

This chapter remains the most controversial of the Guidelines. The readers are advised to carefully review the availabledata and Comments provided within this Supplement to establish their own perspective on this evolving matter.

Michael L.J. Apuzzo


TABLE 9.1. Summary of Reports on Treatment with Methyprednisolone after Acute Cervical Spinal Cord Injurya


Bracken et al., 1984

(11)

Multicenter, double-blind randomized trial

comparing MP (1000 mg/d versus 100 mg/d for

11 d) in treatment of 330 ASCI patients (NASCIS

I study).

III

(Study design, data presentation,

interpretation, and analysis

flaws)

No treatment effect at 6 wk and 6 mo postinjury. No control group.


(15)

1-yr follow-up of NASCIS I study. III



flaws)

No significant difference in neurological recovery of motor or sensory

function 1-yr postinjury.


(14)

Multicenter, randomized, double-blind, placebo-

controlled trial comparing MP with naloxone

and placebo in treatment of 487 ASCI patients

(NASCIS II study).

III



flaws)

Significant improvement in motor change scores (P � 0.03), and

sensation change scores (P � 0.02) at 6 mo postinjury for patients treated

with MP within 8 h of injury.


(13)

1-yr follow-up of NASCIS II study. III



flaws)

Significant improvement in motor change scores 1 year postinjury for

patients treated with MP within 8 h of injury (P � 0.03). Administration of

MP detrimental if given more than 8 h after injury.

Galandiuk et al.,

1993 (21)

Prospective assessment of 15 patients from 1990

to 1993 with retrospective review of 17 patients

from 1987 to 1990 to assess differences in

treatment outcome with MP compared with

treatment without corticosteroids.

III No difference in neurological outcome between two sets of patients. MP

patients had immune response alterations, higher rate of pneumonia, and

longer hospital stays than patients who did not receive corticosteroids.

Gerhart et al., 1995

(29)

Concurrent cohort comparison study

(population-based) of 363 ASCI patients

managed from 1990 to 1991 and 1993. 188

patients managed with NASCIS II MP compared

with 90 patients with no MP.

III

(Inadequate statistical power)

No differences in neurological outcome using Frankel classification

between MP and No-MP patients. However, may be insufficient numbers

of patients to show significant differences.

George et al., 1995

(28)

Retrospective review of 145 ASCI patients, 80

treated with MP compared with 65 who did not

receive MP.

III No difference in mortality or neurological outcome between groups

despite younger age, less severe injury in MP-treated patients.

Gerndt et al., 1997

(30)

Retrospective review with historical control of

231 ASCI patients. 91 excluded. Comparison of

medical complications among 93 MP patients

compared with 47 who received no

corticosteroid.

III MP-treated patients had significant increases in pneumonia (P � 0.02),

acute pneumonia (P � 0.03), ventilated days (P � 0.04), and ICU stay

(P � 0.45), but no adverse effect on long-term outcome.

Poynton et al., 1997

(39)

Case-control analysis of 71 consecutive ASCI

admissions. 63 available for 13 mo to 57 mo

follow-up. 38 patients treated with MP

compared with 25 referred �8 h after injury

who received no MP.

III Multiple factors influence recovery after SCI. No effect of MP or surgery

on outcome.


(16)

Multicenter, randomized, double-blind trial

comparing MP administered for 24 hr to MP

administered 48 hr and TM in the treatment of

499 ASCI patients (NASCIS III study).

III



flaws)

48 MP patients had improved motor recovery at 6 wk and at 6 mo

compared with 24 MP and 48 TM groups NS. When treatment initiated

between 3 h and 8 h after injury, 48 MP had significant improvement of

motor scores at 6 wk (P � 0.04) and 6 mo (P � 0.01). 48 MP was

associated with high rates of sepsis and pneumonia. No control group.


(17)

1-yr follow-up of NASCIS III study. III



flaws)

Recovery rates equal in all 3 groups when treatment initiated within 3 h

of injury. When treatment initiated between 3 h and 8 h, 24 MP patients

had diminished recovery, 48 MP patients had increased motor recovery

(P � 0.053).

Pointillart et al., 2000

(38)

Multicenter, prospective, randomized clinical

trial of 106 ASCI patients treated with MP,

nimodipine, neither, or both.

III

(Inadequate statistical power)

No significant difference in neurological outcome at 1-yr follow-up

between groups. Incomplete ASCI had significant improvement below

level of injury compared to complete patients (P � 0.0001). Higher

incidence of infectious complications among patients receiving

corticosteroids (NS).

Matsumoto et al.,

2001 (36)

Prospective, randomized, double-blind study

comparing incidence of medical complications

among 46 ASCI patients, 23 treated with MP, 23

with placebo.

I MP patients had higher incidence of complications (56.5% versus 34.8%).

Respiratory complications (P � 0.009) and gastrointestinal bleed

(P � 0.036) were most significant between groups. No data on

neurological improvement.

a ASCI, acute spinal cord injury; NASCIS, National Acute Spinal Cord Injury Study; MP, methylprednisolone; ICU, intensive care unit; SCI,spinal cord injury; TM, tirilazad mesylate; NS, not significant.



completely known, they have the potential to stabilize mem-brane structures, maintain the blood-spinal cord barrier po-tentially reducing vasogenic edema, enhance spinal cordblood flow, alter electrolyte concentrations at the site of in-jury, inhibit endorphin release, scavenge damaging free rad-icals, and limit the inflammatory response after injury (2, 47,48, 50, 51). After considerable positive study in the laboratory,methylprednisolone was studied in human SCI patients in amulticenter, randomized, double-blinded clinical trial initi-ated in 1979. The first National Acute Spinal Cord InjuryStudy (NASCIS I) (11), reported in 1984, compared the effi-cacy of administration of a 100-mg bolus of methylpred-nisolone and then 100 mg daily thereafter for 10 days withadministration of a 1000-mg bolus and then 1000 mg daily for10 days in 330 acute injury patients assessed 6 weeks and 6months after injury. There was no control group. The studyrevealed no difference in neurological recovery (motor orsensory function) between the treatment groups at either 6weeks or 6 months after injury. Motor scores were determinedfrom the examination of seven muscle groups on each side ofthe body scored on a 6-point scale. Sensory function wasassessed using a 3-point scale of dermatomal light touch andpinprick sensation. The authors reported the motor and sen-sory scores from the right side of the body only. There was noanatomic level injury limit (superior to T12 vertebral level, forexample) in the study to include only SCI patients and ex-clude primary cauda equina injuries or “mixed” central andcauda equina injuries that might occur with a lower fractureinjury (e.g., T12–L1 or L1–L2 injuries). The study did notrequire a minimum motor impairment for inclusion; hence,patients with normal motor examinations and those withminimal neurological deficits were included in the study if theattending physician determined that the patient had an SCI ofany severity. In 1985, the same group of investigators reportedon the 1-year follow-up of these study patients (15). No dif-ferences in motor or sensory outcome were identified betweenthe two treatment groups.

Animal studies of the efficacy of methylprednisolone afterexperimental SCI suggested that the doses of methylpred-nisolone used in the NASCIS I investigation were too low to

demonstrate a significant difference in outcome (2, 14, 19, 50,51). A multicenter NASCIS II trial was initiated in 1985 usinga much higher dose of methylprednisolone (30 mg/kg as abolus and then 5.4 mg/kg/h infusion for 23 h). These patientswere compared with similarly injured patients who receivedeither naloxone (5.4 mg/kg bolus and then an infusion of 4.0mg/kg/h for 23 h) or placebo. Patients had to be randomizedto one of three treatment arms within 12 hours of ASCI. Theresults of NASCIS II were reported in 1990 (14). Four hundredeighty-seven patients were entered into the study; 162 re-ceived methylprednisolone, 154 were given naloxone, and 171patients were in the placebo control group. The authors re-ported that the administration of methylprednisolone within8 hours of injury was associated with a significant improve-ment in motor function (neurological change scores, right sideof body only, P � 0.03), and in sensation (pinprick, P � 0.02;light touch, P � 0.03) at the 6-month follow-up, comparedwith patients receiving methylprednisolone more than 8hours after injury and patients receiving naloxone or placebo.No similar significant improvements were noted at the6-week follow-up, either motor or sensory. Motor scores weredetermined from the examination of seven muscle groups oneach side of the body scored on a scale of 0 to 5 points.Sensory function was assessed using a 3-point scale of der-matomal light touch and pinprick sensation. The NASCIS IIstudy reported on the motor scores from the right side of thebody only. Bilateral sensory scores were provided. Like theNASCIS I study, there was no anatomic level injury limit inthe study (superior to T12 vertebral level, for example), toensure that only SCI patients were included for study (11, 15).Similarly, NASCIS II did not require a minimum motor im-pairment for inclusion; hence, patients with normal motorexaminations and those with minimal neurological deficitswere included. No outcome measures involving patient func-tion were used in this study. In 1992, NASCIS investigatorsreported on the 1-year follow-up of NASCIS II study patients(13). They reported statistically significant improvement inmotor scores on the right side of the body for 62 of 487 studypatients (P � 0.03). These 62 patients received methylpred-nisolone within 8 hours of injury. Significant right body motor

TABLE 9.2. Summary of Reports on Treatment with GM-1 Ganglioside after Acute Spinal Cord Injurya


Geisler et al., 1991 (25) Prospective, randomized, double-blindtrial of GM-1 ganglioside in 37 humanASCI patients. All received 250-mg MPbolus followed by 125 mg every 6 h �72 h before randomization (placebogroup).

I GM-1 ganglioside enhances recovery ofneurological function, significantdifference in recovery compared withMP group (P � 0.047). Insufficientnumbers of patients to draw meaningfulconclusions. No true placebo group.

Geisler et al., 2001 (23) Prospective, randomized, double-blind,stratified multicenter trial of GM-1ganglioside in 760 ASCI patients. Allreceived MP per NASCIS II protocol(placebo group).

I No significant differences inneurological recovery identifiedbetween GM-1-treated patients andMP-treated patients at 26-wk follow-up.Trend for earlier recovery in GM-1-treated patients. No true placebo group.

a ASCI, acute spinal cord injury; NASCIS, National Acute Spinal Cord Injury Study; MP, methylprednisolone.

Pharmacological Therapy in Acute Cervical Spinal Cord Injury S65


score improvement was identified in two of three categoriesof patients, plegic patients with total sensory loss (P � 0.019)and paretic patients with variable sensory loss (P � 0.024), butnot among plegic patients with partial sensory loss (P �0.481). There were no significant improvements in motorchange scores described among the remaining 421 patientsentered in the study. There were no significant differences insensory scores for any treatment group or categories of pa-tients despite the differences reported at the 6-monthfollow-up for patients receiving methylprednisolone within 8hours of injury. Patients treated more than 8 hours after injurywith methylprednisolone or naloxone experienced less recov-ery of motor function compared with placebo treatment pa-tients. The authors concluded that treatment with the studydose of methylprednisolone administered within 8 hours ofinjury improves neurological outcome and is therefore indi-cated in the treatment of patients with ASCI. The use of studydose methylprednisolone in patients was not associated withharmful side effects compared with patients in the othertreatment groups, although the authors reported an increasedincidence of wound infection and gastrointestinal bleedingamong corticosteroid-treated patients. Treatment with meth-ylprednisolone beyond 8 hours after injury was notrecommended.

There are several flaws in the NASCIS II study, and criti-cism has been offered on several methodological, scientific,and statistical issues (18, 19, 22, 31, 32, 35, 37, 40–42, 44–46, 51).The investigators described two a priori hypotheses: thattreatment effect would be influenced by how soon the drugwas given after injury and by the severity of injury. Patientswere considered eligible for inclusion if they were admitted tothe study and randomized to treatment within 12 hours ofinjury. At some point, patient outcome was stratified accord-ing to the timing of methylprednisolone administration (�8 h,�8 h). Some reviewers have requested examination of the rawdata to look for time-related diminishing effects of methyl-prednisolone administration relative to injury rather than as-signment of an “all or nothing” time cutoff (18, 32, 37, 40, 42,51). Analysis of results of the entire population of patientsaccording to the second a priori hypothesis was not providedby the authors (18, 31, 37, 40, 42, 51). Analysis using thesecond hypothesis was accomplished on the group of patientspreviously stratified according to the first hypothesis. It maybe that the two hypotheses are fully independent, yet nojustification for this assumption was offered (31, 40). Thestudy did not offer a standardized medical treatment regimenfor all ASCI patients in this study. The medical managementof study patients including monitoring, blood pressure aug-mentation, respiratory care, deep venous thrombosis prophy-laxis, nutritional support, and initiation of rehabilitation ac-tivities was neither consistent within centers nor consistentfrom center to center (18, 22, 31, 37). Similarly, surgical treat-ment offered to patients in the NASCIS II study was notconsistent from center to center (19, 31, 35, 51). There was nodescription of surgical approaches used for specific pathologyor documentation of the timing of surgical intervention forindividual patients. There was no consideration given to theindependent effect that either aggressive medical manage-

ment or surgery had, or may have had, on outcome (18, 19, 22,31, 35, 37, 51).

The most important and significant criticism of the NASCISII study is the failure to measure patient functional recovery(e.g., functional independence measure [FIM]) to determinewhether the modest improvement reported in neurological ex-amination (change in motor scores) in the methylprednisolone-treated patients had meaningful clinical significance (18, 32, 35,37, 44). It is unclear from the change in score data providedwhether the improvement had any clinical significance to theinjured patients (1, 18, 32, 35, 37, 44–46). One of the most frequentcriticisms of the reported NASCIS II results is the failure toprovide scientific data on which statistical comparisons weremade (18, 19, 31, 32, 37, 40–42, 46, 51). As with the NASCIS Istudy, only right-sided motor scores were reported in NASCISII, but bilateral sensory scores were reported. Change in motorscore (improvement) on the right side only of ASCI patients hasbeen cited by the study authors as a significant neurologicalbenefit associated with methylprednisolone administrationgiven at study doses within 8 hours of injury and assessed at6-month and 1-year follow-up (P � 0.03) (13, 14). These findingswere observed in only a small subset of study patients (18, 31, 37,41). Was this an a priori hypothesis of the investigators, and wasthe result significant for the whole population of patients? If so,then the finding stands and the post hoc subgroup analysissuggests which subgroup receives the benefit. If, however, theentire result is from a post hoc hypothesis and analysis and issignificant only for the subgroup and not for all of the patientsanalyzed together, then it is a weak suggestive finding. This isnot made clear by the authors. Reviewers have argued againstthe use of right-side only motor scores, and particularly thechange of score results in NASCIS II publications (18, 22, 31, 32,40, 41). The lack of evidence describing left-sided motor scoresand total body motor scores in NASCIS II is confusing (4, 8–10,12, 50).

Also confusing is the reported difference in change of motorscore outcome for patients with incomplete SCI who were inthe placebo treatment arm. Patients with incomplete SCIs inthe NASCIS II study who received placebo more than 8 hoursafter injury had significantly better neurological recovery thandid patients who received placebo within 8 hours of injury(13, 18, 32, 42). Additionally, the neurological recovery curvegenerated for patients with incomplete SCIs treated withmethylprednisolone within 8 hours of injury is virtually iden-tical to that of patients with incomplete SCIs treated withplacebo beyond 8 hours after injury. The benefit of treatmentwith respect to neurological recovery (motor change score)with methylprednisolone given within 8 hours of injuryseems equal to treatment with placebo more than 8 hours afterinjury (18, 37, 42).

Statistical criticisms of the NASCIS II results are many (18,19, 22, 31, 32, 40–42, 45, 46, 51). They include potential inter-pretive errors, problematic statistical comparisons, simplifica-tion of subgroup analysis from the pre-planned 15 categoriesto 3 seemingly arbitrarily determined categories, an improperand incomplete presentation of odds ratios, and a post hocanalysis of study data including only 127 patients (62 meth-ylprednisolone, 65 placebo) treated within 8 hours of injury,



rather than the entire study population of 487 patients (18, 19,22, 31, 32, 40–42, 45, 46, 51). NASCIS II was designed andimplemented to be a randomized, controlled, double-blindedclinical study in an attempt to generate Class I evidence on theefficacy of methylprednisolone and naloxone after ASCI inhuman subjects. The lack of a measure of functional signifi-cance, the dependence on post hoc analyses, and the absenceof an analysis of surgical treatment diminish the quality andusefulness of the evidence provided by these studies.

In 1993, Galandiuk et al. (21) described 32 patients withcervical or upper thoracic ASCIs managed in an urban traumacenter. Fourteen patients who received NASCIS II doses ofmethylprednisolone within 8 hours of injury were comparedwith 18 ASCI patients with similar injuries managed withoutcorticosteroids. The authors reported no difference in neurolog-ical outcome between the two sets of patients but noted thatmethylprednisolone-treated patients had immune response al-terations (lower percentage and density of monocyte Class IIantigen expression and lower T-cell helper/suppressor cell ra-tios), a higher rate of pneumonia (79% versus 50%), and longerhospital stays (44.4 d versus 27.7 d) than similar ASCI patientsthey managed without administration of corticosteroids. Al-though the conclusions drawn by the authors are interesting,they have little scientific power. The mix of historical patientswith contemporary patients, the lack of a prospective design,and the haphazard assignment and assessment of patients dilutethe quality of the evidence provided.

Bracken and Holford (8) described the effect of timing ofmethylprednisolone on neurological recovery in NASCIS IIstudy patients in 1993. They concluded from post hoc analysisof the NASCIS II data that methylprednisolone administeredto patients within 8 hours of ASCI improves neurologicalfunction below the level of the spinal cord lesion in patientsinitially diagnosed as having complete or incomplete injuries.The majority of the improvements they reported were amongpatients with incomplete SCIs at admission. Complete injurypatients demonstrated very little recovery below the level ofinjury irrespective of treatment. Their post hoc analysis alsoconfirmed that methylprednisolone administered more than 8hours after injury may be associated with a worse neurolog-ical outcome.

This 1993 article (8) refers to and references the 1-yearfollow-up NASCIS II study data but only describes patientgroups and offers percentages (18, 42). It provides neither newevidence nor the numbers of patients on whom Bracken andHolford based their conclusions. Although the result that theauthors describe is positive (methylprednisolone adminis-tered within 8 h of injury improves spinal cord function inpatients with SCI), it was identified in a very small subgroupof patients, which raises questions as to its true weight andvalidity. The manner in which the data and conclusions werepresented is ambiguous and suggests that this was a positiveresult reflected by analysis of the entire NASCIS II studypopulation (n � 487) (18). In fact, it was only a subgroupanalysis of the population of patients who received methyl-prednisolone within 8 hours of injury (n � 62), compared withthose who received placebo within 8 hours of injury (n � 65).Forty-five methylprednisolone-treated patients had complete

injuries and demonstrated very little change in function belowthe level of injury. The same is true for 43 similar (complete)patients who received placebo (no significant difference). Theactual differences described by the authors are based on 17methylprednisolone patients compared with 22 placebo-treated patients, all of whom had incomplete SCI and hadtherapy initiated within 8 hours of injury (18).

Their report (8) does help to clarify the issue of recovery offunction (motor score change) in NASCIS II patients withcomplete injuries at admission who received methylpred-nisolone within 8 hours of injury. The NASCIS II results at 1year cite a significant improvement in motor function forpatients who received methylprednisolone at study doseswithin 8 hours of injury compared with placebo-treated pa-tients (P � 0.03) (13). For the patients who had completeinjuries who met the early treatment criteria (n � 45), thesignificance of improvement (change in motor score) was P �0.019, compared with similar patients who received placebo.Bracken and Holford’s (8) post hoc analysis revealed no sig-nificant difference in recovery below the level of the lesion inthese patients compared with placebo-treated patients. Thissuggests that the primary improvements in function identi-fied in the NASCIS II study for patients with complete spinalinjuries treated within 8 hours were at the level of injury,likely root recovery, rather than a significant gain in spinalcord function (18). Again, the relationship between any suchrecovery and an improvement in patient function is un-known, irrespective of the sample size, because the study didnot use functional outcome assessments (18, 35, 37).

In 1994, Duh et al. (20) reported on the effect of surgery onoutcome among NASCIS II study patients. In all, 298 of 487study patients underwent 303 operative procedures, 56 viathe anterior approach and 247 via the posterior approach. Theauthors examined the influence of surgery on neurologicaloutcome across all study groups of patients at time periods ofless than 25 hours, 26 to 50 hours, 51 to 100 hours, 101 to 200hours, and more than 200 hours. They found that the mostseverely injured patients were less likely to be treated surgi-cally. The authors did not identify significant differences inoutcome, motor or sensory, with surgical treatment, eitherearly or late. Functional recovery was not measured.

Gerhart et al. (29), in 1995, reported a population-based,concurrent cohort comparison study of 363 ASCI survivorstreated in Colorado. Two hundred eighteen patients weremanaged between May 1990 and December 1991, and 145injury patients were managed 2 years later in 1993. Of 218patients managed in 1990 to 1991, 100 (46%) were treatedaccording to the NASCIS II protocol. Fifty-one patients (23%)received no methylprednisolone, and 67 patients (31%) re-ceived another corticosteroid, were given an incorrect dose, orhad insufficient data. In the 1993 study population, 61% ofASCI patients (n � 88) received methylprednisolone accord-ing to NASCIS II protocol. Thirty-nine patients (27%) receivedno methylprednisolone and 18 patients (13%) were givenanother corticosteroid, received an incorrect dose, or hadinsufficient data. The authors reported no significant differ-ences in outcome as assessed by the Frankel scale at the timeof hospital discharge when 188 patients who received proto-



col methylprednisolone (appropriate dose and timing) werecompared with those (n � 90) who did not receive any meth-ylprednisolone during treatment. This was true for the com-bined population of patients and for both the 1990 to 1991 andthe 1993 patient populations. It does not seem, however, thatadequate numbers of patients were analyzed by the authors,substantially diluting the statistical power of their findings.

In 1995, George et al. (28) reported their experience withASCI patients at Michigan State University from 1989 through1992. One hundred forty-five patients were described, 80 ofwhom were treated with methylprednisolone per the NASCISII protocol (MP group) and 65 of whom did not receivemethylprednisolone (No-MP group). Admission, discharge,and follow-up neurological assessments were accomplishedaccording to the FIM instrument. Fifteen patients were ex-cluded from review, leaving 130 patients (75 MP, 55 No-MP).The MP group was significantly younger than the No-MPgroup (30 yr versus 38 yr, P � 0.05). Although the meantrauma scores were similar between the two groups, the MPpatients had a significantly lower injury severity score (ISS)than the No-MP patients (P � 0.05). The authors found nodifferences in mortality or neurological outcome between pa-tients treated with methylprednisolone and those who werenot. Despite older age and higher injury severity score, theNo-MP group had better mobility at the time of hospitaldischarge. Admission mobility scores were similar (MP � 5.99versus No-MP � 5.90), but the mobility scores differed sig-nificantly on hospital discharge (MP � 5.16 versus No-MP �4.67, P � 0.05). The authors argued that the MP patient grouphad a more favorable opportunity for improvement than theNo-MP patient group owing to younger age and lower ISSscores; however, neurological improvements in the MP groupcompared with the No-MP group were not observed. It isunclear from the study why most patients did not receivecorticosteroid therapy, and this is the weakness of a nonran-domized study in which patient assignment to treatment mayintroduce bias. For example, an examination of the data indi-cates that the worst neurologically injured patients at admis-sion were more likely to have received methylprednisolone.The findings of no difference in neurological examinationimprovement or functional recovery in this group seem torefute the finding of neurological improvement in NASCIS IIpatients who received methylprednisolone less than 8 hoursafter injury compared with those who did not receive thedrug.

Gerndt et al. (30), in 1997, reported a retrospective review of231 patients with ASCI for the purpose of examining medicalcomplications. Ninety-one patients were excluded becausethey received corticosteroids outside the NASCIS II protocol.One hundred forty patients were reviewed, comparing 93patients who received methylprednisolone per the NASCIS IIprotocol with a historical control group of 47 patients whoreceived no corticosteroid during treatment. The patientgroups were similar with respect to age and injury severity.The authors found significant differences (increases) in theincidence of pneumonia (P � 0.02, 2.6-fold increase), particu-larly acute pneumonia (P � 0.03, 4-fold increase), ventilateddays (P � 0.04), and ICU length of stay (P � 0.045) in

methylprednisolone-treated patients compared with thosewho did not receive corticosteroids during treatment. Non-corticosteroid-treated patients had a higher incidence of uri-nary tract infections (P � 0.01). Methylprednisolone-treatedpatients had decreased general care floor length of stay (P �0.02) and rehabilitation length of stay (P � 0.035). The authorsconcluded that methylprednisolone may increase the inci-dence of early infection, particularly pneumonia, in ASCIpatients but has no adverse effect on long-term outcome. In1997, Poynton et al. (39) described 71 consecutive ASCI pa-tients managed at the National Spinal Trauma Unit in Dublin,Ireland. They attempted a case-control analysis of ASCI pa-tients treated with methylprednisolone (n � 38) comparedwith patients who did not receive methylprednisolone (n �25) and provided follow-up from 13 months to 57 monthsafter injury. Patients who did not receive methylprednisolonewere referred more than 8 hours after injury. The authorsconcluded that multiple factors influenced outcome afterASCI. They found no difference in neurological outcomewhen they compared patients who received methylpred-nisolone with those who did not.

The results of the third NASCIS study (NASCIS III) werepublished in 1997 (16). NASCIS III was a double-blind,randomized clinical trial comparing the efficacy of methyl-prednisolone administered for 24 hours with that ofmethylprednisolone administered for 48 hours and tirilazadmesylate administered for 48 hours. There was no placebogroup. Entry criteria were similar to those described for NAS-CIS II study patients. Patients were assessed neurologicallyaccording to NASCIS I and II (change in motor and sensoryscores) and by change in FIM at 6 weeks and 6 months. Fourhundred ninety-nine patients were entered into the study, 166in the 24-hour methylprednisolone group (24 MP), 167 in the48-hour tirilazad mesylate group (48 TM), and 166 in the48-hour methylprednisolone group (48 MP). The authors re-ported that patients in the 48 MP group showed improvedmotor recovery at 6 weeks (P � 0.09) and at 6 months (P �0.07) follow-up compared with 24 MP patients and 48 TMpatients. When therapy was initiated between 3 and 8 hoursafter injury, the effect of the 48 MP regimen on change inmotor score was significant at 6 weeks (P � 0.04) and at 6months (P � 0.01) follow-up compared with patients in the 24MP and 48 TS treatment groups. 48 MP patients had moreimprovement in FIM at the 6-month follow-up (P � 0.08)compared with patients in the other two treatment groups. 48MP treatment patients also had higher rates of severe sepsis (P� 0.07) and severe pneumonia (P � 0.02). When treatmentwas initiated within 3 hours of injury, the same recoverypattern was observed in all three treatment groups. The au-thors concluded that patients with ASCI who receive methyl-prednisolone within 3 hours of injury should be maintainedon the 24 MP regimen. When methylprednisolone is admin-istered 3 to 8 hours after injury, they recommended the 48 MPregimen.

In 1998, the 1-year follow-up results of the NASCIS III trialwere reported (17). The authors reported that for patientstreated within 3 hours of injury, recovery rates at 1 year wereequal in all three treatment groups. For patients treated be-



tween 3 and 8 hours after injury, 24 MP patients had dimin-ished motor recovery and 48 MP patients had increased motorrecovery at 1 year (P � 0.053). They noted no significantdifference in functional outcome as measured by FIM in anytreatment group. The authors concluded that if methylpred-nisolone is administered to patients with ASCI within 3 hoursof injury, 24-hour maintenance is recommended. If methyl-prednisolone is administered 3 to 8 hours after injury, theyrecommended that a 48-hour maintenance regimen be fol-lowed. These final recommendations seem to be based onmotor recovery score improvement alone (P � 0.053).

Predominant criticisms of the NASCIS III study and thereported results focus on three major issues: determination ofoptimum timing of therapy, method of motor assessment ofSCI patients, and insignificant differences in motor recoveryscores and functional outcome measures among study pa-tients (18, 19, 32, 33, 37, 51). For optimum timing of therapy,time-to-treatment data were not offered or explained. Like the8-hour time for treatment cutoff “result” that came from theNASCIS II study, the “within 3 hours of injury” versus the “3to 8 hours after injury” timeframes reported in NASCIS IIIseem arbitrary (18, 32, 37). It is not intuitive or likely that the3-hour treatment time is an “all or nothing” time periodsupported by physiological evidence. With respect to themethod of motor assessment and reporting, like the NASCISII study, NASCIS III motor scores were reported as change inmotor scores from the right side of the body. Left-side motorscores and total body motor scores were not provided. Thefailure to provide this study’s scientific evidence (particularlyin light of the NASCIS I and II criticisms) suggests that thechanges in right-side only motor scores are the only findingsthat approach significance at 1 year (P � 0.053) and argueagainst the meaningful nature of the data as interpreted andprovided by the authors (18, 32, 37). Finally, the clinical sig-nificance of the changes in motor scores between groups, inlight of the nonsignificant differences in patient function asdetermined by FIM scores, is not evident. NASCIS III patientswho received 48 MP treatment had a 2-fold higher incidenceof severe pneumonia, a 4-fold higher incidence of severesepsis, and a 6-fold higher incidence of death due to respira-tory complications than patients in the 24 MP treatment group(8, 32). These differences, although not statistically significant,raise questions about the safety of the 48-hour treatmentstrategy proposed for patients with ASCI treated within 3 to 8hours of injury. Additional important criticisms of theNASCIS III trial include those levied against both the NASCISI and II studies (i.e., lack of standardized medical treatment,lack of a minimum motor impairment for inclusion [hence,normal motor function patients admitted to the study], novertebral level of injury cutoff, and unclear statistical meth-odology, analysis, and data interpretation) (18, 32, 37). NAS-CIS III was designed and implemented to be a randomized,double-blind clinical study in an attempt to generate Class Ievidence on the efficacy of methylprednisolone, offered intwo different treatment regimens, and tirilazad mesylate afterASCI in human subjects. The absence of evidence for func-tional improvement in any group argues against the clinicalrelevance of any of these regimens.

Wing et al. (49) examined the effect of methylprednisoloneadministered per the NASCIS II protocol on avascular necro-sis (AVN) of the femoral heads of 91 ASCI patients, 59 whoreceived the corticosteroid, and 32 who did not. The authorsfound no case of AVN in their study population and estimatethe relative risk of AVN with high-dose 24-hour methylpred-nisolone therapy to be less than 5%.

In 2000, Pointillart et al. (38) reported the results of a pro-spective, randomized clinical trial designed to evaluate thesafety and effect of nimodipine, methylprednisolone, or bothversus no pharmacological therapy in 106 ASCI patients. Pa-tients were randomly assigned to one of four treatmentgroups, methylprednisolone per NASCIS II protocol (M), ni-modipine (N), both methylprednisolone and nimodipine(MN), and neither medication (P). Blinded neurological as-sessment was accomplished via the American Spinal CordInjury Association (ASIA) score at initiation of treatment andat 1-year follow-up. The authors performed early spinal de-compression and stabilization as indicated. One hundred pa-tients were available at 1-year follow-up. There was no sig-nificant difference in outcome among the four treatmentgroups for any of the ASIA scores recorded. Patients in allfour treatment groups demonstrated significant neurologicalimprovement at the 1-year follow-up compared with admis-sion (P � 0.0001). Two-way analysis of variance revealed nointeraction between methylprednisolone and nimodipine.There was a significant difference in recovery below the levelof injury among patients with complete SCIs compared withthose with incomplete injuries (P � 0.0001). Improvementamong complete injury patients, when present, involved thelevel of the lesion and the two adjacent caudal levels. Thegreatest neurological improvements were identified in incom-plete injury patients. There was no significant difference inneurological outcome for patients who underwent surgerywithin 8 hours of injury, patients treated surgically between 8and 24 hours after injury, and those managed without sur-gery. The incidence of infectious complications was higheramong patients treated with methylprednisolone comparedwith those who did not receive corticosteroids (66% versus45%), but this difference was not significant. The authorsconcluded that pharmacological therapy offered no addedbenefit to patients with ASCIs. Unfortunately, sample sizecalculations are not provided by the authors, and thereforethe statistical power of the study to show a significant benefitof the treatment(s) is unknown. In addition, indications forsurgery and for timing of surgery were not provided, poten-tially adding bias. The failure to show a difference betweengroups in this study may be explained by these potentialstudy design flaws.

A number of published critiques of the NASCIS data andtheir presentation in support of the use of methylprednisolonein the management of patients with ASCI have been offered(1, 18, 19, 22, 31–33a, 35, 37, 40–42, 44–46, 51). A recent medicalevidence-based review is provided by Short et al. (46). Theseauthors conclude, after review of the medical literature on theuse of methylprednisolone for ASCI (animal and human ex-perimental studies, including randomized human clinical tri-als), that the available evidence does not support the use of



methylprednisolone in the treatment of ASCI. A number ofreviews that support the use of methylprednisolone afterASCI have also been published, including a Cochrane Data-base of Systematic Reviews (3–5, 6a, 7, 9, 10, 50).

In 2001, Matsumoto et al. (36) reported their results of aprospective, randomized double-blind clinical trial compar-ing methylprednisolone with placebo in the treatment of pa-tients with acute cervical SCI. The authors focused on poten-tial medical complications after ASCI. Forty-six patients wereincluded in the study: 23 treated with methylprednisolone perthe NASCIS II protocol were compared with 23 patients in aplacebo treatment group. Complications associated with ther-apy were noted at 2-month follow-up. Patients treated withmethylprednisolone had a higher incidence of complicationscompared with placebo-treated patients (56.5% versus 34.8%).Respiratory complications (P � 0.009) and gastrointestinalcomplications (P � 0.036) were the most significant betweenthe two treatment populations. The authors concluded thatpatients with ASCI treated with methylprednisolone (partic-ularly older patients) are at increased risk for pulmonary andgastrointestinal complications and deserve special care. Thisincidence of medical complications using methylprednisolonefor 24 hours seems clinically important. The NASCIS III studydemonstrated that these complications are even higher for48-hour methylprednisolone administration as describedabove (17). This calls into question the use of corticosteroidsfor any timeframe, but especially for the 48-hour duration.

Finally, a review of the data in a large number of patients inthe most recent GM-1 ganglioside trial who had methylpred-nisolone alone according to NASCIS II and III protocols didnot confirm the findings of the NASCIS II and III trials (23).This is described in detail in the section below on the GM-1ganglioside trials.

In summary, the available medical evidence does not supporta significant clinical benefit from the administration of methyl-prednisolone in the treatment of patients after ASCI for either 24or 48 hours duration. Three North American, multicenter ran-domized clinical trials have been completed and several otherstudies have been accomplished addressing this issue (11, 13–17,21, 28, 29, 38, 39). The neurological recovery benefit of methyl-prednisolone when administered within 8 hours of ASCI hasbeen suggested but not convincingly proven. The administrationof methylprednisolone for 24 hours has been associated with asignificant increase in severe medical complications. This is evenmore striking for methylprednisolone administered for 48 hours.In light of the failure of clinical trials to convincingly demon-strate a significant clinical benefit of administration of methyl-prednisolone, in conjunction with the increased risks of medicalcomplications associated with its use, methylprednisolone in thetreatment of acute human SCI is recommended as an option thatshould only be undertaken with the knowledge that the evi-dence suggesting harmful side effects is more consistent than thesuggestion of clinical benefit.

GM-1 ganglioside

GM-1 ganglioside has been evaluated in both animal andhuman studies of ASCI (2, 26, 27, 47, 48). In 1991, Geisler et al.

(25) described the results of a prospective, randomized,placebo-controlled, double-blind trial of GM-1 ganglioside inthe treatment of human patients with ASCI. Of 37 patientsentered into the study, 34 were available for 1-year follow-up(16 GM-1 patients, 18 placebo). All patients received a 250-mgbolus of methylprednisolone and then 125 mg every 6 hoursfor 72 hours. GM-1 patients were administered 100 mg ofGM-1 ganglioside per day for 18 to 32 days, with the first doseprovided within 72 hours of injury. Neurological evaluationwas accomplished with Frankel scale and ASIA motor scoreassessments. The authors reported that GM-1 ganglioside-treated patients had significant improvements in the distribu-tion of Frankel grades from baseline to 1-year follow-up (P �0.034) and significantly improved ASIA motor scores com-pared with placebo-treated patients (P � 0.047) (26, 27). Therecovery of motor function in GM-1 ganglioside-treated pa-tients was thought to caused by recovery of strength in par-alyzed muscles rather than strengthening of paretic muscles.There were no adverse effects attributed to the administrationof the study drug. The authors concluded that GM-1 gangli-oside enhances neurological recovery in human patients afterSCI and deserves further study.

In 1992, a multicenter GM-1 ganglioside ASCI study wasinitiated. It was a prospective, double-blind, randomized andstratified trial that enrolled 797 patients by study end in early1997 (23). All patients received methylprednisolone per theNASCIS II protocol. Patients were randomized into threeinitial study groups: placebo, low-dose GM-1 (300-mg loadingdose and then 100 mg/d for 56 d), and high-dose GM-1(600-mg loading dose and then 200 mg/d for 56 d). Placebo orGM-1 was administered at the conclusion of the 23-hourmethylprednisolone infusion. Patients were assessed usingthe modified Benzel Classification and the ASIA motor andsensory examinations at 4, 8, 16, 26, and 52 weeks after injury.Aggressive medical and surgical management paradigmswere used. Patients had to have an acute, nonpenetrating SCI(anatomic vertebral level C2 through T11) of at least moderateseverity (no neurologically normal or nearly normal patients).The primary efficacy assessment was the proportion of pa-tients who improved at least two grades from baseline exam-ination (defined as “marked recovery”), at Week 26 of thestudy. Secondary efficacy assessments included the timecourse of marked recovery, the ASIA motor score, and ASIAsensory evaluations, relative and absolute sensory levels ofimpairment, and assessments of bladder and bowel function.A planned interim analysis of the first 180 patients resulted inthe addition of stratification by patient age and discontinua-tion of the high-dose GM-1 treatment strategy because of anearly trend for higher mortality. At the study conclusion, 37patients were judged ineligible, leaving 760 patients for pri-mary efficacy analysis. The authors found no significant dif-ference in mortality between treatment groups (23). The au-thors did not identify a higher proportion of patients withmarked recovery in motor function at 26 weeks when theycompared GM-1-treated patients to the placebo treatmentgroup in their primary efficacy analysis. The time course ofrecovery indicated earlier attainment of marked recovery inGM-1-treated patients. The authors concluded that, despite



the lack of statistical significance in the primary analysis,numerous positive secondary analyses indicate that GM-1ganglioside is a useful drug in the management of ASCI (23).The placebo group within this study of GM-1 represents agroup of 322 patients who received methylprednisolonewithin 8 hours of injury. Of interest, these 322 patients (mea-sured in a similar, albeit, more detailed manner as NASCIS IIpatients) did not demonstrate the previously published neu-rological examination improvement found in 62 NASCIS IIpatients treated within the same timeframe (13, 14). Similarly,218 of these patients received 24 hours methylprednisolonetreatment within 3 hours of injury, as suggested in NASCISIII, and did not show the same neurological examinationmotor improvement as the 75 NASCIS III patients who re-ceived the same regimen (16, 17). The authors could notconfirm the NASCIS findings that timing of methylpred-nisolone therapy had an impact on spinal cord recovery. Thisfurther brings into question the conclusions of the NASCIS IIand III methylprednisolone trials.

In summary, the available medical evidence does not sup-port a significant clinical benefit from the administration ofGM-1 ganglioside in the treatment of patients after ASCI. TwoNorth American multicenter, randomized clinical trials havebeen completed addressing this issue (23, 29). The neurolog-ical recovery benefit of GM-1 ganglioside when administeredfor 56 days after the administration of methylprednisolonewithin 8 hours of ASCI has been suggested but not convinc-ingly proven. At present, GM-1 ganglioside (a 300-mg loadingdose and then 100 mg/d for 56 d), when initiated after theadministration of methylprednisolone given within 8 hours ofinjury (NASCIS II protocol), is recommended as an option inthe treatment of adult patients with ASCI.


Given the problems associated with the many trials at-tempting to answer the questions surrounding the use ofpharmacological agents in acute spinal cord-injured patients,it is clear that more research is required. Issues such as ade-quate numbers of patients to achieve statistical power, a pla-cebo group as one of the treatment arms, standardized med-ical and surgical protocols to diminish bias, careful collectionof relevant outcome data, especially functional outcomes, andappropriate statistical analyses need to be further addressed apriori. Research into all potentially promising pharmacologicalagents, including, but not limited to, tirilazad mesylate, nalox-one, methylprednisolone, and GM-1, should be undertaken.


REFERENCES

1. Acland RH, Anthony A, Inglis GS, Walton DI, Xiong X: Methylpred-nisolone use in acute spinal cord injury. N Z Med J 9:99, 2001 (letter).

2. Amar PA, Levy ML: Pathogenesis and pharmacological strate-gies for mitigating secondary damage in acute spinal cord injury.Neurosurgery 44:1027–1040, 1999.

3. Bracken MB: National Acute Spinal Cord Injury Study of methyl-prednisolone or naloxone. Neurosurgery 28:628–629, 1991 (letter).

4. Bracken MB: Methylprednisolone and spinal cord injury.J Neurosurg 93:175–177, 2000 (letter).

5. Bracken MB: The use of methylprednisolone. J Neurosurg 93:340–341, 2000 (letter).

6. Bracken MB: High dose methylprednisolone must be given for 24 or48 hours after acute spinal cord injury. BMJ 322:862–863, 2001 (letter).

6a. Bracken MB: Methylprednisolone and acute spinal cord injury: Anupdate of the randomized evidence. Spine 26(24 Suppl):S47–S54, 2001.

7. Bracken MB: Pharmacological interventions for acute spinal cordinjury. Cochrane Database Syst Rev 1:1–32, 2001.

8. Bracken MB, Holford TR: Effects of timing of methylprednisoloneor naloxone administration on recovery of segmental and long-tractneurological function in NASCIS 2. J Neurosurg 79:500–507, 1993.

9. Bracken MB, Holford TR: Response: Treatment of spinal cordinjury. J Neurosurg 80:954–955, 1994 (letter).

10. Bracken MB, Aldrich EF, Herr DL, Hitchon PW, Holford TR,Marshall LF, Nockels RP, Pascale V, Shepard MJ, Sonntag VKH,Winn HR, Young W: Clinical measurement, statistical analysis,and risk-benefit: Controversies from trials of spinal injury.J Trauma 48:558–561, 2000.

11. Bracken MB, Collins WF, Freeman DF, Shepard MJ, Wagner FW,Silten RM, Hellenbrand KG, Ransohoff J, Hunt WE, Perot PL Jr,Grossman RG, Green BA, Eisenberg HM, Rifkinson N, GoodmanJH, Meagher JN, Fischer B, Clifton GL, Flamm ES, Rawe SE:Efficacy of methylprednisolone in acute spinal cord injury.JAMA 251:45–52, 1984.

12. Bracken MB, Shepard MJ, Collins WF, Holford TR, Baskin DS,Flamm E, Eisenberg HM, Leo-Summers L, Maroon JC, Marshall LF,Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC Jr, Wilberger JL,Winn HR, Young W: Response: Methylprednisolone for spinal cordinjury. J Neurosurg 77:325–327, 1992 (letter).

13. Bracken MB, Shepard MJ, Collins WF Jr, Holford TR, Baskin DS,Eisenberg HM, Flamm E, Leo-Summers L, Maroon JC, Marshall LF,Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC Jr, Wilberger JE,Winn HR, Young W: Methylprednisolone or naloxone treatmentafter acute spinal cord injury: 1-year follow-up data—Results of theSecond National Acute Spinal Cord Injury Study. J Neurosurg76:23–31, 1992.

14. Bracken MB, Shepard MJ, Collins WF, Holford TR, Young W,Baskin DS, Eisenberg HM, Flamm E, Leo-Summers L, Maroon J,Marshall LF, Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC,Wilberger JE, Winn HR: A randomized, controlled trial of meth-ylprednisolone or naloxone in the treatment of acute spinal cordinjury: Results of the Second National Acute Spinal Cord InjuryStudy (NASCIS-2). N Engl J Med 322:1405–1411, 1990.

15. Bracken MB, Shepard MJ, Hellenbrand KG, Collins WF, Leo-Summers L, Freeman DF, Wagner FC, Flamm ES, Eisenberg HM,Goodman JH, Perot PL Jr, Green BA, Grossman RG, Meagher JN,Young W, Fischer B, Clifton GL, Hunt WE, Rifkinson N: Meth-ylprednisolone and neurological function 1 year after spinal cordinjury: Results of the National Acute Spinal Cord Injury Study.J Neurosurg 63:704–713, 1985.

16. Bracken MB, Shepard MJ, Holford TR, Leo-Summers L, AldrichEF, Fazl M, Fehlings M, Herr DL, Hitchon PW, Marshall LF,Nockels RP, Pascale V, Perot PL Jr, Piepmeier JM, Sonntag VKH,Wagner F, Wilberger JE, Winn HR, Young W: Administration ofmethylprednisolone for 24 or 48 hours or tirilazad mesylate for48 hours in the treatment of acute spinal cord injury: Results ofthe Third National Acute Spinal Cord Injury Randomized Con-trolled Trial—National Acute Spinal Cord Injury Study. JAMA277:1597–1604, 1997.



17. Bracken MB, Shepard MJ, Holford TR, Leo-Summers L, AldrichEF, Fazl M, Fehlings MG, Herr DL, Hitchon PW, Marshall LF,Nockels RP, Pascale V, Perot PL Jr, Piepmeier JM, Sonntag VKH,Wagner F, Wilberger JE, Winn HR, Young W: Methylpred-nisolone or tirilazad mesylate administration after acute spinalcord injury: 1-year follow-up—Results of the Third NationalAcute Spinal Cord Injury Randomized Controlled Trial.J Neurosurg 89:699–706, 1998.

18. Coleman WP, Benzel D, Cahill DW, Ducker T, Geisler F, GreenB, Gropper MR, Goffin J, Madsen PW III, Maiman DJ, Ondra SL,Rosner M, Sasso RC, Trost GR, Zeidman S: A critical appraisal ofthe reporting of the National Acute Spinal Cord Injury Studies(II and III) of methylprednisolone in acute spinal cord injury.J Spinal Disord 13:185–199, 2000.

19. Ducker TB, Zeidman SM: Spinal cord injury: Role of steroidtherapy. Spine 19:2281–2287, 1994.

20. Duh MS, Shepard MJ, Wilberger JE, Bracken MB: The effective-ness of surgery on the treatment of acute spinal cord injury andits relation to pharmacological treatment. Neurosurgery 35:240–249, 1994.

21. Galandiuk S, Raque G, Appel S, Polk HC Jr: The two-edgedsword of large-dose steroids for spinal cord trauma. Ann Surg218:419–427, 1993.

22. Geisler FH: Commentary on NASCIS-2. J Spinal Disord 5:132–133,1992 (comment).

23. Geisler FH, Coleman WP, Grieco G, Dorsey FC, Poonian D, TheSygen Study Group: The GM1 ganglioside multi-center acutespinal cord injury study. Spine 26(24 Suppl):S87–S98, 2001.

24. Geisler FH, Dorsey FC, Coleman WP: Correction: Recovery of motorfunction after spinal cord injury—A randomized, placebo-controlledtrial with GM-1 ganglioside. N Engl J Med 325:1659–1660, 1991.

25. Geisler FH, Dorsey FC, Coleman WP: Recovery of motor func-tion after spinal cord injury: A randomized, placebo-controlledtrial with GM-1 ganglioside. N Engl J Med 324:1829–1838, 1991.

26. Geisler FH, Dorsey FC, Coleman WP: GM-1 ganglioside in hu-man spinal cord injury. J Neurotrauma 9[Suppl 2]:S517–S530,1992.

27. Geisler FH, Dorsey FC, Coleman WP: GM-1 ganglioside forspinal cord injury. N Engl J Med 326:494, 1992.

28. George ER, Scholten DJ, Buechler CM, Jordan-Tibbs J, Mattice C,Albrecht RM: Failure of methylprednisolone to improve theoutcome of spinal cord injuries. Am Surg 61:659–664, 1995.

29. Gerhart KA, Johnson RL, Menconi J, Hoffman RE, LammertseDP: Utilization and effectiveness of methylprednisolone in apopulation-based sample of spinal cord injured persons. Para-plegia 33:316–321, 1995.

30. Gerndt SJ, Rodriguez JL, Pawlik JW, Taheri PA, Wahl WL,Micheals AJ, Papadopoulos SM: Consequences of high-dose ste-roid therapy for acute spinal cord injury. J Trauma 42:279–284,1997.

31. Hanigan WC, Anderson RJ: Commentary on NASCIS-2. J SpinalDisord 5:125–131, 1992.

32. Hurlbert RJ: Methylprednisolone for acute spinal cord injury: Aninappropriate standard of care. J Neurosurg 93[Suppl 1]:1–7, 2000.

33. Hurlbert RJ: The use of methylprednisolone. J Neurosurg93[Suppl 1]:340–341, 2000 (letter).

33a. Hurlbert RJ: The role of steroids in acute spinal cord injury: Anevidence-based analysis. Spine 26(24 Suppl):S39–S46, 2001.

34. Landi G, Ciccone A: GM-1 ganglioside for spinal cord injury.N Engl J Med 326:493, 1992 (letter).

35. Lyons MK, Partington MD, Meyer FB: A randomized, controlledtrial of methylprednisolone or naloxone in the treatment of acutespinal-cord injury. N Engl J Med 323:1207–1208, 1990.

36. Matsumoto T, Tamaki T, Kawakami M, Yoshida M, Ando M,Yamada H: Early complications of high-dose methylpred-nisolone sodium succinate treatment in the follow-up of acutecervical spinal cord injury. Spine 26:426–430, 2001.

37. Nesathurai S: Steroids and spinal cord injury: Revisiting theNASCIS 2 and NASCIS 3 trials. J Trauma 45:1088–1093, 1998.

38. Pointillart V, Petitjean ME, Wiart L, Vital JM, Lassie P, ThicoipeM, Dabadie P: Pharmacological therapy of spinal cord injuryduring the acute phase. Spinal Cord 38:71–76, 2000.

39. Poynton AR, O’Farrell DA, Shannon F, Murray P, McManus F,Walsh MG: An evaluation of the factors affecting neurological re-covery following spinal cord injury. Injury 28:545–548, 1997.

40. Rosner MJ: National acute spinal cord injury study of methyl-prednisolone or naloxone. Neurosurgery 28:628, 1991 (letter).

41. Rosner MJ: Methylprednisolone for spinal cord injury.J Neurosurg 77:324–325, 1992.

42. Rosner MJ: Treatment of spinal cord injury. J Neurosurg 80:954–955, 1994.

43. Schönhöfer PS: GM-1 ganglioside for spinal cord injury. N EnglJ Med 326:493, 1992 (letter).

44. Shapiro SA: Methylprednisolone for spinal cord injury.J Neurosurg 77:324–327, 1992.

45. Short DJ: Use of steroids for acute spinal cord injury must bereassessed. BMJ 321:1224, 2000 (letter).

46. Short DJ, El Masry WS, Jones PW: High dose methylpred-nisolone in the management of acute spinal cord injury: A system-atic review from a clinical perspective. Spinal Cord 38:273–286,2000.

47. Tator CH: Experimental and clinical studies of the pathophysi-ology and management of acute spinal cord injury. J SpinalCord Med 19:206–214, 1996.

48. Tator CH: Biology of neurological recovery and functional res-toration after spinal cord injury. Neurosurgery 42:696–708, 1998.

49. Wing PC, Nance P, Connell DG, Gagnon F: Risk of avascularnecrosis following short term megadose methylprednisolonetreatment. Spinal Cord 36:633–636, 1998.

50. Young W, Bracken MB: The Second National Acute Spinal CordInjury Study. J Neurotrauma 9[Suppl 1]:S397–S405, 1992.

51. Zeidman SM, Ling GS, Ducker TB, Ellenbogen RG: Clinicalapplications of pharmacologic therapies for spinal cord injury.J Spinal Disord 9:367–380, 1996.


CHAPTER 10

Deep Venous Thrombosis and Thromboembolism in Patientswith Cervical Spinal Cord Injuries

RECOMMENDATIONSSTANDARDS:• Prophylactic treatment of thromboembolism in patients with severe motor deficits due to spinal cord injury

is recommended.• The use of low-molecular-weight heparins, rotating beds, adjusted dose heparin, or a combination of

modalities is recommended as a prophylactic treatment strategy.• Low-dose heparin in combination with pneumatic compression stockings or electrical stimulation is rec-

ommended as a prophylactic treatment strategy.

GUIDELINES:• Low-dose heparin therapy alone is not recommended as a prophylactic treatment strategy.• Oral anticoagulation alone is not recommended as a prophylactic treatment strategy.

OPTIONS:• Duplex Doppler ultrasound, impedance plethysmography, and venography are recommended for use as

diagnostic tests for deep venous thrombosis in the spinal cord-injured patient population.• A 3-month duration of prophylactic treatment for deep venous thrombosis and pulmonary embolism is

recommended.• Vena cava filters are recommended for patients who do not respond to anticoagulation or who are not

candidates for anticoagulation therapy and/or mechanical devices.

RATIONALE

Deep venous thrombosis (DVT) and pulmonary embo-lism (PE) are problems frequently encountered inpatients who have sustained cervical spinal cord in-

juries (SCIs). Several means of prophylaxis and treatment areavailable, including anticoagulation, pneumatic compressiondevices, and vena cava filters. The purpose of this evidence-based medicine review is to evaluate the literature on themethods of prevention and identification of DVT and PEcomplications in patients after acute cervical SCI.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 through 2001was performed. The following medical subject headings wereused in combination with “spinal cord injury”: “deep venousthrombosis,” “pulmonary embolism,” and “thromboembo-lism.” The search was limited to human studies in the Englishlanguage. This resulted in 129 citations. Duplicate references,reviews, letters, and tangential reports were discarded.Thirty-seven articles dealing with the prophylaxis or treat-

ment of thromboembolic disease in adult SCI patients makeup the basis for this guideline and are summarized in Table10.1. Supporting references included four evidence-based re-views published by various organizations concerned withthromboembolism prophylaxis and treatment in different pa-tient populations. Finally, several series dealing with throm-boembolism in general trauma patients with results germaneto a discussion of SCI patients are included among the refer-ences as supporting documents.


The incidence of thromboembolic complications in the un-treated SCI patient population is high. Depending on injuryseverity, patient age, and the methods used to diagnose athromboembolism, the incidence of thromboembolic eventsranges from 7 to 100% in reported series of patients receivingeither no prophylaxis or inadequate prophylaxis (3, 9, 11, 13,16, 22, 23, 26, 27, 29, 30, 33, 36). Substantial morbidity andmortality have been associated with the occurrence of DVTand PE events in the SCI patient population (6, 7).


TAB

LE10

.1.

Sum

mar

yof

Rep

orts

onD

eep

Ven

ous

Thro

mbo

sis

and

Pulm

onar

yEm

bolis

min

Pati

ents

wit

hC

ervi

cal

Spin

alC

ord

Inju

ries

a

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ceC

lass

Con

clus

ions

Che

net

al.,

1999

(4a)

Larg

epo

pula

tion

ofA

SCI

patie

nts

(164

9)st

udie

dfr

omad

mis

sion

tore

habi

litat

ion

(mea

n19

d)to

disc

harg

e(m

ean,

50d)

.In

cide

nce

ofD

VT

�PE

decl

inin

gov

ertim

e,

but

rem

ains

6.1%

desp

itepr

ophy

laxi

s.

IIID

VT/

PEst

illpr

oble

ms

desp

itepr

ophy

laxi

s.(S

eeM

cKin

ley

etal

.(2

4)

for

follo

w-u

p).

McK

inle

yet

al.,

1999

(24)

Chr

onic

SCI

popu

latio

nst

udie

d.In

cide

nce

ofD

VT

high

est

duri

ng1s

tyr

(2.1

%)

but

then

drop

sof

fto

0.5–

1%/y

rth

erea

fter.

IIIR

isk

ofD

VT/

PEhi

ghes

tdu

ring

1st

yraf

ter

inju

ryan

dth

enri

skdr

ops

sign

ifica

ntly

.

Pow

ell

etal

.,19

99(3

0)In

cide

nce

ofD

VT

inSC

Ipo

pula

tion

(n�

189)

ontr

ansf

erto

reha

bilit

atio

n

(dia

gnos

isw

ithU

S)w

as4.

1%in

grou

pw

hore

ceiv

edpr

ophy

laxi

sve

rsus

16.4

%in

grou

pw

ithou

tpr

ophy

laxi

s.In

prop

hyla

xis

grou

p,D

VTs

only

occu

rred

inpa

tient

s

rece

ivin

ghe

pari

nal

one.

IIPr

ophy

laxi

sde

crea

ses

inci

denc

eof

DV

Tin

SCI

popu

latio

n.H

epar

in

alon

ew

asth

ele

ast

effe

ctiv

em

easu

re.

Rou

ssi

etal

.,19

99(3

2)6

of67

(9%

)of

SCI

patie

nts

deve

lope

dD

VT

desp

itepr

ophy

laxi

sw

ithLM

WH

.

D-d

imer

had

100%

nega

tive

pred

ictiv

eva

lue

com

pare

dw

ithdu

plex

Dop

pler

.

(How

ever

,sp

ecifi

city

only

34%

and

posi

tive

pred

ictiv

eva

lue

13%

).

Ifo

rdi

agno

stic

test

;III

othe

rwis

eIn

cide

nce

ofD

VT

desp

itepr

ophy

laxi

sw

ithLM

WH

still

9%.

D-d

imer

isse

nsiti

vebu

tno

tsp

ecifi

cte

stfo

rD

VT.

Win

emill

eret

al.,

1999

(40)

Ret

rosp

ectiv

est

udy

of42

8SC

Ipa

tient

s.TE

occu

rred

in19

.6%

.C

ompr

essi

on

stoc

king

san

dse

quen

tial

com

pres

sion

devi

ces

low

ered

risk

ofTE

.Ef

fect

sof

low

-

dose

hepa

rin

wer

ese

enin

first

14d

but

wer

eno

tsi

gnifi

cant

.TE

sal

loc

curr

edin

first

150

d.

IIISe

quen

tial

com

pres

sion

devi

ces

and

stoc

king

sre

duce

risk

ofTE

.

Low

-dos

ehe

pari

nm

aybe

effe

ctiv

ein

first

14d

afte

rin

jury

.

Tom

aio

etal

.,19

98(3

5a)

Enox

apar

in(L

MW

H)

vers

ushe

pari

nus

efo

rin

itial

DV

Ttr

eatm

ent

ingr

oup

of6

SCI

patie

nts.

IIIEn

oxap

arin

was

cost

-effe

ctiv

eal

tern

ativ

eto

intr

aven

ous

hepa

rin

for

initi

altr

eatm

ent

ofD

VT.

Har

ris

etal

.,19

96(1

8)R

etro

spec

tive

stud

yof

enox

apar

in(L

MW

H)

in10

5SC

Ipa

tient

s(o

ne-t

hird

inta

ct,

40

com

plet

e).

No

clin

ical

DV

T/PE

in10

5,no

US

evid

ence

in60

.

IIIEn

oxap

arin

issa

fean

def

fect

ive

for

DV

Tpr

ophy

laxi

sin

the

SCI

patie

nt.

Kha

nsar

inia

etal

.,19

95(2

0)R

etro

spec

tive

hist

oric

alco

hort

com

pari

son

ofpr

ophy

lact

icPG

Fin

324

gene

ral

trau

ma

patie

nts.

PGF

grou

pha

dfe

wer

PEs

than

cont

rol

grou

p.

IIIG

reen

field

filte

rsa

fean

def

fect

ive

for

PEpr

ophy

laxi

sin

gene

ral

trau

ma

popu

latio

n.

Gee

rts

etal

.,19

94(1

1)Pr

ospe

ctiv

eev

alua

tion

of71

6tr

aum

apa

tient

s(n

opr

ophy

laxi

s)w

ithV

OP

and

veno

grap

hy.

Inci

denc

eof

DV

Tin

SCI

popu

latio

n(n

�66

)w

as62

%.

IIID

VT

isve

ryco

mm

onin

SCI

patie

nts

ifno

prop

hyla

xis

used

.

Wils

onet

al.,

1994

(39)

Inse

rted

cava

lfil

ters

in15

SCI

patie

nts.

Non

eha

dD

VT

orPE

in1

yr.

Cla

ims

this

resu

ltsu

peri

orto

hist

oric

alco

ntro

ls(n

oev

iden

cepr

esen

ted

tosu

ppor

tth

iscl

aim

).

1-yr

pate

ncy

rate

was

81%

.

IIIIn

sert

ion

ofca

val

filte

rsap

pear

sto

besa

fein

SCI

patie

nts.

Gre

enet

al.,

1994

(12)

His

tori

cal

coho

rtco

mpa

riso

nof

LMW

Han

dst

anda

rdan

dad

just

eddo

sehe

pari

n

prop

hyla

xis.

Trau

ma

patie

nts

trea

ted

with

8-w

kco

urse

ofLM

WH

had

few

er

blee

ding

epis

odes

(P�

0.05

)an

dTE

com

plic

atio

ns(P

�0.

15)

than

thos

etr

eate

d

with

hepa

rin.

IIILM

WH

may

besa

fer

and

mor

eef

fect

ive

for

prop

hyla

xis

than

min

idos

eor

adju

sted

dose

hepa

rin.

Gun

duz

etal

.,19

93(1

6)31

SCI

patie

nts

onlo

w-d

ose

hepa

rin

ther

apy

unde

rwen

tve

nogr

aphy

.In

cide

nce

of

DV

Tw

as53

.3%

.

IIIIn

cide

nce

ofD

VT

high

inSC

Ipa

tient

sde

spite

low

-dos

ehe

pari

n

(ther

apy

star

ted

onre

habi

litat

ion

unit)

.

Bur

nset

al.,

1993

(3)

Pros

pect

ive

asse

ssm

ent

ofD

VT

inhi

gh-r

isk

trau

ma

patie

nts

with

US.

Foun

d

inci

denc

eof

21%

(12/

57)

desp

itelo

w-d

ose

hepa

rin

orpn

eum

atic

boot

sin

85%

.

IIID

VT

isco

mm

onde

spite

use

oflo

w-d

ose

hepa

rin

orpn

eum

atic

boot

s.

Lam

bet

al.,

1993

(23)

287

chro

nica

llyin

jure

dSC

Ipa

tient

sfo

llow

ed.

Ove

rall

inci

denc

eof

TEev

ents

was

10%

,m

ost

even

tsin

first

6m

o.

IIIPr

ophy

lact

icth

erap

yno

tne

cess

ary

beyo

nd6

mo

inSC

Ipo

pula

tion.

Kul

karn

iet

al.,

1992

(22)

100

SCI

patie

nts

pros

pect

ivel

ytr

eate

dw

ithlo

w-d

ose

hepa

rin.

26%

inci

denc

eof

clin

ical

lyde

tect

edD

VT

(9%

PE)

note

d.

IIID

VT

and

PEin

cide

nce

sign

ifica

ntde

spite

low

-dos

esu

bcut

aneo

us

hepa

rin.

Mer

liet

al.,

1992

(25)

Hep

arin

�pn

eum

atic

stoc

king

seq

ual

tohi

stor

ical

cont

rols

ofhe

pari

n�

stim

ulat

ion

and

bette

rth

anhi

stor

ical

cont

rols

ofhe

pari

nor

plac

ebo

inSC

Ipa

tient

s.

IILo

w-d

ose

hepa

rin

�pn

eum

atic

hose

safe

and

effe

ctiv

eas

DV

T

prop

hyla

xis

inSC

Ipa

tient

s.

War

ing

and

Kar

unas

,19

91(3

6)La

rge

data

base

(141

9)of

SCI

patie

nts.

Inci

denc

eof

DV

Tw

as14

.5%

,PE

4.6%

.

Seve

rity

ofin

jury

was

apr

edic

tor

ofD

VT

and

age

was

apr

edic

tor

ofPE

.N

o

men

tion

mad

eof

prop

hyla

ctic

mea

sure

s.

IIID

VT

and

PEar

esi

gnifi

cant

prob

lem

sin

SCI

popu

latio

n.A

gean

d

inju

ryse

veri

tyne

edto

bead

dres

sed

inst

udie

sco

mpa

ring

trea

tmen

t

mod

aliti

es.



TAB

LE10

.1.

Con

tinu

ed

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ceC

lass

Con

clus

ions

Yel

nik

etal

.,19

91(4

1)Pr

ospe

ctiv

est

udy

of12

7SC

Ipa

tient

sw

ithph

lebo

grap

hy.

29/1

27ha

dD

VT

on

adm

issi

onto

reha

bilit

atio

nun

it.O

f87

patie

nts

with

initi

ally

nega

tive

stud

ies,

12

deve

lope

dD

VT

desp

itepr

ophy

laxi

sfo

rup

to80

d.

IIIIn

cide

nce

ofD

VT

inSC

Ipo

pula

tion

ishi

gh,

and

high

-ris

kpe

riod

exte

nds

toen

dof

3rd

mo.

Aut

hors

reco

mm

end

peri

odic

scre

enin

g

with

phle

bogr

aphy

.

Bal

shi

etal

.,19

89(1

)C

ase

seri

esof

13qu

adri

pleg

icpa

tient

sw

hoha

dve

naca

val

filte

rspl

aced

for

DV

T

orPE

.A

bnor

mal

ities

ofth

efil

ter

wer

ede

tect

edin

5/11

patie

nts

who

had

follo

w-u

p

x-ra

ys:

2pa

tient

sre

quir

edla

paro

tom

yto

rem

ove

filte

rs,

4ha

ddi

stal

mig

ratio

n,an

d

2ha

dna

rrow

ing

ofdi

amet

eras

soci

ated

with

cava

loc

clus

ion.

9of

thes

e11

patie

nts

wer

etr

eate

dw

ithth

e“q

uad

coug

h”te

chni

que.

IIIFi

lter

plac

emen

tm

aybe

asso

ciat

edw

ithsi

gnifi

cant

long

-ter

m

mor

bidi

tyin

the

SCI

popu

latio

n,pa

rtic

ular

lyth

ose

requ

irin

g

aggr

essi

vepu

lmon

ary

toile

t.

DeV

ivo

etal

.,19

89(7

)Ep

idem

iolo

gica

lst

udy

ofca

uses

ofde

ath

for

SCI

patie

nts.

Hig

hest

ratio

sof

actu

alto

expe

cted

caus

esof

deat

hw

ere

for

pneu

mon

ia,

PE,

and

sept

icem

ia.

The

risk

ratio

for

TEdr

oppe

dsi

gnifi

cant

lyaf

ter

the

1st

mo

post

inju

rybu

tre

mai

ned

elev

ated

at6

mo

post

inju

ry.

IIITE

isa

sign

ifica

ntpr

oble

mfo

rpa

tient

sw

hosu

rviv

eSC

I.Pe

riod

of

grea

test

risk

isin

first

few

mon

ths

afte

rin

jury

,bu

tri

skco

ntin

ues

even

afte

r6

mo.

Gre

enet

al.,

1988

(13)

Ran

dom

ized

cont

rolle

dtr

ial

oflo

w-d

ose

vers

usad

just

eddo

sehe

pari

nin

SCI

patie

nts.

Rat

eof

TElo

wer

inad

just

eddo

segr

oup

(7%

vers

us31

%)

(inte

ntto

trea

t,

P�

NS)

,bu

tal

soha

dhi

gher

rate

ofbl

eedi

ngco

mpl

icat

ions

(7/2

9).

IA

djus

ted

dose

hepa

rin

mor

eef

fect

ive

than

low

-dos

ehe

pari

n,

blee

ding

mor

eco

mm

onin

adju

sted

dose

grou

p.

Mer

liet

al.,

1988

(26)

Pros

pect

ive

rand

omiz

edtr

ial

ofpl

aceb

ove

rsus

min

idos

ehe

pari

nve

rsus

hepa

rin

�

elec

tric

alst

imul

atio

nin

grou

pof

48SC

Ipa

tient

s.H

epar

ingr

oup

�pl

aceb

ogr

oup

at50

%,

stim

ulat

edgr

oup

sign

ifica

ntly

few

erD

VT.

ILo

w-d

ose

hepa

rin

nobe

tter

than

plac

ebo,

hepa

rin

�el

ectr

ical

stim

ulat

ion

muc

hbe

tter

for

DV

Tpr

ophy

laxi

sin

SCI

patie

nts.

Wei

ngar

den

etal

.,19

88(3

8a)

Ret

rosp

ectiv

ere

view

of14

8SC

Ipa

tient

s.10

had

docu

men

ted

DV

Tor

PE.

Of

6

patie

nts

who

had

adeq

uate

reco

rds,

all

6ha

dfe

ver

asa

pres

entin

gsi

gn,

4ha

dno

othe

rcl

inic

alsi

gns

reco

rded

.A

llep

isod

esoc

curr

edin

first

12w

k.

IIIFe

ver

may

indi

cate

TEdi

seas

ein

SCI

patie

nts.

Bec

ker

etal

.,19

87(2

)R

ando

miz

edtr

ial

ofro

tatin

gve

rsus

nonr

otat

ing

beds

inth

eac

ute

setti

ngaf

ter

SCI

(10

d).

n�

15pa

tient

s.Pl

ethy

smog

raph

yan

dfib

rino

gen

leg

scan

sus

ed.

IR

otat

ing

beds

redu

ceth

ein

cide

nce

ofD

VT

duri

ngth

efir

st10

daf

ter

SCI.

Tato

r,19

87(3

3)17

%in

cide

nce

ofD

VT

inse

ries

of20

8SC

Ipa

tient

s.In

cide

nce

was

high

erin

oper

ated

patie

nts

(23%

)co

mpa

red

with

nono

pera

ted

(10%

).U

seof

prop

hyla

xis

is

not

men

tione

d.

IIIPa

tient

sre

quir

ing

surg

ery

may

have

high

erin

cide

nce

ofD

VT.

Chu

etal

.,19

85(5

)C

ompa

riso

nbe

twee

nD

oppl

erU

S,V

OP,

and

clin

ical

exam

inat

ion

inSC

Ipa

tient

s.

All

had

sens

itivi

tyan

dsp

ecifi

city

of10

0%in

smal

l(n

�21

)se

ries

.O

vera

ll

inci

denc

e,19

%(C

lass

IIIbe

caus

eno

gold

stan

dard

used

).

IIID

oppl

erU

S,V

OP,

and

clin

ical

exam

inat

ion

all

good

for

diag

nosi

sof

DV

T.

Myl

lyne

net

al.,

1985

(27)

Com

pare

din

cide

nce

ofD

VT

inim

mob

ilize

dsp

ine-

inju

red

patie

nts

with

and

with

out

para

lysi

s.Th

ose

with

para

lysi

sha

da

100%

DV

Tin

cide

nce

(fibr

inog

en

scan

)ve

rsus

0%fo

rpa

tient

sim

mob

ilize

daf

ter

spin

alfr

actu

rew

ithou

tpa

raly

sis.

IIIIn

cide

nce

ofD

VT

isve

ryhi

ghin

SCI

patie

nts

and

isno

tto

tally

depe

nden

ton

imm

obili

zatio

n.

ElM

asri

and

Silv

er,

1981

(8)

Ret

rosp

ectiv

ere

view

of10

2pa

tient

sw

ithSC

I.Th

ere

wer

e21

epis

odes

ofPE

in19

patie

nts.

No

patie

ntw

ithPE

was

adeq

uate

lyan

ticoa

gula

ted

atth

etim

eof

the

PE

(ora

lph

enin

dion

e).

Onl

y8/

19pa

tient

sha

dev

iden

ceof

DV

Tby

exam

inat

ion

or

VO

P.

IIIA

utho

rsre

com

men

dpr

olon

ged

trea

tmen

t(�

6m

o)in

patie

nts

with

obes

ityor

prio

rhi

stor

yof

DV

T.

Fris

bie

and

Sasa

hara

,19

81(9

)Sm

all,

pros

pect

ive

cont

rolle

dst

udy

oflo

w-d

ose

(500

0un

itstw

ice

daily

)he

pari

n

vers

usco

ntro

lgr

oup.

No

diffe

renc

ein

inci

denc

eof

DV

Tno

ted

(onl

y7%

inea

ch

grou

p).

Aut

hors

sugg

est

prot

ectiv

eef

fect

offr

eque

ntph

ysio

ther

apy.

IIN

odi

ffere

nce

betw

een

low

-dos

ehe

pari

nan

dco

ntro

lgr

oups

inSC

I

patie

nts

rece

ivin

gtw

ice

daily

phys

ioth

erap

y.

Perk

ash

etal

.,19

80(2

8a)

Trea

tmen

tof

8pa

tient

sw

ithTE

disc

usse

d.A

utho

rsus

edhe

pari

nan

dth

enw

arfa

rin

with

reas

onab

lere

sults

.

IIIA

ntic

oagu

latio

nis

effe

ctiv

etr

eatm

ent

for

SCI

patie

nts

with

TE.

Perk

ash

etal

.,19

78(2

9)In

cide

nce

ofTE

in48

SCI

patie

nts

was

18%

.C

linic

alex

amin

atio

nse

nsiti

vity

89%

,

spec

ifici

ty88

%,

nega

tive

pred

ictiv

eva

lue

97%

,po

sitiv

epr

edic

tive

valu

e62

%.

One

-thi

rdof

TEev

ents

occu

rred

�12

wk

afte

rin

jury

.

Ifo

rdi

agno

stic

test

;III

othe

rwis

eC

linic

alex

amin

atio

nap

pear

sto

bequ

itego

odfo

rde

tect

ion

ofD

VT

insu

bacu

tese

tting

.Pe

riod

ofri

skm

ayex

tend

beyo

nd12

wk.

Wat

son,

1978

(38)

Ret

rosp

ectiv

ehi

stor

ical

coho

rtst

udy

inve

stig

atin

glo

w-d

ose

hepa

rin

vers

usno

prop

hyla

xis.

IIIH

epar

ingr

oup

had

few

erTE

com

plic

atio

ns.

No

TEev

ents

afte

r3

mo

desp

itepr

ophy

laxi

sce

ssat

ion

at3

mo.

Deep Venous Thrombosis and Thromboembolism S75


Prophylaxis

Prophylactic therapy has been shown to be effective for theprevention of DVT and PE. In a small randomized study,Becker et al. (2) demonstrated that the use of rotating bedsduring the first 10 days after SCI decreased the incidence ofDVT. Four of five control patients were diagnosed with DVT(by fibrinogen screening) compared with one of 10 treatedpatients. The use of low-dose heparin (5000 units given viasubcutaneous injection twice or three times daily) has beendescribed by several authors (4, 9, 16, 17, 22, 30, 38). Hachen(17) published the results of a retrospective historical compar-ison of low-dose heparin versus oral anticoagulation in agroup of 120 SCI patients. He found a lower incidence ofthromboembolic events in the low-dose heparin group com-pared with the oral anticoagulation group. In 1977, Casas et al.(4) reported the results of a prospective assessment of low-dose heparin in SCI patients. These authors administeredheparin for a mean period of 66 days in 18 SCI patients andnoted no thromboembolic events as detected by clinical ex-amination. Watson (38) reported a lower incidence of throm-boembolic events with the use of low-dose heparin than withno prophylaxis in a retrospective historical cohort study. Fris-bie and Sasahara (9), however, found that low-dose heparindid not affect the incidence of DVT in a prospective study of32 SCI patients compared with treatment with twice dailyphysical therapy alone. These authors thought that the lack ofeffect was due to the very low incidence of DVT in theircontrol group compared with other series because of theaggressive physical therapy paradigm used in their patients.Although they performed venous occlusion plethysmography(VOP) screening with confirmatory venography weekly, theincidence of DVT was only 7% in both groups. An identicalobserved frequency of DVT in both treatment groups cannotbe explained by anything other than that the treatments wereequivalent in this study. This incidence is substantially lowerthan that reported by two separate groups of investigators adecade later. In 1992, Kulkarni et al. (22) reported a muchhigher incidence of DVT (26%) and of PE (9%) in a group of100 SCI patients prospectively treated with low-dose heparin.In 1993, Gunduz et al. (16) reported a 53% incidence of DVTconfirmed by venography in 31 patients they managed withSCI treated with low-dose heparin. In a study published in1999, Powell et al. (30) noted that the incidence of DVT in 189SCI patients receiving prophylaxis was significantly lowerthan that identified in SCI patients who did not receive pro-phylaxis, 4.1% versus 16.4%. They found, in addition, thatDVT in the prophylaxis group occurred in patients who re-ceived low-dose heparin alone.

Several studies have demonstrated that better prophylactictherapies than low-dose heparin exist (13, 25, 26). Green et al.(13) published a randomized controlled study comparinglow-dose versus adjusted dose heparin (dose adjusted to 1.5times normal activated partial thromboplastin time) in SCIpatients. They found that patients treated with adjusted doseheparin had fewer thromboembolic events (7% versus 31%)during the course of their 10-week study, but the patients hada higher incidence of bleeding complications. Merli et al. (26),TA

BLE

10.1

.C

onti

nued

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ceC

lass

Con

clus

ions

Cas

aset

al.,

1978

(4)

Pros

pect

ive

asse

ssm

ent

oflo

w-d

ose

hepa

rin

in18

/21

patie

nts

with

SCI

(mea

n

dura

tion,

66d)

.N

opa

tient

trea

ted

had

sym

ptom

atic

DV

Tor

PE.

No

use

of

US/

plet

hysm

ogra

phy/

veno

grap

hy.

IIILo

w-d

ose

hepa

rin

may

beus

eful

for

prev

entio

nof

sym

ptom

atic

DV

T.

Todd

etal

.,19

76(3

4)U

sed

VO

P,fib

rino

gen

scan

,an

dve

nogr

aphy

tost

udy

20SC

Ipa

tient

sfo

r60

d.

Fibr

inog

ensc

anw

aspo

sitiv

ein

all

patie

nts

but

was

conf

irm

edby

anot

her

test

in

only

half

ofth

eca

ses.

IIID

VT

isco

mm

onin

SCI

popu

latio

n.

Hac

hen,

1974

(17)

Coh

ort

cont

rolle

dtr

ial

oflo

w-d

ose

hepa

rin

(500

0un

itsth

ree

times

daily

)ve

rsus

oral

war

fari

nin

SCI

patie

nts.

Hep

arin

grou

pha

dsi

gnifi

cant

lyfe

wer

TEev

ents

.

IILo

w-d

ose

subc

utan

eous

hepa

rin

bette

rth

anor

alw

arfa

rin

for

prop

hyla

xis

afte

rac

ute

SCI.

Nas

o,19

74(2

8)PE

occu

rred

in4/

26pa

tient

sw

ithac

ute

(�3

mo)

SCI

but

none

occu

rred

in17

patie

nts

with

chro

nic

(�3

mo)

SCI.

IIISC

Ipa

tient

spr

imar

ilyat

risk

duri

ngfir

st3

mo

afte

rin

jury

.

Wat

son,

1968

(37)

Inci

denc

eof

TEco

mpl

icat

ions

per

year

rang

esfr

om8

to40

%in

sam

eun

it(n

o

prop

hyla

xis)

.

IIITE

com

plic

atio

nsar

ea

sign

ifica

ntpr

oble

m,

and

ther

eis

vari

abili

ty

year

toye

arde

spite

iden

tical

trea

tmen

tst

rate

gies

.

aSC

I,sp

inal

cord

inju

ry;A

SCI,

acut

eSC

I;D

VT,

deep

veno

usth

rom

bosi

s;PE

,pul

mon

ary

embo

lism

;LM

WH

,low

-mol

ecul

ar-w

eigh

the

pari

ns;T

E,th

rom

boem

bolis

m;U

S,ul

tras

ound

;PG

F,pr

ophy

lact

icG

reen

field

filte

r;V

OP,

veno

usoc

clus

ion

plet

hysm

ogra

phy;

NS,

not

sign

ifica

nt.



in 1988, reported their findings of the additive protectiveeffects of electrical stimulation in combination with low-doseheparin, heparin alone, and placebo in 48 SCI patients treatedfor 4 weeks. In this Class I prospective, randomized trial, theyfound that the heparin therapy alone group had an incidenceof DVT similar to that of the placebo group. The combinationof low-dose heparin and electrical stimulation significantlydecreased the incidence of DVT (1 of 15 patients comparedwith the other two treatment groups [8 of 16 low-dose heparinalone, and 8 of 17 placebo; P � 0.05]). In 1992, Merli et al. (25)reported that heparin in combination with pneumatic stock-ings was equal to the effectiveness of heparin plus electricalstimulation. The heparin in combination with electrical stim-ulation group and the placebo group for this comparison werea historical cohort. This constitutes Class III evidence becausethe comparison group is a historical one. Winemiller et al. (40)studied a large series of 428 SCI patients with a multivariateanalysis and found that the use of pneumatic compressiondevices for 6 weeks was associated with a significant decreasein thromboembolic events (odds ratio, 0.5; 95% confidenceinterval, 0.28–0.90). Low-dose heparin treatment seemed tohave a protective effect as well; however, the effect of heparinalone was not statistically significant.

Recently, low-molecular-weight heparins (LMWHs) havebeen studied as prophylactic therapy for thromboembolism inSCI patients. Green et al. (12) treated a series of SCI patientswith 8 weeks of LMWH (tinzaparin) and compared the resultswith those in a historical cohort of patients treated with low-dose or adjusted dose heparin. They found that the use ofLMWHs compared favorably with the use of either heparindosing regimen in terms of fewer thromboembolic events (16of 79 in heparin group versus 7 of 68 in LMWH group; P �0.15) and a significant decrease in bleeding complications (9 of79 in heparin group versus 1 of 68 in LMWH group; P � 0.04).More recently, Harris et al. (18) performed a retrospectivestudy of LMWH (enoxaparin) administration in a series of 105patients with spinal injuries. Forty of their 105 patients hadneurologically complete injuries. No patient treated with LM-WHs exhibited clinical or ultrasound evidence of DVT, and nopatient had a PE. Roussi et al. (32) reported a 9% incidence ofDVT in a study involving 69 SCI patients receiving LMWHs,testimony to the fact that no prophylactic therapy is 100%effective.

The use of inferior vena cava (IVC) filters as prophylacticdevices for thromboembolism has been advocated (19, 20, 31,39). Wilson et al. (39) placed caval filters in 15 SCI patientswho were concurrently treated with either low-dose heparinor pneumatic stockings. None had a PE during a 1-yearfollow-up period. The 1-year patency rate of the IVC was 81%.These authors reported that their results are superior to thosein a historical cohort of 111 patients treated without IVCfilters. Seven of the cohort patients had a PE; however, six ofthe seven were not receiving any prophylaxis at the time oftheir PE. The patient who was receiving DVT prophylaxis hada gunshot wound to the spine (39). Khansarinia et al. (20)described a historical cohort study of 108 general traumapatients treated with prophylactic IVC filter placement. Noneof these patients had a PE. They compared this group with a

historical cohort of 216 patients treated (apparently) witheither low-dose heparin or pneumatic compression devicesbefore the use of IVC filters. Thirteen of these 216 had a PE, ofwhich 9 were fatal. The overall mortality of the filter groupwas lower than that of the control group, but this differencewas not significant (16% versus 22%). Tola et al. (35) haveshown that percutaneous IVC filter placement in the ICUsetting is as safe and is less costly than IVC filter placement inthe operating room or the invasive radiology suite. Theseauthors suggest that IVC interruption is an effective means toprevent PE. Placement of filters is not without complications.Balshi et al. (1), Greenfield (15), and Kinney et al. (21) havedescribed distal migration, intraperitoneal erosion, and symp-tomatic IVC occlusion in patients with SCIs treated with IVCfilters. Balshi et al. (1) hypothesized that quadriplegic patientsare at higher risk for complications from IVC filter placementowing to loss of abdominal muscle tone, as well as the use of the“quad cough” maneuver. No study performed to date comparesthe use of prophylactic IVC filters to the use of modern methodsof pharmacological PE and DVT prophylaxis.

Duration of prophylaxis

Most thromboembolic events seem to occur within the first2 to 3 months after injury. Naso (28) described his experiencewith four patients with PEs in a group of 43 SCI patients. Allfour PE events occurred within 3 months injury. Perkash et al.(29) reported an 18% incidence of thromboembolism in aseries of 48 patients with acute SCIs and 2 patients withtransverse myelitis. Only one patient had a new onset PE 3months after injury; 2 other patients had recurrent PEs3 months after injury owing to existing DVT. Lamb et al. (23)described a series of 287 SCI patients. The overall risk ofthromboembolic events in their patient population was 10%.Most events occurred within the first 6 months after injury.Twenty-two of 28 events they documented occurred withinthe first 3 months of injury. El Masri and Silver (8) reported 21documented events of PE in a series of 102 spine-injuredpatients. Twenty of 21 events occurred within the first 3months after SCI. A pulmonary embolism occurred in a pa-tient with a history of PE whose therapeutic anticoagulationwas discontinued for gallbladder surgery. DeVivo et al. (7)found that patients had a 500-fold higher risk of dying fromPE in the first month after SCI than age- and sex-matchednoninjured patients. This risk decreased with time but re-mained approximately 20 times higher than for normativecontrols 6 months after injury. McKinley et al. (24) studiedchronic spine-injured patients in a rehabilitation center settingand found an incidence of DVT of 2.1% in the first year afterinjury. This incidence decreased to between 0.5 and 1% peryear thereafter. From these data, it is apparent that mostthromboembolic events (DVT and PE) occur within 3 monthsof acute spinal injury. Although late thromboembolic eventscan occur, the risk of these events must be balanced againstthe cost and risks of indefinite anticoagulation. Prolongedprophylactic anticoagulation therapy is not without risk andis associated with bleeding complications (12, 13). Most stud-ies addressing prophylactic treatment for DVT and PE have



used treatment courses of 8 to 12 weeks with success. Forthese reasons, it is recommended that prophylactic treatmentbe continued for 8 to 12 weeks, in SCI patients without othermajor risk factors (previous thromboembolic events, obesity,advanced age), for DVT and PE. Prophylactic treatment maybe discontinued earlier in patients with useful motor functionin the lower extremities, as these patients seem to be at lessrisk for DVT and PE (6, 27).

Diagnosis

The diagnosis of DVT in various studies has been made onthe basis of clinical criteria, Doppler ultrasound examination,VOP, venography, fibrinogen scanning, or D-dimer measure-ment (2–4, 8–11, 13, 14, 16, 17, 22, 26–30, 32, 34, 36–38, 41).Although venography may be considered a “gold standard”examination for DVT, venography is not possible in all pa-tients, is invasive, and is expensive (11). Gunduz et al. (16)reported a 10% incidence of adverse effects from venographyincluding postvenographic phlebitis and allergic reactions.Doppler ultrasound examination and VOP are both less inva-sive, less expensive, and more broadly applicable (11, 30). Thesensitivity and specificity of these examinations when com-pared with venography has been generally reported to rangefrom 80 to 100% (5). Chu et al. (5) compared Doppler ultra-sound and VOP with the clinical examination and found allthree to agree 100% of the time in a small series of 21 patients.Perkash et al. (29) studied a series of 48 SCI patients with dailyphysical examinations and weekly VOP. They found that thesensitivity of the clinical examination compared with VOPwas 89%; the specificity was 88%, the negative predictivevalue was 97%, and the positive predictive value was 62%.Other authors have described the increased sensitivity of fi-brinogen scanning and the use of D-dimer measurements forthe diagnosis of DVT. Increased sensitivity is associated withdecreased specificity. For example, Roussi et al. (32) reported100% sensitivity and 100% negative predictive value withD-dimer determinations compared with Doppler ultrasoundand the clinical examination. However, the specificity ofD-dimer was only 34%, and the positive predictive value wasonly 13%. Similarly, Todd et al. (34) found that fibrinogenscanning was positive in all 20 patients studied in a prospec-tive fashion, but the diagnosis of DVT was confirmed byanother test in only half of the cases. Overall, no single test iscompletely applicable, accurate, and sensitive for the detec-tion of DVT in the SCI patient population. Furthermore, asubstantial number of patients who have PEs are found tohave negative lower extremity venograms (8, 11). The Con-sortium for Spinal Cord Medicine (6) has recommended theuse of ultrasound for the study of patients with suspectedDVT and venography for use when clinical suspicion is strongand the ultrasound examination is negative. On the basis ofthe medical evidence available, these recommendations seemto be sound.

SUMMARY

Thromboembolic disease is a common occurrence in pa-tients who have sustained a cervical SCI, and it is associated

with significant morbidity. Class I medical evidence existsdemonstrating the efficacy of several means of prophylaxisfor the prevention of thromboembolic events. Therefore, pa-tients with SCI should be treated with a regimen aimed atprophylaxis. Although low-dose heparin therapy has beenreported to be effective as prophylaxis for thromboembolismin several Class III studies, other Class I, II, and III medicalevidence indicates that better alternatives than low-dose hep-arin therapy exist. These alternatives include the use ofLMWH, adjusted dose heparin, or anticoagulation in conjunc-tion with pneumatic compression devices or electrical stimu-lation. Oral anticoagulation alone does not seem to be aseffective as these other measures used for prophylaxis.

The incidence of thromboembolic events seems to decreaseover time, and the prolonged use of anticoagulant therapy isassociated with a definite incidence of bleeding complica-tions. There are many reports of the beneficial effects of theprophylaxis therapy for 6 to 12 weeks after SCI. Very fewthromboembolic events occur beyond 3 months after injury.For these reasons, it is recommended that prophylactic ther-apy be discontinued after 3 months unless the patient is athigh risk (previous thromboembolic events, obesity, ad-vanced age). It is reasonable to discontinue therapy earlier inpatients with retained lower extremity motor function afterSCI, because the incidence of thromboembolic events in thesepatients is substantially lower than in those patients withmotor complete injuries.

Caval filters seem to be efficacious for the prevention of PEin SCI patients. The relative efficacy of caval filters versusprophylactic combination therapy with LMWH and pneu-matic compression stockings has not been studied. Cavalfilters are associated with long-term complications in SCIpatients, although these complications are relatively rare. Ca-val filters are recommended for SCI patients who have hadthromboembolic events despite anticoagulation and for SCIpatients with contraindications to anticoagulation and/or theuse of pneumatic compression devices.

There are several methods available for the diagnosis ofDVT. Venography is considered the “gold standard,” but it isinvasive, not applicable to all patients, and associated withintrinsic morbidity. Duplex Doppler ultrasound and VOPhave been reported to have sensitivities of approximately 90%and are not invasive. It is reasonable to use these noninvasivetests for the diagnosis of DVT and to reserve venography forthe rare situation when clinical suspicion is high and theresults of VOP and ultrasound testing are negative.


Although thromboembolic events in the SCI patient areassociated with significant morbidity, no study has demon-strated improved outcomes in SCI patients as a result ofsurveillance testing for them. A prospective study evaluating6-month outcomes in patients treated with prophylaxis withor without surveillance ultrasound imaging would be a sub-stantial and potentially cost-saving contribution to theliterature.



Caval filters seem to be effective in preventing PEs, andmany institutions are using these devices as first-tier preven-tive therapy without trying other preventive measures. Cavalfilters have not been compared with LMWHs or combinationtherapy with anticoagulants and pneumatic compression de-vices for efficacy in the SCI patient population. As filters doseem to be associated with long-term morbidity in a fractionof SCI patients, a prospective study needs to be performed toestablish whether the potential increase in protection againstPE offsets the risks for long-term complications. A studycomparing the use of vena caval filters prophylactically ver-sus other modes of prevention with the use of filters placedonly after failure of alternative methods should be instituted,including cost-effectiveness outcomes of each mode of pre-vention currently used in SCI patients.


REFERENCES

1. Balshi JD, Cantelmo NL, Menzoian JO: Complications of cavalinterruption by Greenfield filter in quadriplegics. J Vasc Surg9:558–562, 1989.

2. Becker DM, Gonzalez M, Gentili A, Eismont F, Green BA: Pre-vention of deep venous thrombosis in patients with acute spinalcord injuries: Use of rotating treatment tables. Neurosurgery20:675–677, 1987.

3. Burns GA, Cohn SM, Frumento RJ, Degutis LC, Hammers L:Prospective ultrasound evaluation of venous thrombosis inhigh-risk trauma patients. J Trauma 35:405–408, 1993.

4. Casas ER, Sanchez MP, Arias CR, Masip JP: Prophylaxis ofvenous thrombosis and pulmonary embolism in patients withacute traumatic spinal cord lesions. Paraplegia 15:209–214, 1977.

4a. Chen D, Apple DF, Hudson LM, Bode R: Medical complicationsduring acute rehabilitation following spinal cord injury: Currentexperience of the Model Systems. Arch Phys Med Rehabil 80:1397–1401, 1999.

5. Chu DA, Ahn JH, Ragnarsson KT, Helt J, Folcarelli P, Ramirez A:Deep venous thrombosis: Diagnosis in spinal cord injured pa-tients. Arch Phys Med Rehabil 66:365–368, 1985.

6. Consortium for Spinal Cord Medicine: Prevention of thromboem-bolism in spinal cord injury. Spinal Cord Med 20:259–283, 1997.

7. DeVivo MJ, Kartus PL, Stover SL, Rutt RD, Fine PR: Cause ofdeath for patients with spinal cord injuries. Arch Intern Med149:1761–1766, 1989.

8. El Masri WS, Silver JR: Prophylactic anticoagulant therapy inpatients with spinal cord injury. Paraplegia 19:334–342, 1981.

9. Frisbie JH, Sasahara AA: Low dose heparin prophylaxis for deepvenous thrombosis in acute spinal cord injury patients: A con-trolled study. Paraplegia 19:343–346, 1981.

10. Frisbie JH, Sharma GV: Pulmonary embolism manifesting asacute disturbances of behavior in patients with spinal cord in-jury. Paraplegia 32:570–572, 1994.

11. Geerts WH, Code KI, Jay RM, Chen E, Szalai JP: A prospectivestudy of venous thromboembolism after major trauma. N EnglJ Med 331:1601–1606, 1994.

12. Green D, Chen D, Chmiel JS, Olsen NK, Berkowitz M, Novick A,Alleva J, Steinberg D, Nussbaum S, Tolotta M: Prevention ofthromboembolism in spinal cord injury: Role of low molecularweight heparin. Arch Phys Med Rehabil 75:290–292, 1994.

13. Green D, Lee MY, Ito VY, Cohn T, Press J, Filbrandt PR,VandenBerg WC, Yarkony GM, Meyer PR Jr: Fixed- vs adjusted-dose heparin in the prophylaxis of thromboembolism in spinalcord injury. JAMA 260:1255–1258, 1988.

14. Green D, Lee MY, Lim AC, Chmiel JS, Vetter M, Pang T, Chen D,Fenton L, Yarkony GM, Meyer PR Jr: Prevention of thromboem-bolism after spinal cord injury using low-molecular-weight hep-arin. Ann Intern Med 113:571–574, 1990.

15. Greenfield LJ: Does cervical spinal cord injury induce a higherincidence of complications after prophylactic Greenfield filterusage? J Vasc Interv Radiol 8:719–720, 1997.

16. Gunduz S, Ogur E, Mohur H, Somuncu I, Acjksoz E, Ustunsoz B:Deep vein thrombosis in spinal cord injured patients. Paraplegia31:606–610, 1993.

17. Hachen HJ: Anticoagulant therapy in patients with spinal cordinjury. Paraplegia 12:176–187, 1974.

18. Harris S, Chen D, Green D: Enoxaparin for thromboembolismprophylaxis in spinal injury. Am J Phys Med Rehabil 75:326–327, 1996.

19. Jarrell BE, Posuniak E, Roberts J, Osterholm J, Cotler J,Ditunno J: A new method of management using the Kim-RayGreenfield filter for deep venous thrombosis and pulmonaryembolism in spinal cord injury. Surg Gynecol Obstet 157:316–320, 1983.

20. Khansarinia S, Dennis JW, Veldenz HC, Butcher JL, Hartland L:Prophylactic Greenfield filter placement in selected high-risktrauma patients. J Vasc Surg 22:231–235, 1995.

21. Kinney TB, Rose SC, Valji K, Oglevie SB, Roberts AC: Doescervical spinal cord injury induce a higher incidence of compli-cations after prophylactic Greenfield inferior vena cava filterusage? J Vasc Interv Radiol 7:907–915, 1996.

22. Kulkarni JR, Burt AA, Tromans AT, Constable PD: Prophylac-tic low dose heparin anticoagulant therapy in patients withspinal cord injuries: A retrospective study. Paraplegia 30:169–172, 1992.

23. Lamb GC, Tomski MA, Kaufman J, Maiman DJ: Is chronic spinalcord injury associated with increased risk of venous thrombo-embolism? J Am Paraplegia Soc 16:153–156, 1993.

24. McKinley WO, Jackson AB, Cardenas DD, DeVivo MJ: Long-term medical complications after traumatic spinal cord injury: Aregional model systems analysis. Arch Phys Med Rehabil 80:1402–1410, 1999.

25. Merli GJ, Crabbe S, Doyle L, Ditunno JF, Herbison GJ: Mechan-ical plus pharmacological prophylaxis for deep vein thrombosisin acute spinal cord injury. Paraplegia 30:558–562, 1992.

26. Merli GJ, Herbison GJ, Ditunno JF, Weitz HH, Henzes JH, ParkCH, Jaweed MM: Deep vein thrombosis: Prophylaxis in acutespinal cord injured patients. Arch Phys Med Rehabil 69:661–665, 1988.

27. Myllynen P, Kammonen M, Rokkanen P, Bostman O, Lalla M,Laasonen E: Deep venous thrombosis and pulmonary embolismin patients with acute spinal cord injury: A comparison withnonparalyzed patients immobilized due to spinal fractures.J Trauma 25:541–543, 1985.

28. Naso F: Pulmonary embolism in acute spinal cord injury. ArchPhys Med Rehabil 55:275–278, 1974.

28a. Perkash A: Experience with the management of deep vein throm-bosis in patients with spinal cord injury: Part II—A critical evalu-ation of the anticoagulant therapy. Paraplegia 18:2–14, 1980.

29. Perkash A, Prakash V, Perkash I: Experience with the manage-ment of thromboembolism in patients with spinal cord injury:Part I—Incidence, diagnosis and role of some risk factors. Para-plegia 16:322–331, 1978.



30. Powell M, Kirshblum S, O’Connor KC: Duplex ultrasoundscreening for deep vein thrombosis in spinal cord injured pa-tients at rehabilitation admission. Arch Phys Med Rehabil 80:1044–1046, 1999.

31. Quirke TE, Ritota PC, Swan KG: Inferior vena caval filter use inU.S. trauma centers: A practitioner survey. J Trauma 43:333–337,1997.

32. Roussi J, Bentolila S, Boudaoud L, Casadevall N, Vallee C,Carlier R, Lortat-Jacob S, Dizien O, Bussel B: Contribution ofD-dimer determination in the exclusion of deep venous throm-bosis in spinal cord injury patients. Spinal Cord 37:548–552,1999.

33. Tator CH, Duncan EG, Edmonds VE, Lapczak LI, Andrews DF: Com-parison of surgical and conservative management in 208 patientswith acute spinal cord injury. Can J Neurol Sci 14:60–69, 1987.

34. Todd JW, Frisbie JH, Rossier AB, Adams DF, Als AV, ArmeniaRJ, Sasahara AA, Tow DE: Deep venous thrombosis in acutespinal cord injury: A comparison of 125I fibrinogen leg scanning,impedance plethysmography and venography. Paraplegia 14:50–57, 1976.

35. Tola JC, Holtzman R, Lottenberg L: Bedside placement of infe-rior vena cava filters in the intensive care unit. Am Surg 65:833–837, 1999.

35a. Tomaio A, Kirshblum SC, O’Connor KC, Johnston M: Treatmentof acute deep vein thrombosis in spinal cord injured patientswith enoxaparin: A cost analysis. J Spinal Cord Med 21:205–210,1998.

36. Waring WP, Karunas RS: Acute spinal cord injuries and theincidence of clinically occurring thromboembolic disease. Para-plegia 29:8–16, 1991.

37. Watson N: Venous thrombosis and pulmonary embolism inspinal cord injury. Paraplegia 6:113–121, 1968.

38. Watson N: Anti-coagulant therapy in the prevention of venousthrombosis and pulmonary embolism in the spinal cord injury.Paraplegia 16:265–269, 1978.

38a. Weingarden SI, Weingarden DS, Belen J: Fever and thromboembolicdisease in acute spinal cord injury. Paraplegia 26:35–42, 1988.

39. Wilson JT, Rogers FB, Wald SL, Shackford SR, Ricci MA: Prophy-lactic vena cava filter insertion in patients with traumatic spinalcord injury: Preliminary results. Neurosurgery 35:234–239, 1994.

40. Winemiller MH, Stolp-Smith KA, Silverstein MD, Therneau TM:Prevention of venous thromboembolism in patients with spinalcord injury: Effects of sequential pneumatic compression andheparin. J Spinal Cord Med 22:182–191, 1999.

41. Yelnik A, Dizien O, Bussel B, Schouman-Claeys E, Frija G, Pan-nier S, Held JP: Systematic lower limb phlebography in acutespinal cord injury in 147 patients. Paraplegia 29:253–260, 1991.

Drawings by Leonardo da Vinci, revealing how thecervical musculature stabilizes the cervical spinal col-umn. Leonardo’s investigations were some of the ear-liest into the relationship of the spinal anatomy andits mechanics. Courtesy, Dr. Edwin Todd, Pasadena,California.


CHAPTER 11

Nutritional Support after Spinal Cord Injury

RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS: Nutritional support of patients with spinal cord injuries is recommended. Energy expenditure is best

determined by indirect calorimetry in these patients because equation estimates of energy expenditure andsubsequent caloric need tend to be inaccurate.

RATIONALE

Hypermetabolism, an accelerated catabolic rate, andrampant nitrogen losses are consistent sequelae tomajor trauma, particularly acute traumatic brain in-

jury and acute spinal cord injury (ASCI) (7, 9–11, 13, 18, 19,22). A well-documented hypermetabolic, catabolic injury cas-cade is initiated immediately after central nervous systeminjury that results in depletion of whole-body energy stores,loss of lean muscle mass, reduced protein synthesis, and,ultimately, loss of gastrointestinal mucosal integrity and com-promise of immune competence (6, 9, 10, 12, 18, 19, 22).Severely brain-injured and spinal cord injury (SCI) patients,therefore, are at risk for prolonged nitrogen losses and ad-vanced malnutrition within 2 to 3 weeks after injury withresultant increased susceptibility for infection, impairedwound healing, and difficulty weaning from mechanical ven-tilation (5, 10, 13, 18, 19, 22). These factors, added to theinherent immobility, denervation, and muscle atrophy asso-ciated with SCI, provide the rationale for nutritional supportof SCI patients after trauma.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The following medical subject headings wereused in combination with “spinal cord injury”: “nutrition”and “nutritional support.” Approximately 105 citations wereacquired. Non-English language citations were deleted, andtitles and abstracts of the remaining publications were re-viewed. We focused on the specific issue of nutrition andhuman patients with ASCI and identified 18 articles. Relevantmanuscripts and reviews describing nutritional support ofhead-injured patients and several reports describing the nu-tritional status of chronic SCI patients are included among the23 citations in the reference list. These efforts identified oneClass II study and four Class III studies that describe metab-olism, nitrogen wasting, and the effect of feeding on nitrogenbalance and serum biochemistries in patients after ASCI.

These articles are summarized in Table 11.1. There were nostudies that examined the effects of nutritional support onoutcome after ASCI.


Hypermetabolism, catabolism, and accelerated nitrogenlosses are well-recognized complications that follow trau-matic injury (10, 11, 19, 22). They have been identified andstudied in human patients who have sustained traumaticbrain injuries and SCI. A number of publications have de-scribed the increased energy requirements and nitrogen lossesof patients after acute head injury (7, 8, 10–12, 17, 18, 22, 23).Fewer studies have focused on hypermetabolism, catabolism,and nitrogen losses after ASCI (6, 13, 14, 20, 22). Althoughthere are metabolic similarities between isolated traumaticbrain injury and severe isolated SCI, it seems that there maybe important biological differences between the two centralnervous system injury types that have a bearing on supple-mental nutritional therapy (6, 13, 14, 18–20, 22).

Severe head injury is associated with a resting energy ex-penditure (REE) of approximately 140% of predicted normalbasal energy expenditure (BEE) (8, 10, 11, 17, 18, 22, 23).Indirect calorimetry is the most widely used reliable means todetermine individual energy requirements in hospitalized pa-tients after traumatic injury (10, 18, 19, 22). It requires the useof a portable metabolic cart and uses a technique that mea-sures respiratory gas exchange and the rate of oxygen use ina given patient. It provides an estimate of energy expenditureby the patient by determining the known caloric yield from 1L of oxygen based on differences in oxygen consumption andcarbon dioxide production. It is performed at the bedside inthe intensive care unit in severely injured patients. Metabolicexpenditure is expressed as a percentage of normal BEE at rest(predicted). Indirect calorimetry is typically performed oncedaily for the first several days after injury and periodicallythereafter (10, 18, 19, 22). The Harris-Benedict equation, withactivity and stress of injury variables, has been shown topredict energy expenditure after a traumatic brain injury with


reasonable accuracy without indirect calorimetry (10, 14, 19,20, 22).

Nutritional support of head-injured patients is typicallybegun within days of admission and is guided by the meta-bolic information provided by indirect calorimetry and bypredicted energy expenditure values derived by equation.Hypermetabolism, accelerated catabolism, and excessive ni-trogen losses continue for at least 2 weeks after injury (8–11,18, 19, 22). The exact duration of this response to injury isunknown, may vary among similar patients, and can be af-fected by other traumatic injuries, pancreatitis, infection, orsepsis (2, 10, 18, 19, 22). Nutritional support in this setting isdesigned to provide nitrogen-rich, high-energy supplementalfuel to blunt excess catabolism and preserve energy stores,muscle mass, gastrointestinal integrity, and immune compe-tence (6, 10, 18, 19, 22). Nitrogen balance is difficult—oftenimpossible—to achieve, particularly within the first week ofinjury (7, 11, 13, 14, 20). Matching nutritional replacementwith caloric needs, therefore, has become the primary goal ofnutritional therapy.

The extent of neuronal connectivity and the neurogenicstimuli (muscle tone) to the musculoskeletal system seemsimportant to the level of metabolic expenditure after injury ofthe central nervous system (1, 3, 4, 13–16, 18–21, 22). Agitated,combative head-injured patients, for example, can have REElevels as high as 200% of expected BEE levels (10, 11, 18, 22).Conversely, pharmacological paralysis of head-injured pa-tients has been associated with reductions in resting energyexpenditure by 20 to 30% (10, 11, 18, 22). Patients who havesustained isolated ASCIs often have increased metabolic ex-penditure compared with normative energy expenditure lev-els (13, 14, 18–20, 22). However, because of the paralysis and

flaccidity associated with ASCI, measured REE values in thesepatients are considerably lower than those predicted by theHarris-Benedict equation based on age, sex, body surfacearea, activity, and injury severity (14, 19, 20, 23). Patients withthe greatest neurological deficits and the least muscle toneafter SCI (high-cervical level quadriplegic patients) havelower measured REE values than those found in patients withincomplete spinal injuries or lower spinal cord injuries (tho-racic level paraplegic patients) (13, 14, 19, 20, 22). Kaufman etal. (13), in 1985, described their experience with eight ASCIpatients managed at the University of Texas. They notedaccelerated nitrogen losses and ongoing negative nitrogenbalance greater than expected. Differences in initial andfollow-up nutritional assessments revealed deterioration innutritional status during the 2-week period of observation,partly caused by inadequate supply of protein and calories.Infective complications and prolonged respiratory supportwere common. The authors concluded that muscle atrophymight play an important role in the accelerated nitrogenlosses they identified in patients with paralysis due to com-plete SCI and that improved nutritional support might reducemedical complications after ASCI.

Young et al. (22) reported four quadriplegic ASCI patientsassessed with indirect calorimetry. They found that indirectcalorimetry provided more accurate REE values for their pa-tients than Harris-Benedict equation estimates, even if theequation estimates did not incorporate injury and activityfactors. The authors also noted marked daily nitrogen lossesand negative nitrogen balance. They concluded that equationestimates of REE values of SCI patients overestimate meta-bolic expenditure, and they emphasized the importance of

TABLE 11.1. Summary of Reports on Nutritional Support after Spinal Cord Injurya


ClassConclusions

Cruse et al., 2000 (5) Comparison of nutritional,immune, endocrine status in 15ASCI patients versus 16 matchedcontrols.

II SCI patients have hormonal changes, poornutritional status, and decreased immunefunction compared with controls.

Rodriguez et al., 1997 (20) Prospective assessment andtreatment of 12 ASCI patients.

III REE less than predicted, marked“obligatory” nitrogen losses due toflaccidity and atrophy of denervatedmuscle after SCI.

Kearns et al., 1992 (14) Prospective assessment of 10 ASCIpatients over 4-wk period ofobservation.

III Exaggerated nitrogen and 3-methylhistidineexcretion, marked weight loss. Lower REEthan predicted after SCI.

Young et al., 1987 (22) 4 ASCI patients assessed viaindirect calorimetry.

III Indirect calorimetry best means todetermine energy expenditure after ASCI.

Kaufman et al., 1985 (13) Assessment of nutritional status of8 SCI patients over 2-wk period ofobservation.

III Deterioration in nutritional status despiteattempted treatment. Marked nitrogenlosses. Increased infectious and respiratorycomplications.

a ASCI, acute spinal cord injury; SCI, spinal cord injury; REE, resting energy expenditure.



indirect calorimetry in predicting energy expenditure afterASCI.

Kearns et al. (14) prospectively assessed and provided nu-tritional support to 10 ASCI patients they managed and mon-itored for 4 weeks. Their 1992 report documents the use ofindirect calorimetry to determine REE and provide matchedcaloric supplementation. All patients had isolated SCIs with-out associated head injury or other organ system trauma.Initial measured REE values were 10% below predicted BEElevels. All patients experienced exaggerated nitrogen and3-methylhistidine losses, indicating excessive lean body massand muscle loss. A 10% decrease in body weight accompaniedthese losses despite caloric replacement matched to or exceed-ing measured REE values for each patient. The specifics ofnutrition administration (mix and route of delivery) were notpresented. The authors noted an increase in REE over time inpart attributable to reductions in body weight and in part dueto return of muscle tone. The authors concluded that acuteisolated SCI is associated with lower REE values comparedwith predicted values. ASCI patients have exaggerated nitro-gen and 3-methylhistidine losses owing to atrophy of dener-vated muscle. The authors attributed the reduced metabolicactivity seen in these patients to the flaccidity of denervatedmusculature after severe SCI and noted that as muscle lossand weight reductions progress, REE increases, particularly ifrecovery of motor function and/or return of muscle toneoccurs.

Rodriguez et al. (20) studied the metabolic response to SCIin 12 acute trauma patients. Assessment and nutritional sup-port were instituted immediately after injury and continuedfor 4 weeks after injury. Harris-Benedict estimations of energyexpenditure were compared with values obtained from indi-rect calorimetry in each patient. All patients had acceleratednitrogen losses and negative nitrogen balance. Eleven of the12 patients had negative nitrogen balance for the entire 4weeks of therapy despite matched caloric replacement. Thesingle patient in whom nitrogen balance was realized had anincomplete SCI. The Harris-Benedict equation with activityfactor of 1.2 and a stress/injury factor of 1.6 consistentlyoverestimated energy expenditure in these 12 patients andwould have resulted in excessive feeding. The authors con-cluded that large nitrogen losses after severe SCI are “oblig-atory” as a result of atrophy and wasting of denervatedmusculature below the level of injury. Patients with completetraumatic myelopathy had greater obligatory nitrogen lossesthan patients with incomplete SCI. They recommended thatindirect calorimetry be used as the energy expenditure assess-ment method after SCI, particularly in the early postinjuryperiod. If the Harris-Benedict equation is used in these pa-tients in this setting, they recommend that the activity factorshould be eliminated and the stress/injury factor of the equa-tion should be reduced.

Cruse et al. (5) examined the neurological, immune, endo-crine, and nutritional status of 15 male SCI patients and com-pared them with 16 healthy age-matched control subjects. Thetiming of assessment in relation to SCI for each patient was notspecified. Their report described decreased natural and adaptiveimmune responses in the SCI patient population beginning

within 2 weeks of injury that reached a nadir 3 months afterinjury. They noted increased adrenocorticotropic hormone andplasma cortisol levels, decreased zinc, albumin, and prealbuminserum levels, surface marker changes in both lymphocytes andgranulocytes, and decreased adhesion molecule-binding abilityafter SCI compared with findings in healthy control patients.They concluded that patients with severe ASCI have decreasedimmune function, impaired nutritional status, and a decreasednumber of adhesion molecules, all of which occurs within weeksafter acute injury. The authors note that these hormonal alter-ations, nutritional deficiencies, and changes in immune functionmay increase susceptibility to infection and may contribute todelayed wound healing.

The change in energy expenditure identified in patientsafter ASCI seems to persist long after the initial injury andrecovery phase (1, 3, 4, 15, 16, 19, 21). Several investigatorshave noted long-standing reductions in REE in SCI patients,reductions that correlate with the degree of neurological in-jury and the extent of lean body mass loss after paralysis (1, 3,4, 15, 16, 19, 21). Cox et al. (4) measured energy expenditure instable non-ASCI patients in the rehabilitation setting. Theyreported that quadriplegic patients required 22.7 kcal/kg/dcompared with 27.9 kcal/kg/d for paraplegic patients theystudied. Most investigators conclude that equation methodsto estimate energy expenditure in SCI patients are inaccurate,in both the acute and chronic settings (15, 16, 20–22).

There has been no report assessing the efficacy of the route offeeding (parenteral or enteral) for SCI patients in the acute set-ting. The literature on nutritional support for head injury pa-tients supports using the enteral route for nutritional supple-mentation if the gut is functional (8, 10, 11, 18–20, 22). Thisgeneral policy seems to have been followed by investigators ofnutritional support for ASCI patients (13, 14, 20, 22). The poten-tial benefits of enteral feeding over parenteral delivery includemaintenance of gut integrity and function, reduced expense,lower risk of infection, and avoidance of complications related tothe intravenous catheter (8, 10, 11, 18, 19, 22). Nasoduodenal ornasojejunal feeding tubes usually allow full caloric, high-nitrogen, high-volume feeding within days of injury. In patientswith bowel injury, mechanical bowel obstruction, or prolongedileus, it is recommended that parenteral nutrition be initiateduntil the bowel recovers and conversion to enteral nutrition canbe accomplished (8, 10, 18, 19, 22).

There has been no report assessing the mix or compositionof nutritional supplementation for SCI patients. The literatureon nutritional support for head injury patients suggests be-ginning with a high-nitrogen enteral or parenteral solutioncontaining at least 15% of calories as protein, no more than15% glucose/dextrose, a minimum of 4% of total energyneeds as essential fatty acids, and the addition of vitamins,essential elements, and trace minerals (10, 12, 17, 18, 19, 22,23). No study has been published that has examined the effectof nutritional support on neurological outcome after ASCI.

SUMMARY

Alterations in metabolism occur after ASCI, but the markedhypermetabolic response seen after acute traumatic brain in-

Nutritional Support after Spinal Cord Injury S83


jury seems to be blunted in SCI patients by the flaccidity ofdenervated musculature after spinal cord transection or in-jury. As a result, REE is lower than predicted after ASCI.Equation estimates of REE in these patients have proven to beinaccurate; therefore, indirect calorimetry is the recom-mended technique to assess energy expenditure in both theacute and chronic settings. Protein catabolism does occur afteracute, severe SCI, and marked losses in lean body mass due tomuscle atrophy result in huge nitrogen losses, prolongednegative nitrogen balance, and rapid weight loss. Nutritionalsupport of the SCI patient to meet caloric and nitrogen needs,not to achieve nitrogen balance, is safe and may reduce thedeleterious effects of the catabolic, nitrogen-wasting processthat occurs after acute SCI.

KEY ISSUES FOR FUTURE INVESTIGATION:

An assessment of the timing, route of administration, andcomposition of nutritional therapy on outcome, both neuro-logical and medical, should be performed. This could be bestaccomplished with a multicenter case-control study.


REFERENCES

1. Barboriak JJ, Rooney CB, El Ghatit AZ, Spuda K, Anderson AJ:Nutrition in spinal cord injury patients. J Am Paraplegia Soc6:32–36, 1983.

2. Carey ME, Nance FC, Kirgis HD, Young HF, Megison LC Jr, KlineDG: Pancreatitis following spinal cord injury. J Neurosurg 47:917–922, 1997.

3. Clarke KS: Caloric costs of activity in paraplegic persons. ArchPhys Med Rehabil 47:427–435, 1966.

4. Cox SA, Weiss SM, Posuniak EA, Worthington P, Prioleau M,Heffley G: Energy expenditure after spinal cord injury: An eval-uation of stable rehabilitating patients. J Trauma 25:419–423, 1985.

5. Cruse JM, Lewis RE, Dilioglou S, Roe DL, Wallace WF, Chen RS:Review of immune function, healing of pressure ulcers, and nu-tritional status in patients with spinal cord injury. J Spinal CordMed 23:129–135, 2000.

6. Cruse JM, Lewis RE, Roe DL, Dilioglou S, Blaine MC, Wallace WF,Chen RS: Facilitation of immune function, healing of pressureulcers, and nutritional status in spinal cord injury patients. ExpMol Pathol 68:38–54, 2000.

7. Frankenfield DC, Smith JS, Cooney RN: Accelerated nitrogen lossafter traumatic injury is not attenuated by achievement of energybalance. JPEN J Parenter Enteral Nutr 21:324–329, 1997.

8. Grahm TW, Zadrozny DB, Harrington T: The benefits of earlyjejunal hyperalimentation in the head injured patient. Neuro-surgery 25:729–735, 1989.

9. Hadley MN: Hypermetabolism after CNS trauma: Arresting the“injury cascade.” Nutrition 5:143, 1989.

10. Hadley MN: Hypermetabolism following head trauma: Nutri-tional considerations, in Barrow DL (ed): Complications and Se-quelae of Head Injury (Neurosurgical Topics series). Park Ridge,AANS, 1992, pp 161–168.

11. Hadley MN, Grahm TW, Harrington T, Schiller WR, McDermottMK, Posillico DB: Nutritional support and neurotrauma: A criti-cal review of early nutrition in forty-five acute head injury pa-tients. Neurosurgery 19:367–373, 1986.

12. Hulsewe KW, van Acker BA, von Meyenfeldt MF, Soeters PB:Nutritional depletion and dietary manipulation: Effects on theimmune response. World J Surg 23:536–544, 1999.

13. Kaufman HH, Rowlands BJ, Stein DK, Kopaniky DR, GildenbergPL: General metabolism in patients with acute paraplegia andquadriplegia. Neurosurgery 16:309–313, 1985.

14. Kearns PJ, Thompson JD, Werner PC, Pipp TL, Wilmot CB: Nu-tritional and metabolic response to acute spinal cord injury. JPENJ Parenter Enteral Nutr 16:11–15, 1992.

15. Levine AM, Nash MS, Green BA, Shea JD, Aronica MJ: An exam-ination of dietary intakes and nutritional status of chronic healthyspinal cord injured individuals. Paraplegia 30:880–889, 1992.

16. Peiffer SC, Blust P, Leyson JF: Nutritional assessment of the spinalcord injured patient. J Am Diet Assoc 78:501–505, 1981.

17. Rapp RP, Young B, Twyman D, Bivins BA, Haack D, Tibbs PA,Bean JR: The favorable effect of early parenteral feeding on sur-vival in head injured patients. J Neurosurg 58:906–912, 1983.

18. Riley KO, May AK, Hadley MN: Neurological injury and nutri-tional support, in Batjer HH, Loftus CM (eds): Textbook of Neuro-logical Surgery: Principles and Practice. Philadelphia, LippincottWilliams & Wilkins (in press).

19. Rodriguez DJ, Benzel EC: Nutritional support, in Benzel EC (ed):Spine Surgery: Techniques, Complication Avoidance, and Management.New York, Churchill Livingstone, 1999, vol 2, pp 1321–1331.

20. Rodriguez DJ, Benzel EC, Clevenger FW: The metabolic responseto spinal cord injury. Spinal Cord 35:599–604, 1997.

21. Sedlock DA, Laventure SJ: Body composition and resting energyexpenditure in long term spinal cord injury. Paraplegia 28:448–454, 1990.

22. Young B, Ott L, Rapp RP, Norton J: The patient with criticalneurological disease. Crit Care Clin 3:217–233, 1987.

23. Young B, Ott L, Twyman D, Norton J, Rapp R, Tibbs P, Haack D,Brivins B, Dempsey R: The effect of nutritional support on out-come from severe head injury. J Neurosurg 67:668–676, 1987.


CHAPTER 12

Management of Pediatric Cervical Spine and SpinalCord Injuries

RECOMMENDATIONSDIAGNOSTIC:Standards: There is insufficient evidence to support diagnostic standards.Guidelines:• In children who have experienced trauma and are alert, conversant, have no neurological deficit, no midline

cervical tenderness, and no painful distracting injury, and are not intoxicated, cervical spine x-rays are notnecessary to exclude cervical spine injury and are not recommended.

• In children who have experienced trauma and who are either not alert, nonconversant, or have neurological deficit,midline cervical tenderness, or painful distracting injury, or are intoxicated, it is recommended that anteroposteriorand lateral cervical spine x-rays be obtained.

Options:• In children younger than age 9 years who have experienced trauma, and who are nonconversant or have an

altered mental status, a neurological deficit, neck pain, or a painful distracting injury, are intoxicated, or haveunexplained hypotension, it is recommended that anteroposterior and lateral cervical spine x-rays be obtained.

• In children age 9 years or older who have experienced trauma, and who are nonconversant or have an altered mentalstatus, a neurological deficit, neck pain, or a painful distracting injury, are intoxicated, or have unexplained hypoten-sion, it is recommended that anteroposterior, lateral, and open-mouth cervical spine x-rays be obtained.

• Computed tomographic scanning with attention to the suspected level of neurological injury to excludeoccult fractures or to evaluate regions not seen adequately on plain x-rays is recommended.

• Flexion/extension cervical x-rays or fluoroscopy may be considered to exclude gross ligamentous instabilitywhen there remains a suspicion of cervical spine instability after static x-rays are obtained.

• Magnetic resonance imaging of the cervical spine may be considered to exclude cord or nerve rootcompression, evaluate ligamentous integrity, or provide information regarding neurological prognosis.

TREATMENT:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options:• Thoracic elevation or an occipital recess to prevent flexion of the head and neck when restrained supine on an

otherwise flat backboard may allow for better neutral alignment and immobilization of the cervical spine in childrenyounger than 8 years because of the relatively large head in these younger children and is recommended.

• Closed reduction and halo immobilization for injuries of the C2 synchondrosis between the body andodontoid is recommended in children younger than 7 years.

• Consideration of primary operative therapy is recommended for isolated ligamentous injuries of the cervicalspine with associated deformity.

RATIONALE

There are distinct, unique aspects of the management ofchildren with potential injuries of the cervical spinalcolumn and cervical spinal cord compared with adult

patients that warrant specific recommendations. The methodsof preadmission immobilization necessary to approximate

“neutral” cervical spinal alignment in a young child differfrom those methods commonly used for adults. The spinalinjury patterns among young children differ from those thatoccur in adults. The diagnostic studies and images necessaryto exclude a cervical spine injury in a child may be differentthan in the adult as well. The interpretation of pediatric x-raystudies must be made with knowledge of age-related devel-


opment of the osseous and ligamentous anatomy. Methods ofreduction, stabilization, and subsequent treatment, surgicaland nonsurgical, must be customized to each child, takinginto account the child’s degree of physical maturation and hisor her specific injury. The purpose of this review is to addressthe unique aspects of children with real or potential cervicalspinal injuries and provide recommendations regarding theirmanagement.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject headings “spinal cord inju-ries” and “child” yielded 1022 citations. These citations werereviewed in combination with “cervical vertebra,” “spinalinjuries,” and “child,” which yielded 152 citations. Non-English language citations were deleted. The remaining ab-stracts were reviewed for those that described children whohad sustained or were being evaluated for a cervical spinalcord or cervical spinal column injury. Articles describing theclinical aspects and management of children were used togenerate these guidelines. Case reports were excluded. Of the58 articles meeting selection criteria, none were Class I stud-ies. One Class II study addressed diagnostic imaging in chil-dren. All remaining articles were case series representingClass III evidence. Summaries of these 58 articles are providedin Tables 12.1 and 12.2. In addition, articles germane to thetopic but not meeting criteria for inclusion in the EvidentiaryTables, such as general review articles, are referenced in theScientific Foundation section and are included in thereferences.


Preadmission immobilization

The primary goal of preadmission management of pediatricpatients with potential cervical spine or spinal cord injury(SCI) is to prevent further injury. Along with ensuring anadequate airway, ventilation, and perfusion, spinal immobi-lization likely plays an important role in preventing furtherinjury to the vertebral column and spinal cord. Immobiliza-tion of the child’s cervical spine in the neutral position isdesired. To achieve neutral alignment of the cervical spine inchildren younger than 8 years, allowances must be made forthe relatively large head compared with the torso, whichforces the neck into a position of flexion when the head andtorso are supine on a flat surface (39). Nypaver and Treloar(39) prospectively evaluated 40 children younger than 8 yearsseen in an emergency room for reasons other then head andneck trauma and assessed them with respect to neutral posi-tioning on a backboard. They found that all 40 children re-quired elevation of the torso to eliminate positional neckflexion and achieve neutral alignment as determined by twoindependent observers. The mean amount of elevation re-quired was 25 mm. Children younger than 4 years requiredmore elevation than those 4 years or older (P � 0.05). Becauseof these findings, it was recommended that, when immobiliz-

ing children younger than 8 years, either the torso be elevatedor an occipital recess be created to achieve a more neutralposition for immobilization of the cervical spine. In a separatereport, Treloar and Nypaver (60) similarly found that semi-rigid cervical collars placed on children younger than 8 yearsdid not prevent this positional forced flexion when placedsupine on standard, rigid spinal boards.

Herzenberg et al. (26) studied 10 children younger than 7years with cervical spine injuries who were positioned on abackboard. All had anterior angulations or translation at theinjured segment, which was reduced by allowing neck exten-sion into a more neutral position. They suggested that align-ment of the patient’s external auditory meatus with his/hershoulders would help to achieve neutral cervical spine posi-tioning. Curran et al. (8), however, found no correlation be-tween age and degree of cervical kyphosis identified in chil-dren transported on backboards. They did note, however, that30% of children had more than 10 degrees of kyphosis asdetermined by Cobb angle measurements between C2 and C6.No specific technique or device allowed superior neutral po-sitioning of the cervical spine in patients they studied. Noneof their patients were immobilized on boards with an occipitalrecess or thoracic padding. Huerta et al. (27) evaluated avariety of immobilization devices on children, infants, andchild-sized mannequins. They concluded that no collar pro-vided “acceptable immobilization” when used alone. Theyfound that the combination of a modified half-spine board,rigid cervical collar, and tape was the most effective means ofimmobilizing the cervical spine for transport in children.Schafermeyer et al. (51), however, cautioned that immobiliza-tion techniques that use taping across the torso to secure thechild to the spine board may have deleterious effects onrespiratory function. They studied 51 healthy children 6 to 15years of age by measuring forced vital capacity (FVC). FVCdecreased when the child moved from the upright to thesupine position. Taping across the torso to secure the volun-teer to the spine board caused further reductions in FVC of 41to 96% (mean, 80%), compared with the supine FVC withouttape. The authors cautioned that this restriction of FVC mightbe enough to create respiratory insufficiency in some traumapatients.

In summary, when spinal immobilization is indicated forchildren for transportation, the type of immobilization shouldtake into account the child’s age and physical maturity. Itshould allow for the relatively larger head with respect to thetorso in younger children. Although ideal spinal immobiliza-tion of pediatric patients who have sustained trauma seems tobe provided by a combination of a spinal board, rigid collar,and tape, these immobilization techniques may negativelyinfluence the child’s respiratory function.

Imaging

After immobilization and transport to an acute care facility,initial clinical evaluation and medical/hemodynamic supportand the need for and type of radiographic assessment must bedecided and undertaken. Several authors have evaluated theindications for radiographic assessment of children with a



potential cervical spinal injury (4, 32). Laham et al. (32) inves-tigated the role of cervical spine x-ray evaluation of 268 chil-dren with apparent isolated head injuries. They retrospec-tively divided the children into high-risk (n � 133) and low-risk (n � 135) groups. High-risk children were those who

were incapable of verbal communication, either because ofage (younger than 2 yr) or head injury, and those who hadneck pain. They used the “three-view approach” of antero-posterior, lateral, and open-mouth x-rays. They discovered nocervical spine injuries in the low-risk group but 10 injuries

TABLE 12.1. Summary of Reports on Diagnosis of Pediatric Cervical Spinal Injuriesa


ClassConclusions

Viccellio et al., 2001 (62) Prospective multicenter evaluation of cervicalspine x-rays obtained in 3065 childrenincurring trauma. Low-risk criteria of absenceof: neck tenderness, painful distracting injury,altered alertness, neurological deficit, orintoxication.

II No child fulfilling all 5 low-risk criteria had acervical spine injury. X-rays may not be necessary toclear the cervical spine in children fulfilling all 5criteria.

Ralston et al., 2001 (45) Blinded review of 129 children with bluntcervical trauma who had flexion and extensionx-rays.

III Flexion and extension views with normal cervicalspine x-rays or with only loss of cervical lordosis didnot unmask any new abnormalities.

Buhs et al., 2000 (6) Multi-institutional review of pediatric cervicalspine injuries and the x-rays needed to achievea diagnosis.

III Lateral cervical x-ray was diagnostic in 13/15children aged �9 yr. In no child �9 yr old was theopen-mouth view the diagnostic study. In only 1/36children �9 yr old was the open-mouth view thediagnostic study.

Dwek and Chung, 2000(11a)

Retrospective review of 247 children with ahistory of trauma who had flexion andextension cervical spine x-rays.

III All children (91%) with normal static cervical x-rayshad normal flexion/extension x-rays.

Swischuk et al., 2000 (59) Survey of pediatric radiologists regarding use ofopen-mouth view of the odontoid.

III �50% response. Approximately 40% of respondentsdid not use open-mouth views in children.

Scarrow et al., 1999 (50) Performed flexion/extension cervicalfluoroscopy with SSEP monitoring in 15comatose pediatric patients.

III None had radiographic abnormalities. 3 childrenhad changes in the SSEPs; 1 of these 3 children wasstudied with MRI, and findings were normal.

Shaw et al., 1999 (56) Retrospective review of the cervical x-rays in138 trauma patients �16 yr old.

III 22% incidence of pseudosubluxation of C2 on C3.Median age of pseudosubluxation group was 6.5 yrversus 9 yr for those without pseudosubluxation.

Berne et al., 1999 (3) 58 patients with severe blunt trauma underwenthelical CT of entire cervical spine.

III 20 had cervical spine injuries. Plain x-rays missed 8injuries. CT missed 2 injuries.

Keiper et al., 1998 (29) Retrospective review of evaluating 52 childrenby MRI with suspected cervical spine trauma orinstability without fracture.

III There were 16 abnormal studies. The most commonabnormality was posterior ligamentous injury. 4children underwent surgical stabilization. The MRIfindings caused the surgeon to extend the length ofstabilization in all 4 cases.

Davis et al., 1993 (9) Retrospective review of 15 children with spinalcord injury underwent MRI 12 h to 2 mo afterinjury. 7 with SCIWORA.

III MRI findings correlated with prognosis. Hemorrhagiccord contusions and cord “infarction” wereassociated with permanent deficits. No compressivelesions in SCIWORA cases. Normal MRI findingswere associated with no myelopathy.

Schleehauf et al., 1989 (51) 104 “high-risk” patients underwent CT asscreening tool for cervical spine injury.

III Sensitivity overall was 0.78. Sensitivity was 1.0 forunstable injuries not able to be seen by plain x-rays.2 upper cervical subluxations without fracture weremissed.

Kowalski et al., 1987 (31) 8 patients with occipitalatlantoaxial problemsand 6 normal subjects were studied with CT.

III CT findings looked similar for those with C1–C2rotary subluxation to normal subjects with theirheads maximally turned. CT with the head turned tothe contralateral side differentiated rotarysubluxation from normals and spasmodic torticollis.

Cattell and Filtzer, 1965 (7) Lateral upright cervical x-rays in neutral,flexion, and extension in 160 randomlyselected children aged 1–16 yr.

III C2–C3 subluxation was moderate to marked in 24%,predominantly in children �8 yr of age. Theatlantodens interval was �3 mm during flexion in20% of children �8 yr of age.

a SSEP, somatosensory evoked potential; CT, computed tomography; MRI, magnetic resonance imaging; SCIWORA, spinal cord injury withoutradiographic abnormality.

Management of Pediatric Cervical Spinal Injuries S87


TAB

LE12

.2.

Sum

mar

yof

Rep

orts

onTr

eatm

ent

ofPe

diat

ric

Cer

vica

lSp

inal

Inju

ries

a

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ce

Cla

ssC

oncl

usio

ns

Eler

aky

etal

.,20

00(1

2)R

etro

spec

tive

revi

ewof

102

child

ren

with

cerv

ical

spin

alin

juri

es.

III30

child

ren

(30%

)w

ere

trea

ted

surg

ical

ly.

Ode

ntet

al.,

1999

(40)

Rev

iew

of15

youn

gch

ildre

nw

ithod

onto

idin

juri

es.

III6

with

neur

olog

ical

defic

itsha

dce

rvic

otho

raci

cco

rdin

juri

es.

Exte

rnal

imm

obili

zatio

nw

asa

succ

essf

ulpr

imar

yth

erap

y.3

child

ren

who

unde

rwen

top

erat

ions

asth

eir

prim

ary

ther

apy

expe

rien

ced

com

plic

atio

ns.

Schw

arz,

1998

(52)

Are

view

of4

child

ren

pres

entin

g�

3m

oaf

ter

the

onse

tof

C1–

C2

rota

ry

subl

uxat

ion.

III2

child

ren

had

irre

duci

ble

subl

uxat

ions

.1

child

had

recu

rren

tsu

blux

atio

nin

aM

iner

vaca

st.

1

child

was

succ

essf

ully

trea

ted

with

clos

edre

duct

ion

and

8w

kin

aM

iner

vaca

st.

Suba

chet

al.,

1998

(58)

Are

view

of20

child

ren

with

C1–

C2

rota

rysu

blux

atio

n.III

4re

duce

dsp

onta

neou

sly.

15/1

6tr

eate

dw

ithtr

actio

nre

duce

din

am

ean

of4

d.6

child

ren

requ

ired

fusi

onbe

caus

eof

recu

rren

tsu

blux

atio

nor

irre

duci

ble

subl

uxat

ion.

No

child

expe

rien

ced

recu

rren

tsu

blux

atio

nif

redu

ced

with

in21

dof

sym

ptom

onse

t.

Finc

han

dB

arne

s,19

98(1

6)R

etro

spec

tive

revi

ewof

32ch

ildre

nw

ithm

ajor

cerv

ical

spin

ein

juri

es.

III8

child

ren

(25%

)w

ere

trea

ted

surg

ical

ly.

All

achi

eved

unio

nor

radi

olog

ical

stab

ility

.N

o

neur

olog

ical

dete

rior

atio

nfr

omsu

rger

yor

clos

edre

duct

ion.

Ope

rate

don

ligam

ento

usin

juri

es.

Rei

nges

etal

.,19

98(4

7)R

epor

tof

prim

ary

C1–

C2

fusi

onin

ayo

ung

child

with

anod

onto

id

inju

ryan

dlo

wer

cerv

ical

cord

inju

ry.

IIIN

one

urol

ogic

alim

prov

emen

t.Su

cces

sful

fusi

on.

Trel

oar

and

Nyp

aver

,19

97(6

0)M

easu

rem

ent

ofce

rvic

alsp

ine

flexi

onin

child

ren

with

sem

i-ri

gid

colla

rs

onsp

inal

boar

ds.

IIISe

mi-

rigi

dco

llars

did

not

prev

ent

the

cerv

ical

spin

efr

ombe

ing

forc

edin

tofle

xion

inch

ildre

n�

8

yrol

dw

hen

ona

spin

albo

ard.

Lui

etal

.,19

96(3

3)R

etro

spec

tive

revi

ewof

C1–

C2

inju

ries

in22

child

ren.

12ch

ildre

nha

d

odon

toid

inju

ries

.9

child

ren

had

ligam

ento

usin

juri

es(A

AD

s)on

ly.

IIIFl

exio

n/ex

tens

ion

x-ra

ysw

ere

need

edto

diag

nose

4od

onto

idin

juri

esan

d6

AA

Ds.

9/12

odon

toid

inju

ries

redu

ced

easi

ly.

5/7

odon

toid

inju

ries

wer

etr

eate

dsu

cces

sful

lyw

ithha

lo.

2od

onto

id

inju

ries

wer

eop

erat

edon

imm

edia

tely

.2

odon

toid

inju

ries

faile

dex

tern

alim

mob

iliza

tion.

5A

AD

s

wer

ein

itial

lytr

eate

dw

ithsu

rgic

alfu

sion

.2

AA

Ds

initi

ally

trea

ted

with

halo

requ

ired

surg

ical

stab

iliza

tion.

Giv

ens

etal

.,19

96(2

0)R

evie

wof

34ch

ildre

nw

ithce

rvic

alsp

ine

inju

ries

over

a3-

yrpe

riod

.III

18in

juri

esoc

curr

edbe

low

C3.

The

leve

lof

inju

rydi

dno

tco

rrel

ate

with

age.

You

ngag

eis

not

asso

ciat

edw

ithex

clus

ivel

yup

per

cerv

ical

spin

ein

juri

es.

Turg

utet

al.,

1996

(61)

Ret

rosp

ectiv

ere

view

of82

child

ren

with

spin

alco

rdor

colu

mn

inju

ries

.III

14ch

ildre

n(1

7%)

wer

etr

eate

dsu

rgic

ally

.

Dor

man

set

al.,

1995

(11)

Are

view

of37

child

ren

with

halo

ring

san

dve

sts

aged

3–16

yr.

Arb

itrar

ilydi

vide

din

toth

ose

�10

yrol

d,an

dol

der.

IIIO

vera

ll68

%co

mpl

icat

ion

rate

.Pi

n-si

tein

fect

ion

was

the

mos

tco

mm

onco

mpl

icat

ion.

Puru

lent

infe

ctio

nsoc

curr

edm

ore

freq

uent

lyin

the

olde

rgr

oup.

Bot

hlo

osen

ing

and

infe

ctio

noc

curr

ed

mor

efr

eque

ntly

inth

ean

teri

orpi

nsi

tes.

Men

ticog

lou

etal

.,19

95(3

7)R

etro

spec

tive

case

seri

esof

15ne

onat

esw

ithbi

rth-

rela

ted

high

cerv

ical

cord

inju

ries

.

IIIA

ll15

wer

ece

phal

icpr

esen

tatio

nsin

whi

chfo

rcep

san

dat

tem

pted

rota

tion

wer

eus

ed.

All

but

1

wer

eap

neic

atbi

rth.

Cur

ran

etal

.,19

95(8

)Pr

ospe

ctiv

est

udy

of11

8ch

ildre

nw

hoar

rive

dim

mob

ilize

dto

asi

ngle

emer

genc

yro

om.

The

cerv

ical

spin

eal

ignm

ent

was

mea

sure

dan

d

com

pare

dw

ithag

ean

dty

peof

imm

obili

zatio

n.

IIIN

oco

rrel

atio

nw

ithde

gree

ofky

phos

isor

lord

osis

was

foun

dw

ithag

e.30

%ha

da

kyph

osis

of

�10

degr

ees.

No

sing

leim

mob

iliza

tion

tech

niqu

ew

assu

peri

or.

Schw

arz

etal

.,19

94(5

3)R

evie

wof

10ch

ildre

nw

ithve

rteb

ral

frac

ture

san

dky

phot

ican

gula

tion.

IIITh

eky

phot

ican

gula

tion

rem

aine

dun

chan

ged

orw

orse

ned

whe

nex

tern

alim

mob

iliza

tion

alon

e

(n�

7)or

dors

alfu

sion

(n�

1)w

asus

ed.

Onl

yth

ose

unde

rgoi

nga

vent

ral

fusi

on(n

�2)

had

a

stab

lere

duct

ion

ofth

eky

phot

icde

form

ity.

Nyp

aver

and

Trel

oar,

1994

(39)

40ch

ildre

nw

ere

plac

edon

spin

ebo

ards

,an

dob

serv

ers

judg

edw

heth

er

the

cerv

ical

spin

ew

asin

the

“neu

tral

”po

sitio

n.

IIIC

hild

ren

�8

yrof

age

requ

ired

tors

oel

evat

ion

toac

hiev

ene

utra

lal

ignm

ent.

Chi

ldre

n�

4yr

ofag

e

requ

ired

the

grea

test

amou

ntof

elev

atio

n.

Laha

met

al.,

1994

(32)

Div

ided

head

-inj

ured

child

ren

into

high

-(�

2yr

ofag

e,

nonc

omm

unic

ativ

e,or

with

neck

pain

)an

dlo

w-r

isk

grou

psfo

rce

rvic

al

spin

ein

jury

.

IIIN

oce

rvic

alsp

ine

inju

ries

dete

cted

inth

elo

w-r

isk

grou

p.10

inju

ries

(7.5

%)

wer

ede

tect

edin

the

high

-ris

kgr

oup.

Fotte

ret

al.,

1994

(17)

Rep

ort

ofbi

rth-

rela

ted

spin

alco

rdin

juri

esim

aged

with

ultr

asou

ndan

d

MR

I.

IIIA

neon

ate

with

com

plet

ein

jury

had

norm

alpl

ain

x-ra

ysw

ithsp

inal

ultr

asou

ndsh

owin

g

inho

mog

eneo

usec

hoge

nici

tyan

ddi

srup

ted

cord

surf

ace.

Ane

onat

ew

ithan

inco

mpl

ete

inju

ryha

d

inta

ctco

rdsu

rfac

ew

ithin

crea

sed

cord

echo

geni

city

.M

RI

corr

obor

ated

thes

efin

ding

s.

Mar

kset

al.,

1993

(36)

Rev

iew

of8

child

ren,

ages

3m

oto

12yr

,im

mob

ilize

din

aha

loja

cket

for

6w

kto

12m

o(m

ean,

2m

o).

IIITh

eon

lyco

mpl

icat

ion

was

aja

cket

chan

gere

quir

edfo

ra

fore

ign

body

(coi

n).

Onl

y3

ofth

ese

child

ren

had

cerv

ical

inst

abili

ty.



TAB

LE12

.2.

Con

tinu

ed

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ce

Cla

ssC

oncl

usio

ns

Shac

ked

etal

.,19

93(5

4)R

etro

spec

tive

revi

ewof

6ch

ildre

n(3

–14

yrol

d)w

ithce

rvic

alsp

ine

inju

ries

trea

ted

via

anan

teri

orap

proa

ch.

IIIA

utog

raft

with

out

inst

rum

enta

tion

afte

rco

rpec

tom

yw

asus

ed.

They

wer

est

abili

zed

post

oper

ativ

ely

with

hard

colla

ror

Min

erva

cast

.A

llw

ithso

lidfu

sion

s,go

odal

ignm

ent,

and

norm

alce

rvic

al

grow

th.

Follo

w-u

p,3–

8yr

.

Gro

gaar

det

al.,

1993

(23)

Atla

ntoa

xial

rota

rysu

blux

atio

nde

scri

bed

in9

child

ren.

8w

ere

diag

nose

dw

ithin

5d,

1w

asdi

agno

sed

afte

r8

wk.

III8

child

ren

wer

etr

eate

dsu

cces

sful

lyw

ith“m

ild”

trac

tion

and

then

aco

llar

for

4–6

wk.

1ch

ild

pres

entin

gla

tere

quir

ed1

wk

oftr

actio

nfo

rre

duct

ion.

Ther

ew

ere

2re

disl

ocat

ions

.A

llev

entu

ally

heal

edin

alig

nmen

tw

ithou

tsu

rger

y.

Man

daba

chet

al.,

1993

(35)

13ch

ildre

nw

ithax

isin

juri

esw

ere

revi

ewed

.10

wer

etr

eate

dpr

imar

ily

with

clos

edre

duct

ion

and

halo

imm

obili

zatio

n.

III8

ofth

e10

trea

ted

prim

arily

with

clos

edre

duct

ion

and

halo

imm

obili

zatio

nfu

sed.

2re

quir

ed

surg

ical

stab

iliza

tion

and

fusi

on.

Mac

Kin

non

etal

.,19

93(3

4)R

etro

spec

tive

case

seri

esof

22ne

onat

esw

ithbi

rth-

rela

ted

spin

alco

rd

inju

ries

.Th

eyex

clud

edne

onat

esw

ithSC

IWO

RA

.

IIIA

ll14

with

high

cerv

ical

inju

ries

had

ceph

alic

pres

enta

tions

with

atte

mpt

edfo

rcep

sro

tatio

n.A

ll6

with

cerv

icot

hora

cic

inju

ries

had

bree

chpr

esen

tatio

ns.

Bot

hne

onat

esw

ithth

orac

olum

bar

inju

ries

wer

epr

emat

ure.

Ros

sitc

han

dO

akes

,19

92(4

8)R

etro

spec

tive

revi

ewof

5ne

onat

esw

ithpe

rina

tal

spin

alco

rdin

jury

.N

o

flexi

on/e

xten

sion

view

sre

port

ed.

III4

ofth

e5

had

noab

norm

ality

onst

atic

spin

alx-

rays

.R

espi

rato

ryin

suffi

cien

cyan

dhy

poto

nia

wer

e

com

mon

sign

s.M

yelo

gram

sw

ere

unre

veal

ing.

All

3w

ithhi

ghce

rvic

alin

juri

esdi

edby

age

3yr

.

Ose

nbac

han

dM

enez

es,

1992

(41)

Ret

rosp

ectiv

ere

view

of17

9ch

ildre

nw

ithsp

inal

inju

ries

.III

59(3

3%)

unde

rwen

tsu

rgic

altr

eatm

ent

for

irre

duci

ble

unst

able

inju

ries

.83

%of

thos

etr

eate

d

surg

ical

lyw

ere

�9

yrof

age.

No

child

with

com

plet

eor

seve

repa

rtia

lm

yelo

path

yre

gain

edus

eful

func

tion.

Rat

hbon

eet

al.,

1992

(46)

Ret

rosp

ectiv

ere

view

of12

child

ren

with

pres

umed

spin

alco

rd

conc

ussi

ondu

ring

athl

etic

sw

ere

inve

stig

ated

for

the

pres

ence

ofce

rvic

al

sten

osis

.

III3

had

aTo

rgra

tio�

0.8,

and

4ha

da

cana

lan

tero

post

erio

rdi

amet

er�

13.4

mm

.M

RI

was

not

used

toev

alua

tefo

rst

enos

is.

Ham

ilton

and

Myl

es,

1992

(25)

Ret

rosp

ectiv

ere

view

ofal

lpe

diat

ric

spin

alin

juri

esdu

ring

14-y

rpe

riod

.

73ch

ildre

nha

dce

rvic

alin

juri

es.

IIISu

rger

yw

aspe

rfor

med

in26

%of

child

ren.

13%

ofch

ildre

nw

ithfr

actu

rean

dno

subl

uxat

ion,

50%

with

subl

uxat

ion

alon

e,an

d57

%w

ithfr

actu

rean

dsu

blux

atio

nw

ere

trea

ted

surg

ical

ly.

Of

39

child

ren

with

com

plet

em

yelo

path

y,4

impr

oved

one

ortw

oFr

anke

lgr

ades

.

Scha

ferm

eyer

etal

.,19

91(5

5)FV

Cw

asst

udie

din

heal

thy

child

ren

whe

nup

righ

t,su

pine

,an

dsu

pine

tape

dto

asp

inal

boar

d.

IIITa

ping

the

child

toth

esp

inal

boar

dca

used

FVC

tode

crea

se41

–96%

(mea

n,80

%)

ofsu

pine

FVC

.

Boh

net

al.,

1990

(4)

16/1

9ch

ildre

npr

esen

ting

with

abse

ntvi

tal

sign

sor

seve

rehy

pote

nsio

n

unex

plai

ned

bybl

ood

loss

unde

rwen

tpo

stm

orte

mex

amin

atio

n.

III13

/16

had

cord

lace

ratio

nor

tran

sect

ion.

2of

thes

ech

ildre

nha

da

norm

alce

rvic

alx-

ray.

Gas

kill

and

Mar

lin,

1990

(18)

6ch

ildre

nag

ed2–

4yr

wer

epl

aced

inM

iner

vaja

cket

sfo

rce

rvic

alsp

ine

inst

abili

ty.

III1

child

had

skin

brea

kdow

nof

the

chin

.Ea

ting

and

othe

rda

ilyac

tiviti

esw

ere

not

impa

ired

.2

wer

epl

aced

inM

iner

vaja

cket

saf

ter

com

plic

atio

nsof

halo

ring

and

vest

imm

obili

zatio

n.

Phill

ips

and

Hen

sing

er,

1989

(44)

Are

view

of23

child

ren

with

C1–

C2

rota

rysu

blux

atio

n.III

16ch

ildre

nse

enw

ithin

1m

oof

onse

tha

dei

ther

spon

tane

ous

redu

ctio

nor

redu

ced

with

trac

tion.

Of

the

7ch

ildre

npr

esen

ting

with

sym

ptom

sfo

r�

1m

o,1

subl

uxat

ion

was

irre

duci

ble,

and

4ha

d

recu

rren

tsu

blux

atio

ns.

Kaw

abe

etal

.,19

89(2

8)R

evie

wof

the

radi

olog

yof

17ch

ildre

nw

ithC

1–C

2ro

tary

subl

uxat

ion.

IIIC

lass

ified

acco

rdin

gto

Fiel

ding

and

Haw

kins

as10

Type

I,5

Type

II,2

Type

III,

and

noTy

peIV

.

Ben

zel

etal

.,19

89(2

)A

com

pari

son

ofce

rvic

alm

otio

nof

inju

red

patie

nts

(onl

y1

child

)

imm

obili

zed

inha

loan

dM

iner

vaja

cket

s.

IIITh

eM

iner

vaja

cket

allo

wed

less

mot

ion

than

the

halo

jack

etat

ever

yle

vel

exce

ptC

1–C

2.

Bau

met

al.,

1989

(1)

Are

view

com

pari

ngth

eha

loco

mpl

icat

ions

in13

child

ren

and

80

adul

ts.

III39

%co

mpl

icat

ion

rate

inch

ildre

nve

rsus

8%in

adul

ts.

The

child

ren

had

4pi

n-si

tein

fect

ions

and

1in

ner

tabl

ecr

ania

lpi

npe

netr

atio

n.

Mub

arak

etal

.,19

89(3

8)R

evie

wof

3ch

ildre

n�

2yr

old

who

wer

epl

aced

inha

lori

ngs

for

2–31

⁄2m

o.

III10

pins

tight

ened

“fin

ger-

tight

”in

a7-

mo-

old

child

,an

d2

in/lb

ina

16-

and

24-m

o-ol

d.2

of3

deve

lope

dm

inor

pin-

site

infe

ctio

nsne

cess

itatin

gpi

nre

mov

al.

Her

zenb

erg

etal

.,19

89(2

6)R

epor

ted

10ch

ildre

n�

7yr

ofag

ew

ithce

rvic

alsp

ine

inju

ries

posi

tione

don

afla

tba

ckbo

ard.

IIITh

ein

juri

esw

ere

ante

rior

lyan

gula

ted

ortr

ansl

ated

whe

non

afla

tba

ckbo

ard

beca

use

the

head

was

forc

edin

tofle

xion

.El

evat

ing

the

tors

oal

low

edfo

rm

ore

neut

ral

alig

nmen

tan

dre

duct

ion

of

the

inju

red

segm

ent.

Evan

san

dB

ethe

m,

1989

(14)

Rev

iew

of24

cons

ecut

ive

cerv

ical

spin

ein

juri

esin

child

ren

�18

yrol

d.III

Hal

fof

the

child

ren

had

inju

ries

atC

3or

abov

e.1

child

was

trea

ted

with

lam

inec

tom

yan

d2

with

fusi

on.

Frac

ture

she

aled

in21

/22

with

nono

pera

tive

ther

apy.

Bir

ney

and

Han

ley,

1988

(3a)

Ret

rosp

ectiv

ere

view

of61

child

ren

with

cerv

ical

spin

ein

juri

es.

23of

thes

ein

juri

esw

ere

C1–

C2

rota

rysu

blux

atio

n.

IIIR

otar

ysu

blux

atio

nun

asso

ciat

edw

ithne

urol

ogic

alde

ficit.

The

defo

rmity

reso

lved

with

halte

r

trac

tion

(n�

10)

orce

rvic

albr

acin

g.1

child

had

are

curr

ence

.A

child

with

tran

sver

selig

amen

t

disr

uptio

nw

astr

eate

dsu

cces

sful

lyw

itha

soft

colla

ron

ly.



TAB

LE12

.2.

Con

tinu

ed

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ce

Cla

ssC

oncl

usio

ns

Had

ley

etal

.,19

88(2

4)R

etro

spec

tive

revi

ewof

122

child

ren

with

spin

alin

juri

es.

Ther

ew

ere

97

cerv

ical

inju

ries

.

IIIO

nly

12ce

rvic

alin

juri

esw

ere

trea

ted

surg

ical

ly.

Hue

rta

etal

.,19

87(2

7)Ev

alua

tion

ofth

eim

mob

iliza

tion

ofco

mm

erci

ally

avai

labl

ein

fant

and

pedi

atri

cce

rvic

alco

llars

.

IIIN

oco

llar

used

alon

epr

ovid

edac

cept

able

imm

obili

zatio

n.Th

eus

eof

am

odifi

edha

lf-sp

ine

boar

d,

rigi

dco

llar,

and

tape

prov

ided

the

best

imm

obili

zatio

n.

Penn

ecot

etal

.,19

84(4

3)R

evie

wof

16ch

ildre

nw

ithlig

amen

tous

inju

ries

ofth

ece

rvic

alsp

ine.

5

with

C1–

C2

inju

ries

.

IIIO

fth

e11

child

ren

with

inju

ries

belo

wC

2,8

unde

rwen

tsu

rgic

alst

abili

zatio

n.Th

eyre

com

men

ded

a3-

mo

tria

lof

exte

rnal

imm

obili

zatio

nin

child

ren

with

ligam

ento

usin

juri

esbu

tno

neur

olog

ical

defic

itor

disl

ocat

ion.

El-K

hour

yet

al.,

1984

(13)

Rev

iew

of3

child

ren

with

C1–

C2

rota

rysu

blux

atio

n.III

All

3w

ere

trea

ted

succ

essf

ully

with

trac

tion

orm

anua

lre

duct

ion

with

in24

hof

pres

enta

tion.

One

child

had

recu

rren

tsu

blux

atio

nth

ene

xtda

yan

dw

astr

eate

dsu

cces

sful

lyw

ithm

anua

lre

duct

ion.

Exte

rnal

orth

oses

wer

eus

edfo

r10

wk,

3m

o,an

d4

mo,

resp

ectiv

ely.

Koo

pet

al.,

1984

(30)

Ret

rosp

ectiv

ere

view

of13

child

ren

with

cerv

ical

inst

abili

tytr

eate

dw

ith

post

erio

rar

thro

desi

san

dha

loim

mob

iliza

tion.

Onl

y3

had

trau

mat

ic

lesi

ons.

III1

faile

dfu

sion

whe

nba

nk-b

one

was

used

.O

ther

ssu

cces

sful

lyfu

sed

with

auto

geno

usili

accr

est

or

rib.

Inte

rnal

wir

ing

used

in2

child

ren.

Ave

rage

halo

imm

obili

zatio

nw

as15

0d.

Sher

ket

al.,

1978

(57)

Rep

ort

of11

child

ren

with

odon

toid

inju

ries

,an

dre

view

of24

from

the

liter

atur

e.

IIIM

ajor

ityof

inju

red

odon

toid

sar

ean

gled

ante

rior

ly.

All

but

1ch

ildw

astr

eate

dsu

cces

sful

lyw

ith

exte

rnal

imm

obili

zatio

n.

Fiel

ding

and

Haw

kins

,19

77(1

5)Th

era

diog

raph

icfin

ding

sof

17pa

tient

sw

ithat

lant

oaxi

alro

tary

fixat

ion

are

desc

ribe

dan

dcl

assi

fied

into

4ty

pes.

III4

clas

ses

ofC

1–C

2ro

tary

subl

uxat

ion

wer

ede

scri

bed,

Type

sI–

IV.

Type

I:od

onto

idac

tsas

pivo

tw

ithco

mpe

tent

tran

sver

selig

amen

t.

Type

II:on

ela

tera

lar

ticul

arpr

oces

sac

tsas

pivo

tw

ithup

to5

mm

ofan

teri

ordi

spla

cem

ent.

Type

III:

both

C1

infe

rior

face

tsar

esu

blux

edan

teri

orly

with

�5

mm

ofan

teri

ordi

spla

cem

ent,

whi

chsu

gges

tsan

inco

mpe

tent

tran

sver

selig

amen

t.

Type

IV:

post

erio

rdi

spla

cem

ent

with

abse

ntor

inco

mpe

tent

odon

toid

.

Gau

finan

dG

oodm

an,

1975

(19)

Are

view

of3

child

ren

�20

mo

old

with

cerv

ical

spin

ein

juri

es.

2of

thes

ech

ildre

nw

ere

trea

ted

with

trac

tion

deliv

ered

via

22-g

auge

wir

e

plac

edth

roug

hbi

late

ral

pari

etal

burr

hole

s.

IIISu

cces

sful

trac

tion

appl

ied

toth

e10

-wk-

old

and

16-m

o-ol

dch

ild.

�9

lbus

edin

the

10-w

k-ol

d

infa

nt.

No

com

plic

atio

nsw

ere

enco

unte

red

with

the

trac

tion

inpl

ace

for

14an

d41

d,

resp

ectiv

ely.

aA

AD

,at

lant

oaxi

aldi

sloc

atio

n;M

RI,

mag

neti

cre

sona

nce

imag

ing;

SCIW

OR

A,

spin

alco

rdin

jury

wit

hout

radi

ogra

phic

abno

rmal

ity;

FVC

,fo

rced

vita

lca

paci

ty.



(7.5%) in the high-risk group. The authors concluded thatcervical spine x-rays are not necessary in children with iso-lated head injuries who can communicate and have no neckpain or neurological deficit. Bohn et al. (4) emphasized thatunexplained hypotension or absent vital signs in pediatrictrauma patients are likely to result from a severe cervical cordinjury. Therefore, they advocate suspicion for a cervical SCI inchildren with either multisystem trauma or an isolated headinjury with hypotension or cardiopulmonary arrest.

Viccellio et al. (62) evaluated the cervical spines in childrenyounger than 18 years using the National EmergencyX-radiography Utilization Study (NEXUS) decision instru-ment in a Class II prospective multicenter study. They usedfive low-risk criteria. These criteria were: 1) the absence ofmidline cervical tenderness, 2) evidence of intoxication, 3)altered level of alertness, 4) focal neurological deficit, and 5) apainful distracting injury. X-rays were obtained at the discre-tion of the treating physician. When x-rays were obtained, aminimum of three views were obtained. Only those patientsfor whom x-rays were obtained were included in the study. Ifall five criteria were met, the child was considered to be at lowrisk. If any one of the five criteria was present, the child wasconsidered to be at high risk. Of 3065 children evaluated, 603fulfilled the low-risk criteria. None of these 603 children de-fined as low-risk had a documented cervical spine injury byradiographic evaluation. Thirty injuries (0.98%) were docu-mented in children not fulfilling the low-risk criteria. Theauthors concluded that applying the NEXUS criteria to chil-dren would reduce cervical spine x-ray use by 20% and wouldnot result in missed injuries. They cautioned that they hadstudied relatively small numbers of young children youngerthan 2 years (n � 88). Statistically, this created large confi-dence intervals for the sensitivity of their instrument whenapplied to younger children. From this Class II study, they“cautiously” endorsed the application of NEXUS criteria inchildren, particularly those from birth to age 9 years. Theirconclusions are consistent with the Class III evidence previ-ously described by Laham et al. (32) on this topic.

The need for and usefulness of open-mouth odontoid viewsin pediatric trauma patients has been questioned (6, 59). Swis-chuk et al. (59) surveyed 984 pediatric radiologists to deter-mine how many injuries were missed on lateral cervical spinex-rays but detected on an open-mouth view. There were 432responses. One hundred sixty-one respondents did not rou-tinely use open-mouth views. Of the 271 radiologists whoobtained open-mouth views in young children, 191 (70%)would not persist beyond a single attempt. Seventy-one radi-ologists (26%) would make up to five attempts to obtain anadequate image. Twenty-eight (7%) of the 432 respondentsreported missing a total of 46 fractures on the lateral view thatwere detected on the open-mouth view. The types of injurieswere not classified (i.e., odontoid versus C1 injury). The au-thors calculated a missed fracture rate of 0.007 per year perradiologist in their study. They concluded that the open-mouth view x-ray might not be needed routinely in childrenyounger than 5 years. Buhs et al. (6) also investigated theusefulness of open mouth views in children. They performeda multi-institutional retrospective review of a large metropol-

itan population of patients younger than 16 years who wereassessed for cervical spine trauma over a 10-year period.Fifty-one children with cervical spinal injuries were identi-fied. The lateral cervical spine x-ray confirmed the diagnosisin 13 of 15 children younger than 9 years. In none of the 15younger patients did the open-mouth view provide the diag-nosis. In only 1 of 36 patients in the group aged 9 to 16 yearswas the open-mouth view the diagnostic study (a Type IIIodontoid injury). The authors concluded that the open-mouthview x-ray is not necessary for clearing the cervical spine inchildren younger than 9 years.

Lui et al. (33), in their review of 22 children with C1–C2injuries, commented that flexion/extension x-rays were re-quired to “identify the instability” of traumatic injuries to thedens in 4 of 12 children with odontoid fractures, and in 6 of 9children with purely ligamentous injuries resulting in atlan-toaxial dislocation. The authors did not state whether anabnormality on the static x-ray led to the dynamic studies, orwhether the initial static studies were normal. Because theydid not describe flexion/extension x-rays as part of theirroutine for the assessment of children with potential cervicalspine injuries, it is likely that some imaging or clinical findingprompted the decision to obtain dynamic films in thesechildren.

The experience of Ruge et al. (49) highlighted the propen-sity for upper cervical injuries in children younger than 9years. They reported no injuries below C3. Evans and Bethem(14) described 24 children with cervical spine injuries. In halfof the patients, the injury was at C3 or higher. Givens et al.(20), however, described the occurrence of important injuriesoccurring at all levels of the cervical spine in young children.They described 34 children with cervical spine injuries. Therewas no correlation of level of injury with age. Two of thechildren they managed had injuries at C7–T1. Hence, it wouldbe dangerous to assume that lower cervical spine injuries donot occur in young children and irresponsible to discount theneed for adequate imaging of the lower cervical spine andcervical-thoracic junction in these young patients.

Scarrow et al. (50) attempted to define a protocol to evaluatethe cervical spine in obtunded children after trauma. Theyused somatosensory evoked responses during flexion/extension fluoroscopy. Of the 15 children evaluated with thisprotocol, none showed pathological motion during flexion/extension fluoroscopy. Three children were thought to have achange in the evoked responses during flexion/extension.Only one of the three children with an abnormal evokedresponse underwent magnetic resonance imaging (MRI) thatwas normal. Their investigation failed to demonstrate anyusefulness for evoked responses, flexion/extension fluoros-copy, or MRI of the cervical spine in the evaluation of thecervical spine in children with altered mental status aftertrauma. Larger numbers of children investigated in this man-ner might define a role for one or more of these diagnosticmaneuvers, but as yet there is no evidence to support theiruse.

Ralston et al. (45) retrospectively analyzed the cervicalspine x-rays of 129 children who had flexion/extension x-raysperformed after an initial static x-ray. They found that if the



static x-ray was normal or depicted only loss of lordosis, theflexion/extension views would reveal no abnormality. Theauthors concluded that the value of the dynamic x-rays wasconfirmation of cervical spinal stability when there was aquestionable finding on the static lateral x-ray. Likewise,Dwek and Chung (11a) retrospectively reviewed 247 childrenwith a history of trauma in whom flexion-extension cervicalx-rays were obtained. Flexion-extension studies were normalin all children who had normal neutral x-rays. The authorsconcluded that the use of flexion-extension x-rays after ob-taining normal static x-rays is of “questionable use.”

The interpretation of cervical spine x-rays must account forthe age and anatomic maturation of the patient. Commonnormal findings on cervical spine x-rays obtained on youngchildren are pseudosubluxation of C2 on C3, overriding of theanterior atlas in relation to the odontoid on extension, exag-gerated atlantodens intervals, and the radiolucent synchon-drosis between the odontoid and C2 body. These normalfindings can be mistaken for acute traumatic injuries in chil-dren after trauma. Cattell and Filtzer (7) obtained lateralcervical x-rays in neutral, flexion, and extension in 160 ran-domly selected children who had no history of trauma orhead and neck problems. The subjects’ ages ranged from 1 to16 years, with 10 children for each year of age. They found a24% incidence of moderate to marked C2 on C3 subluxation inchildren aged 1 through 7 years. Thirty-two (46%) of 70 chil-dren younger than 8 years had 3 mm or more of anterior-posterior motion of C2 on C3 on flexion/extension x-rays.Fourteen percent of all children had radiographic pseudosub-luxation of C3 on C4. Twenty percent of children aged 1through 7 years had an atlantodens interval of 3 mm or more.Overriding of the anterior arch of the atlas on the odontoidwas present in 20% of children younger than 8 years. Thesynchondrosis between the odontoid and axis body was notedas a lucency in all children imaged up to age 4 years. Thesynchondrosis remained visible in half the children up to age11 years. The authors also described an absence of the normalcervical lordosis in 14% of subjects, most commonly in the age8- to 16-year groups. Shaw et al. (56), in a retrospective reviewof cervical spine x-rays in 138 children younger than 16 yearswho were evaluated after trauma, found a 22% incidence ofradiographic pseudosubluxation of C2 on C3. The only factorthat correlated with the presence of pseudosubluxation intheir study was patient age. The pseudosubluxation grouphad a median age of 6.5 years versus 9 years in the groupwithout this finding. It was identified, however, in children asold as 14 years. Intubation status, injury severity score, andsex had no correlation with pseudosubluxation of C2 on C3.To differentiate between physiological and traumatic subluxa-tions, they recommend a method that involves drawing a linethrough the posterior arches of C1 and C3. In the circum-stance of pseudosubluxation of C2 on C3, the C1–C3 lineshould pass through, touch, or lie up to 1 mm anterior to theanterior cortex of the posterior arch of C2. If the anteriorcortex of the posterior arch of C2 is 2 mm or more behind theline, then a true dislocation (rather than pseudosubluxation)should be assumed.

Keiper et al. (29) reviewed their experience of using MRI inthe evaluation of children with clinical evidence of cervicalspine trauma who had no evidence of fracture by plain x-raysor computed tomography (CT), but who had persistent ordelayed symptoms or instability. There were 16 abnormalMRI examinations in 52 children. Posterior soft tissue andligamentous changes were described as the most commonabnormalities. Only one child had a bulging disc. Four ofthese 52 children underwent surgical treatment. In each of thefour surgical cases, the MRI findings led the surgeon to sta-bilize more levels than otherwise would have been under-taken without the MRI information. Davis et al. (9) describedthe use of MRI in evaluating pediatric SCI and found that itdid not reveal any lesion that would warrant surgical decom-pression. These authors did note, however, that MRI findingswere correlated with neurological outcome. Evidence of he-matomyelia was associated with permanent neurological def-icit. Although little information is available on this subject, itseems that preoperative MRI studies of children with unstablecervical spinal injuries, who require surgical stabilization,may affect the specifics of the surgical management.

There are no studies that have systematically reviewed therole of CT in the evaluation of the cervical spines of pediatricpatients after trauma. In children younger than 10 years withcervical spinal injuries, most patients will have ligamentousinjuries without fracture (10, 12, 24, 25, 41). In older childrenwith cervical spinal injuries, the incidence of a fracture ismuch greater than ligamentous injury without fracture, 80%versus 20%, respectively (14, 62). Therefore, normal osseousanatomy as depicted on an axial computed tomographic im-age should not be used alone to exclude injury to the pediatriccervical spine. Schleehauf et al. (52), in reporting a series ofpediatric and adult trauma patients in 1989, concluded thatCT should not be relied on to exclude ligamentous injuries.They reported two false-negative computed tomographicstudies in patients with C1–C2 ligamentous injuries in theirstudy of the merits of CT to evaluate the cervical spine inhigh-risk trauma patients. The authors favored CT for theevaluation of regions that could not be viewed adequatelywith plain x-rays (e.g., C7–T1) and for the investigation of theosseous integrity of specific vertebrae suspicious for fractureon plain x-rays.

In a series consisting almost entirely of adults, the role ofhelical CT in the evaluation of the cervical spine in “high-risk”patients after severe, blunt, multisystem trauma was prospec-tively studied (3). The plain spine x-rays and computed to-mographic images were reviewed by a radiologist blinded tothe patients and their history. The investigators found 20cervical spine injuries (12 stable, 8 unstable) in 58 patients(34%). Eight of these injuries (five stable, three unstable) werenot detected on plain x-rays. The authors concluded thathelical cervical spinal CT should be used to assess the cervicalspine in high-risk trauma patients. In young children inwhom the entire cervical spine is often easily and accuratelyvisualized on plain x-ray studies, the need for cervical spinalhelical CT is likely not as great. In older high-risk childrenwho have spinal biomechanics and injury patterns more con-



sistent with those of adult trauma patients, helical CT of thecervical spine may be fruitful.

In summary, to clear a child’s cervical spine, Class II andClass III evidence supports obtaining lateral and anteropos-terior cervical spine x-rays in children who have experiencedtrauma and cannot communicate because of age or headinjury, have a neurological deficit, have neck pain, have apainful distracting injury, or are intoxicated. In children whoare alert, have no neurological deficit, no midline cervicaltenderness, and no painful distracting injury, and are notintoxicated, cervical spine x-rays are not necessary to excludecervical spine injury (32, 62). Unexplained hypotensionshould raise the suspicion of SCI. Open-mouth views of theodontoid do not seem to be useful in children younger than 9years. Open-mouth views should be attempted in childrenaged 9 years and older. Flexion/extension studies (fluorosco-py or x-rays) are likely to be unrevealing in children withstatic x-rays proven to be normal. Dynamic studies could beconsidered, however, when the static x-rays or the child’sclinical findings suggest but do not definitively demonstratecervical spinal instability. CT of the cervical spine should beused judiciously to define bony anatomy at specific levels butis not recommended as a means to clear the entire cervicalspine in children. MRI may provide important informationabout ligamentous injury that may influence surgical manage-ment, and it may provide prognostic information regardingneurological outcome.

Injury management

Injury patterns that have a strong predilection for or areunique to children merit discussion because of the specializedmanagement paradigms used to treat them. SCIs withoutradiographic abnormality and atlanto-occipital dislocation in-juries have been addressed elsewhere (see Chapters 13 and14). SCIs secondary to birth-related trauma and epiphysiolysisof the axis are injuries unique to children. Common but notunique to children are C1–C2 rotary subluxation injuries.These entities will be discussed below in light of the availableliterature. It should be noted that there is no informationprovided in the literature describing the medical managementof pediatric patients with SCI. The issue of corticosteroidadministration after acute pediatric SCI, for example, has notbeen addressed. Although prospective, randomized clinicaltrials such as the Second and Third National Acute SpinalCord Injury Studies have evaluated pharmacological therapyafter acute SCI, children younger than 13 years were excludedfrom the study (5).

Neonatal SCI

Birth injuries of the spinal cord occur in approximately 1per 60,000 births (63). The most common level of injury isupper cervical and then cervicothoracic (34). MacKinnon et al.(34) described 22 neonates with birth-related SCIs. The diag-nosis was defined by the following criteria: clinical findings ofacute cord injury for at least 1 day and evidence of spinal cordor spinal column injury by imaging or electrophysiologicalstudies. Fourteen neonates had upper cervical injuries, six had

cervicothoracic injuries, and two had thoracolumbar injuries.All upper cervical cord injuries were associated with cephalicpresentation and the use of forceps for rotational maneuvers.Cervicothoracic injuries were associated with the breech pre-sentation. All infants had signs of “spinal shock,” defined asflaccidity, no spontaneous motion, and no deep tendon re-flexes. Of the nine infants with upper cervical injuries surviv-ing longer than 3 months, seven were alive at last follow-up.Six of these seven are dependent on mechanical ventilation.The two neonates with upper cervical injuries who hadbreathing movements on Day 1 of life were the only twothought to have satisfactory outcomes. All survivors withupper cervical cord injuries whose first respiratory effort wasbeyond the first 24 hours of life have remained ventilator-dependent. Only two children of six who sustained cervico-thoracic SCI lived and remained paraplegic. One requiresmechanical ventilation. Hypoxic and ischemic encephalopa-thy was noted in 9 of 14 newborns with upper cervical cordinjuries, and in 1 of 6 with a cervicothoracic cord injury. Theauthors did not describe any treatment provided for the un-derlying spinal column or cord injury, or whether survivorsexperienced progression of any spinal deformities.

Menticoglou et al. (37), drawing partly from the same pa-tient data as MacKinnon et al. (34), reported 15 neonates withbirth-related upper cervical SCIs. All were associated withcephalic deliveries requiring rotational maneuvers with for-ceps. All but one child was apneic at birth with quadriplegia.There is no description of postinjury spinal column or spinalcord management, medical or surgical, in their report. Ros-sitch and Oakes (48) described five neonates with birth-related SCI. They reported that incorrect diagnoses weremade in four. They consisted of Werdnig-Hoffmann syn-drome, occult myelodysplasia, and birth asphyxia. Only oneneonate had an abnormal plain x-ray (atlanto-occipital dislo-cation). They provided no description of the management ofthe spinal cord or column injuries in these five neonates.Fotter et al. (17) reported the use of bedside ultrasound todiagnose neonatal SCI. They found excellent correlation withMRI studies with respect to the extent of cord injury in theirtwo cases. Pang and Hanley (42) provided the only descrip-tion of an external immobilization device for neonates. Theydescribed a thermoplastic molded device that is contoured tothe occiput, neck, and thorax. Velcro straps cross the foreheadand torso, securing the infant and immobilizing the spinalcolumn.

In summary, cervical instability after birth-related SCI is notaddressed in the literature. The extremely high mortality rateassociated with birth-related SCI may have generated thera-peutic nihilism for this entity, hence the lack of aggressivemanagement. The literature suggests that the presentation ofapnea with flaccid quadriplegia after cephalic presentationwith forceps manipulation is the hallmark of upper cervicalSCI. Absence of respiratory effort within the first 24 hours oflife is associated with dependence on long-term mechanicalventilation. It seems reasonable to treat these neonates withspinal immobilization for a presumed cervical spinal injury.The method and length of immobilization is at presentarbitrary.



Odontoid epiphysiolysis

The neurocentral synchondrosis of C2, which may not fusecompletely until age 7 years, represents a vulnerable site ofinjury in young children (22). The lateral cervical spine x-rayis the diagnostic imaging modality of choice to depict thisinjury. It will often reveal the odontoid process to be angu-lated anteriorly, and rarely posteriorly (57). Although injuriesto the neurocentral or subdental synchondrosis may be seenin children up to age 7 years, it most commonly occurs inpreschool-aged children (35). Mandabach et al. (35) described13 children with odontoid injuries ranging in age from 9months to 7 years. The authors reported that 8 of 10 childrenwho were initially managed with halo immobilization aloneachieved stable fusion. The average time to fusion was 13weeks, with a range of 10 to 18 weeks. Because the injuryoccurs through the epiphysis, it has a high likelihood ofhealing if closed reduction and immobilization are used.Mandabach et al., in their review, cited several other reportsdescribing the successful treatment of young children withodontoid injuries who were managed with a variety of exter-nal immobilization devices. Sherk et al. (57) reported 11 chil-dren with odontoid injuries and reviewed an additional 24from the literature. Only 1 of these 35 children required sur-gical fusion. Although the literature describes the use of Min-erva jackets, soft collars, hard collars, and the halo vest asmeans of external immobilization to achieve successful fusionin young children with odontoid injuries, the halo is the mostwidely used immobilization device in the contemporary lit-erature for these injuries (35, 40, 57).

To obtain injury reduction in these children, Mandabach etal. (35) advocate the application of the halo device underketamine anesthesia and then realignment of the dens utiliz-ing C-arm fluoroscopy. Other reports describe using tractionto obtain alignment before immobilizing the child in an ex-ternal orthosis (22). Compared with halo application and im-mediate reduction and immobilization, traction requires aperiod of bed rest and is associated with the potential risk ofoverdistraction (35).

The literature is scant regarding the operative treatment ofC2 epiphysiolysis. Most reports describe using operative in-ternal fixation and fusion only if external immobilization hasfailed to maintain reduction or achieve stability. Reinges et al.(47) noted that only three “young” children have been re-ported in the literature who have had odontoid injuries pri-marily treated with surgical stabilization. This underscoresthe near universal application of external immobilization asthe primary means of treating odontoid injuries in youngchildren. Odent et al. (40) reported that of the 15 youngchildren with odontoid injuries they managed, three that weretreated with surgical stabilization and fusion experiencedcomplications. The other 12 children with similar injuries,managed without operation, did well. Wang et al. (64) de-scribed using anterior odontoid screw fixation as the primarytreatment option in a 3-year-old child with C2 epiphysiolysis.A hard cervical collar was used postoperatively. Halo immo-bilization was not used either preoperatively or postopera-tively. They successfully used anterior odontoid screw fixa-

tion as the primary treatment in two older children (ages 10and 14 years) and then hard collar immobilization. It is likelythat these two children had true Type II odontoid fracturesand not C2 epiphysiolysis. Likewise, Godard et al. (21) per-formed anterior odontoid screw fixation in a 2-year-old childwith a severe head injury. They used skeletal traction to alignthe fracture preoperatively. The rationale for proceeding tooperative stabilization without an attempt at treatment withexternal immobilization was to avoid the halo orthosis and toallow for more aggressive physiotherapy in this severely in-jured child. They believe that anterior odontoid screw fixationis advantageous because no motion segments are fused, nor-mal motion is preserved, and the need for halo immobiliza-tion is obviated.

For management of injuries of the C2 neurocentral syn-chondrosis, the literature supports the use of closed reductionand external immobilization for approximately 10 weeks. Thisstrategy is associated with an 80% fusion success rate. Al-though primary surgical stabilization of this injury has beenreported, the experience in the literature is limited. Surgicalstabilization seems to play a role when external immobiliza-tion is unable to maintain alignment of the odontoid atop theC2 body. Although both anterior and posterior surgical ap-proaches have been successfully used in this setting, there aremore reports describing posterior C1–C2 techniques than re-ports describing anterior operative techniques.

Atlantoaxial rotary subluxation

Fixed rotary subluxation of the atlantoaxial complex is notunique to children, but it is more common during childhood.It can present after minor trauma, in association with anupper respiratory infection, or without an identifiable incitingevent. The head is rotated to one side with the head tilted tothe other side, causing the so-called cock-robin appearance.The child is unable to turn his or her head past the midline.Attempts to move the neck are often painful. The neurologicalstatus is almost always normal (13, 31, 44, 58).

It can be difficult to differentiate atlantoaxial rotary sublux-ation from other causes of head rotation on clinical groundsalone. Several reports describe the radiographic characteriza-tion and diagnosis of this entity. Fielding and Hawkins (15)described 17 children and adults with atlantoaxial rotary sub-luxation and classified their dislocations into four types basedon radiographic features. Type I was the most common type,identified in 8 of the 17 patients. It was described as unilateralanterior rotation of the atlas pivoting around the dens with acompetent transverse ligament. Type II was identified in fivepatients. It was described as unilateral anterior subluxation ofthe atlas with the pivot being the contralateral C1–C2 facet.The atlantodens interval is increased to no more than 5 mm.Type III is described as anterior subluxation of both C1 facetswith an incompetent transverse ligament. Type IV is posteriordisplacement of C1 relative to C2 with an absent or hypoplas-tic odontoid process.

Kawabe et al. (28) reviewed the x-rays of a series of 17children with C1–C2 rotary subluxation and classified themaccording to Fielding and Hawkins (15). There were 10 Type



I, five Type II, two Type III, and no Type IV subluxations intheir experience. CT has been used to help define the C1–C2complex in cases of suspected rotary subluxation. Kowalski etal. (31) demonstrated the superiority of dynamic computedtomographic studies compared with information obtainedwith static computed tomographic studies. They comparedthe computed tomographic scans of 8 patients with C1–C2pathology to computed tomographic studies of six normalsubjects. The computed tomographic scans obtained with nor-mal subjects maximally rotating their heads could not bedifferentiated from the computed tomographic scans of thosewith known C1–C2 rotary subluxation. When CT was per-formed with the head rotated as far as possible to the con-tralateral side, computed tomographic studies of normal sub-jects could be easily differentiated from those performed onpatients with rotary subluxation.

Type I and Type II subluxation account for most rotaryatlantoaxial subluxations in reports describing these injuries.Grogaard et al. (23) and Subach et al. (58) published retro-spective reviews on the success of conservative therapies inchildren presenting early after C1–C2 rotary subluxation.Grogaard et al. (23) described eight children who presentedwithin 5 days of subluxation, and one child who presented 8weeks after injury. All were successfully treated with closedreduction and immobilization. The child presenting late re-quired 1 week of skeletal traction to achieve reduction andwas ultimately treated with halo immobilization for 10 weeks.The children who presented early had their injuries reducedwith manual manipulation. They were treated in a hard collarfor 4 to 6 weeks. Two patients had recurrent subluxation. Bothwere reduced and treated successfully without surgical inter-vention. Subach et al. (58) reported 20 children with C1–C2rotary subluxation, in whom four injuries reduced spontane-ously. Injury reduction was accomplished in 15 of 16 patientstreated with traction for a mean duration of 4 days. Six chil-dren required fusion because of recurrent subluxation (n � 5)or irreducible subluxation (n � 1). No child experienced re-current subluxation if reduced within 21 days of symptomonset.

El-Khoury et al. (13) reported three children who presentedwithin 24 hours of traumatic rotary subluxation. All threewere successfully treated with traction or manual reductionwithin 24 hours of presentation. One child experienced recur-rent subluxation the next day that was successfully reducedmanually. External orthoses were used from 10 weeks to 4months. Phillips and Hensinger (44) reviewed 23 childrenwith C1–C2 rotary subluxation. Sixteen children were seenwithin 1 month of subluxation onset and either experiencedspontaneous reduction or were reduced with traction. Ofseven children presenting with a duration of symptoms ofmore than 1 month, one subluxation was irreducible, and fourrecurred after initial reduction. Schwarz (53) described fourchildren who presented more than 3 months after the onset ofC1–C2 rotary subluxation. Two children had irreducible sub-luxations. One child had recurrent subluxation despite the useof a Minerva cast. Only one child had successful treatmentwith closed reduction and a Minerva cast immobilization for8 weeks. These experiences highlight the ease and success of

nonsurgical management for these injuries when the sublux-ation is treated early rather than late. If the subluxation iseasily reducible and treated early, 4 weeks in a rigid collarseems to be sufficient for healing. Because C1–C2 rotary sub-luxation can reduce spontaneously in the first week, tractionor manipulation can be reserved for those subluxations thatdo not reduce spontaneously in the first few days. The use ofmore restrictive external immobilization devices (e.g., halovest, Minerva cast) for longer periods of treatment up to 4months has been described in those children presenting late,or those who have recurrent subluxations (44). Operativetreatment for C1–C2 rotary subluxations has been reserved forrecurrent subluxations or those that cannot be reduced byclosed means. Subach et al. (58) operated on 6 of the 20children they reported with rotary subluxation using theseindications. The authors used a posterior approach and ac-complished atlantoaxial fusion. There were no complications,and all fusions were successful.

In summary, the diagnosis of atlantoaxial rotary subluxa-tion is suggested when findings of a “cock-robin” appearanceare present: the patient is unable to turn the head past themidline to the contralateral side and experiences spasm of theipsilateral (the side to which the chin is turned) sternocleido-mastoid muscle (44). Plain cervical spine x-rays may revealthe lateral mass of C1 rotated anterior to the odontoid on alateral view. The anteroposterior x-ray may demonstrate ro-tation of the spinous processes toward the ipsilateral side in acompensatory motion to restore alignment. If the diagnosis ofC1–C2 rotary subluxation is not certain after clinical exami-nation and plain radiographic study, a dynamic computedtomographic study should be considered. It seems that thelonger a C1–C2 rotary subluxation is present before attemptedtreatment, the less likely reduction can be accomplished. Ifreduction is accomplished in these older injuries, it is lesslikely to be maintained. Therefore, rotary subluxations that donot reduce spontaneously should undergo attempted reduc-tion with manipulation or traction. The subsequent period ofimmobilization should be proportional to the length of timethat the subluxation was present before treatment. Surgicalarthrodesis can be considered for those with irreducible sub-luxations, recurrent subluxations, or subluxations present formore than 3 weeks duration.

Other injuries

Lui et al. (33) described nine children with ligamentousinjuries resulting in atlantoaxial dislocation. Unlike childrenwith traumatic injuries to the dens who can be managed withclosed reduction and immobilization, these children with at-lantoaxial dislocation required surgical stabilization and fu-sion. The authors attempted to treat two children with haloimmobilization for 3 months duration; both attempts failed toachieve stability. All nine children with atlantoaxial instabilityrequired operative stabilization and fusion. Finally, Rathboneet al. (46) described a series of 12 children who sustained a“spinal cord concussion” while participating in athleticevents. They found that four of these children had plain spinex-rays consistent with cervical spinal stenosis. The authors



raised the concern that children with congenital cervical ste-nosis may be more susceptible to SCI in contact sports.

Therapeutic cervical spine immobilization

Once an injury to the pediatric cervical spine has been diag-nosed, some form of external immobilization is usually neces-sary to allow for application of traction either to restore align-ment or to immobilize the spine to allow for healing of the injury.This section will discuss the literature available concerningmethods of skeletal traction in children and various externalorthoses used to immobilize the pediatric cervical spine.

Traction for the purpose of restoring alignment or reducingneural compression in children is rarely addressed in the litera-ture. Unique concerns of cervical traction in children exist be-cause of the relatively thinner cranium with a higher likelihoodof inner cranial table penetration, lighter body weight that pro-vides less counterforce to traction, more elastic ligaments, andless well-developed musculature, increasing the potential foroverdistraction. The placement of bilateral pairs of parietal burrholes and passing 22-gauge wire through them to provide apoint of fixation for traction has been described for infants withcervical spinal injuries. Gaufin and Goodman (19) reported aseries of three infants with cervical injuries, two of whom hadinjuries reduced in this fashion. Up to 9 pounds was used in a10-week-old infant and a 16-month-old boy. They experiencedno complications with 14 and 41 days of traction, respectively.Other techniques of cervical traction application in children arenot described in the literature.

Mubarak et al. (38) described halo application in infants forthe purpose of immobilization, but not halo-ring traction.They described three infants aged 7 months, 16 months, and24 months. Ten pins were used in each child. The pins in theyoungest child were inserted to finger tightness only, whereasthe older children had 2 inches per pound of torque applied(38). The children were maintained in the halo devices for 2 to3.5 months. Only the youngest child had a minor complica-tion of frontal pin-site infection necessitating removal of twoanterior pins.

Marks et al. (36) described eight children aged 3 months to12 years who were immobilized in halo vests for 6 weeks to 12months with a mean duration of 2 months. Only three of thesechildren had cervical spinal instability. Five had thoracic spi-nal disorders. The only complication they reported was theneed to remove and replace the vest when a foreign bodybecame lodged under the vest. Dormans et al. (11) reported on37 children aged 3 to 12 years whom they managed in haloimmobilization devices. They had a 68% complication rate.Pin-site infections were most common. They arbitrarily di-vided their patient population into those younger than 10years and those 10 years or older. Purulent pin-site infectionsoccurred more commonly in the older group. Loosening ofpins occurred more commonly in the younger group. Bothloosening and infection occurred more often at the anteriorpin sites. They also reported one incident of dural penetrationand one transient supraorbital nerve injury. Baum et al. (1)compared halo use complications in children and adults. Thecomplication rates in their series were 8% for adults and 39%

for children. The complications reported for the children wereone cranial penetration and four pin-site infections. Althoughthe halo device seems to provide adequate immobilization ofthe cervical spine in children, the rate of minor complicationswith halo use is higher in children than in adults.

Gaskill and Marlin (18) described six children aged 2 yearsto 4 years who had cervical spinal instability managed with athermoplastic Minerva orthosis as an alternative to a haloimmobilization device. Two of the children they describedhad halo devices removed because of complications beforebeing placed in Minerva orthoses. The authors described noproblems with eating or with activities of daily living in thesechildren. Only one child had a minor complication from Min-erva use, a site of skin breakdown. The authors concludedthat immobilization with a thermoplastic Minerva orthosisoffered a reliable and satisfactory alternative to halo immobi-lization in young children.

Benzel et al. (2) analyzed cervical motion during spinal immo-bilization in adults serially treated with halo and Minerva de-vices. They found that the Minerva offered superior immobili-zation at all intersegmental levels of the cervical spine, with theexception of C1–C2. Although this study was carried out inadults with cervical spine instability, it underscores the useful-ness of the Minerva as a cervical immobilization device. Becausea large proportion of pediatric cervical spine injuries occur be-tween the occiput and C2, the Minerva device may not be idealfor many pediatric cervical spine injuries.

In summary, the physical properties of young skin, cranialthickness, and small body size likely contribute to the highercomplication rate among children who require traction orlong-term cervical spinal immobilization, as compared withadults. The literature includes descriptions of options avail-able for reduction and immobilization of cervical spine inju-ries in children, but it does not provide evidence for a singlebest method.

Surgical treatment

There are no reports in the literature that address the topicof early versus late surgical decompression after acute pedi-atric cervical SCI. Pediatric spinal injuries account for only 5%of all vertebral column injuries. Recent reports that describethe management of pediatric spinal injuries have been offeredby Turgut et al. (61), Finch and Barnes (16), and Elaraky et al.(12). These authors managed pediatric spinal injuries opera-tively in 17, 25, and 30% of patients, respectively. The reportby Elaraky et al. (12) in 2000 suggests that operative treatmentof pediatric cervical spine injuries is being used more fre-quently than in the past. Specific details of the operativemanagement, including timing of intervention, the approach(anterior versus posterior), and the method of internal fixationas an adjunct to fusion, are scarce in the literature. Finch andBarnes (16) used primary operative stabilization in most chil-dren they managed with ligamentous injuries of the cervicalspine. They stated that although external immobilization mayhave resulted in ligamentous healing, they elected to inter-nally fixate and fuse such injuries. They based their approachon two cases of ligamentous injuries of the cervical spine that



they managed with external immobilization, which failed toheal and later required operative fusion. Shacked et al. (55)described six children aged 3 years to 14 years who hadcervical spine injuries that they treated surgically via an an-terior approach. They reported successful fusion with goodalignment and normal cervical spine growth in follow-up forall six children. The procedure varied (i.e., total or partialcorpectomy versus discectomy only) depending on the pa-thology. All underwent autograft fusion without instrumen-tation. The authors described severe hyperflexion injury withfracture and avulsion of the vertebral body, fracture-dislocation with disruption of the posterior elements and disc,and major anatomic deformity of the cervical spine with cordcompression as indications for an anterior approach.

Pennecot et al. (43) described 16 children with ligamentousinjuries of the cervical spine. They managed minor ligamentousinjuries (atlantodens interval of 5–7 mm, or interspinous widen-ing without dislocation or neurological deficit) with reductionand immobilization. Of 11 children with injuries below C2, 8required operative treatment with fusion via a posterior ap-proach. They used interspinous wiring techniques in younger(preschool-aged) children and posterior plates and screws inolder children as adjuncts to fusion. All had successful fusion atlast follow-up. All children were immobilized in a plaster or halocast postoperatively. Similarly, Koop et al. (30) described 13children with acute cervical spine injuries who required poste-rior arthrodesis and halo immobilization. They reported success-ful fusion in 12 patients. The single failure was associated withallograft fusion substrate. All other children had autologousgrafts. Internal fixation with wire was used in only two children.Halo immobilization was used for an average of 150 days. Theyreduced the length of postoperative halo immobilization to 100days in their most recent cases. They commented that carefultechnique allowed successful posterior fusion in children withminimal complications. Schwarz et al. (54) described 10 childrenwith traumatic cervical kyphosis. Two children who underwentanterior reconstruction with fusion had successful deformityreduction. All others managed with either external immobiliza-tion with or without traction (n � 7) or posterior fusion (n � 1)had either progression of the posttraumatic deformity or a stableunreduced kyphotic angulation.

In summary, pediatric spinal injuries are relatively infre-quent. Most pediatric spinal injuries are managed withoutoperations. Selection criteria for operative intervention in chil-dren with cervical spine injuries are difficult to glean from theliterature. Anatomic reduction of deformity, stabilization ofunstable injuries and decompression of the spinal cord, andisolated ligamentous injuries associated with deformity areindications for surgical treatment cited by various authors (16,30, 33, 43, 54, 55, 64). These reports provide Class III evidence.

SUMMARY

The available medical evidence does not allow the generationof diagnostic or treatment standards for managing pediatricpatients with cervical spine or cervical SCI. Only diagnosticguidelines and options and treatment options are supported bythis evidence. The literature suggests that obtaining neutral cer-

vical spine alignment in a child may be difficult when standardbackboards are used. The determination that a child does nothave a cervical spine injury on clinical grounds alone is sup-ported by Class II and Class III evidence. When the child is alertand communicative and is without neurological deficit, necktenderness, painful distracting injury, or intoxication, cervicalx-rays are not necessary to exclude cervical spinal injury. Whencervical spine x-rays are used to verify or rule out a cervicalspinal injury in children younger than 9 years, only lateral andanteroposterior cervical spine views need be obtained. The tra-ditional three-view x-ray assessment may increase the sensitivityof plain spine x-rays in children aged 9 years and older. Mostpediatric cervical spine injuries can be effectively treated withoutoperating. The most effective immobilization seems to be accom-plished with either halo devices or Minerva jackets. Halo immo-bilization is associated with acceptable but considerable minormorbidity in children, typically pin-site infection and pin loos-ening. The only specific pediatric cervical spine injury for whichmedical evidence supports a particular treatment paradigm is anodontoid injury in children younger than 7 years. These childrenare effectively treated with closed reduction and immobilization.Primarily ligamentous injuries of the cervical spine in childrenmay heal with external immobilization alone but are associatedwith a relatively high rate of progressive deformity when treatedwithout operating. Pharmacological therapy and intensive careunit management schemes for children with SCI have not beendescribed in the literature.


Prospective epidemiological data may be the best source ofinformation that could lead to methods of prevention byidentifying the more common mechanisms of spinal injury inchildren. Future studies involving pediatric cervical spineinjury patients should be multi-institutional because of theinfrequency of these injuries treated at any single institution.Further defining the indications and methods for cervicalspine clearance in young children (younger than 9 yr) withprospective gathering of data would be a valuable addition tothe literature. The role of flexion/extension x-rays is poorlydefined in the literature, and a prospective evaluation of theirsensitivity and specificity for spinal column injury in specificclinical scenarios would be a valuable addition to the litera-ture. The incidence and clinical significance of complicationsof cervical spine injuries in children, such as syringomyeliaand vertebral artery injury, are unknown and could be stud-ied by prospectively gathering data in a multi-institutionalsetting.

More common injuries, such as odontoid injuries, could bestudied prospectively in a randomized fashion (e.g., closedreduction and immobilization versus anterior screw fixation),although it would be difficult from technical and feasibilitystandpoints. Prospectively collected data could also providethe basis for case-control or other comparative studies togenerate Class II evidence.




REFERENCES

1. Baum JA, Hanley EN Jr, Pullekines J: Comparison of halo com-plications in adults and children. Spine 14:251–252, 1989.

2. Benzel EC, Hadden TA, Saulsbery CM: A comparison of theMinerva and halo jackets for stabilization of the cervical spine.J Neurosurg 70:411–414, 1989.

3. Berne JD, Velmahos GC, El-Tawil Q, Demetriades D, Asensio JA,Murray JA, Cornwell EE, Belzberg H, Berne TV: Value of com-plete cervical helical computed tomographic scanning in identi-fying cervical spine injury in the unevaluable blunt traumapatient with multiple injuries: A prospective study. J Trauma47:896–902, 1999.

3a. Birney TJ, Hanley EN: Traumatic cervical spine injuries in child-hood and adolescence. Spine 14:1277–1282, 1989.

4. Bohn D, Armstrong D, Becker L, Humphreys R: Cervical spineinjuries in children. J Trauma 30:463–466, 1990.


6. Buhs C, Cullen M, Klein M, Farmer D: The pediatric traumaC-spine: Is the “odontoid” view necessary? J Pediatr Surg 35:994–997, 2000.

7. Cattell HS, Filtzer DL: Pseudosubluxation and other normalvariations in the cervical spine in children: A study of onehundred and sixty children. J Bone Joint Surg Am 47A:1295–1309, 1965.


9. Davis PC, Reisner A, Hudgins PA, Davis WE, O’Brien MS:Spinal injuries in children: Role of MR. AJNR Am J Neuroradiol14:607–617, 1993.

10. Dickman CA, Rekate HL, Sonntag VKH, Zabramski JM: Pediat-ric spinal trauma: Vertebral column and spinal cord injuries inchildren. Pediatr Neurosci 15:237–256, 1989.

11. Dormans JP, Criscitiello AA, Drummond DS, Davidson RS:Complications in children managed with immobilization in ahalo vest. J Bone Joint Surg Am 77A:1370–1373, 1995.

11a. Dwek JR, Chung CB: Radiography of cervical spine injury inchildren: Are flexion-extension radiographs useful for acutetrauma? AJR Am J Roentgenol 174:1617–1619, 2000.

12. Eleraky MA, Theodore N, Adams M, Rekate HL, Sonntag VKH:Pediatric cervical spine injuries: Report of 102 cases and reviewof the literature. J Neurosurg 92[Suppl 1]:12–17, 2000.

13. El-Khoury GY, Clark CR, Gravett AW: Acute traumatic rotatoryatlanto-axial dislocation in children: A report of three cases.J Bone Joint Surg Am 66A:774–777, 1984.

14. Evans DL, Bethem D: Cervical spine injuries in children.J Pediatr Orthop 9:563–568, 1989.

15. Fielding JW, Hawkins RJ: Atlanto-axial rotary fixation: Fixedrotatory subluxation of the atlanto-axial joint. J Bone Joint SurgAm 59A:37–44, 1977.

16. Finch GD, Barnes MJ: Major cervical spine injuries in childrenand adolescents. J Pediatr Orthop 18:811–814, 1998.

17. Fotter R, Sorantin E, Schneider U, Ranner G, Fast C, Schober P:Ultrasound diagnosis of birth-related spinal cord trauma: Neo-natal diagnosis and follow-up and correlation with MRI. PediatrRadiol 24:241–244, 1994.

18. Gaskill SJ, Marlin AE: Custom fitted thermoplastic Minervajackets in the treatment of cervical spine instability in preschoolage children. Pediatr Neurosurg 16:35–39, 1990.

19. Gaufin LM, Goodman SJ: Cervical spine injuries in infants.J Neurosurg 42:179–185, 1975.

20. Givens TG, Polley KA, Smith GF, Hardin WD Jr: Pediatric cer-vical spine injury: A three-year experience. J Trauma 41:310–314,1996.

21. Godard J, Hadji M, Raul JS: Odontoid fractures in the child withneurological injury: Direct anterior osteosynthesis with a cortico-spongious screw and literature review. Childs Nerv Syst 13:105–107, 1997.

22. Griffiths SC: Fracture of the odontoid process in children.J Pediatr Surg 7:680–683, 1972.

23. Grogaard B, Dullerud R, Magnaes B: Acute torticollis in childrendue to atlanto-axial rotary fixation. Arch Orthop Trauma Surg112:185–188, 1993.

24. Hadley MN, Zabramski JM, Browner CM, Rekate H, SonntagVKH: Pediatric spinal trauma: Review of 122 cases of spinal cordand vertebral column injuries. J Neurosurg 68:18–24, 1988.

25. Hamilton MG, Myles ST: Pediatric spinal injury: Review of 174hospital admissions. J Neurosurg 77:700–704, 1992.

26. Herzenberg JE, Hensinger RN, Dedrick DK, Phillips WA: Emer-gency transport and positioning of young children who have aninjury of the cervical spine: The standard backboard may bedangerous. J Bone Joint Surg Am 71A:15–22, 1989.

27. Huerta C, Griffith R, Joyce SM: Cervical spine stabilization inpediatric patients: Evaluation of current techniques. Ann EmergMed 16:1121–1126, 1987.

28. Kawabe N, Hirotani H, Tanaka O: Pathomechanism of atlanto-axial rotary fixation in children. J Pediatr Orthop 9:569–574,1989.

29. Keiper MD, Zimmerman RA, Bilaniuk LT: MRI in the assess-ment of the supportive soft tissues of the cervical spine in acutetrauma in children. Neuroradiology 40:359–363, 1998.

30. Koop SE, Winter RB, Lonstein JE: The surgical treatment ofinstability of the upper part of the cervical spine in children andadolescents. J Bone Joint Surg Am 66A:403–411, 1984.

31. Kowalski HM, Cohen WA, Cooper P, Wisoff JH: Pitfalls in theCT diagnosis of atlantoaxial rotary subluxation. AJR Am JRoentgenol 149:595–600, 1987.

32. Laham JL, Cotcamp DH, Gibbons PA, Kahana MD, Crone KR:Isolated head injuries versus multiple trauma in pediatric pa-tients: Do the same indications for cervical spine evaluationapply? Pediatr Neurosurg 21:221–226, 1994.

33. Lui TN, Lee ST, Wong CW, Yeh YS, Tzaan WC, Chen TY, HungSY: C1–C2 fracture-dislocations in children and adolescents.J Trauma 40:408–411, 1996.

34. MacKinnon JA, Perlman M, Kirpalani H, Rehan V, Sauve R,Kovacs L: Spinal cord injury at birth: Diagnostic and prognosticdata in twenty-two patients. J Pediatr 122:431–437, 1993.

35. Mandabach M, Ruge JR, Hahn YS, McLone DG: Pediatric axisfractures: Early halo immobilization, management and outcome.Pediatr Neurosurg 19:225–232, 1993.

36. Marks DS, Roberts P, Wilton PJ, Burns LA, Thompson AG: Ahalo jacket for stabilisation of the paediatric cervical spine. ArchOrthop Trauma Surg 112:134–135, 1993.

37. Menticoglou SM, Perlman M, Manning FA: High cervical spinalcord injury in neonates delivered with forceps: Report of 15cases. Obstet Gynecol 86:589–594, 1995.

38. Mubarak SJ, Camp JF, Vuletich W, Wenger DR, Garfin SR: Haloapplication in the infant. J Pediatr Orthop 9:612–614, 1989.




40. Odent T, Langlais J, Glorion C, Kassis B, Bataille J, Pouliquen JC:Fractures of the odontoid process: A report of 15 cases in chil-dren younger than six years. J Pediatr Orthop 19:51–54, 1999.

41. Osenbach RK, Menezes AH: Pediatric spinal cord and vertebralcolumn injury. Neurosurgery 30:385–390, 1992.

42. Pang D, Hanley EN: Special problems of spinal stabilization inchildren, in Cooper PR (ed): Management of Posttraumatic SpinalInstability (Neurosurgical Topics No. 3). Park Ridge, AANS, 1990,pp 181–206.

43. Pennecot GF, Leonard P, Peyrot Des Gachons S, Hardy JR,Pouliquen JC: Traumatic ligamentous instability of the cervicalspine in children. J Pediatr Orthop 4:339–345, 1984.

44. Phillips WA, Hensinger RN: The management of rotatoryatlanto-axial subluxation in children. J Bone Joint Surg Am71A:664–668, 1989.

45. Ralston ME, Chung K, Barnes PD, Emans JB, Schutzman SA:Role of flexion-extension radiographs in blunt pediatric cervicalspine injury. Acad Emerg Med 8:237–245, 2001.

46. Rathbone D, Johnson G, Letts M: Spinal cord concussion inpediatric athletes. J Pediatr Orthop 12:616–620, 1992.

47. Reinges MH, Mayfrank L, Rohde V, Spetzger U, Gilsbach JM:Surgically treated traumatic synchondrotic disruption of theodontoid process in a 15-month-old girl. Childs Nerv Syst 14:85–87, 1998.

48. Rossitch E Jr, Oakes WJ: Perinatal spinal cord injury. PediatrNeurosurg 18:149–152, 1992.

49. Ruge JR, Sinson GP, McLone DG, Cerullo LJ: Pediatric spinalinjury: The very young. J Neurosurg 68:25–30, 1988.

50. Scarrow AM, Levy EI, Resnick DK, Adelson PD, Sclabassi RJ: Cervicalspine evaluation in obtunded or comatose pediatric trauma patients: Apilot study. Pediatr Neurosurg 30:169–175, 1999.


52. Schleehauf K, Ross SE, Civil ID, Schwab CW: Computed tomog-raphy in the initial evaluation of the cervical spine. Ann EmergMed 18:815–817, 1989.

53. Schwarz N: The fate of missed atlanto-axial rotatory subluxationin children. Arch Orthop Trauma Surg 117:288–289, 1998.

54. Schwarz N, Genelin F, Schwarz AF: Post-traumatic cervical ky-phosis in children cannot be prevented by non-operative meth-ods. Injury 25:173–175, 1994.

55. Shacked I, Ram Z, Hadani M: The anterior cervical approach fortraumatic injuries to the cervical spine in children. Clin Orthop292:144–150, 1993.

56. Shaw M, Burnett H, Wilson A, Chan O: Pseudosubluxation of C2on C3 in polytraumatized children: Prevalence and significance.Clin Radiol 54:377–380, 1999.

57. Sherk HH, Nicholson JT, Chung SM: Fractures of the odontoid processin young children. J Bone Joint Surg Am 60A:921–924, 1978.

58. Subach BR, McLaughlin MR, Albright AL, Pollack IF: Currentmanagement of pediatric atlantoaxial rotatory subluxation.Spine 23:2174–2179, 1998.

59. Swischuk LE, John SD, Hendrick EP: Is the open-mouth odon-toid view necessary in children under 5 years? Pediatr Radiol30:186–189, 2000.

60. Treloar DJ, Nypaver M: Angulation of the pediatric cervicalspine with and without cervical collar. Pediatr Emerg Care13:5–8, 1997.

61. Turgut M, Akpinar G, Akalan N, Ozcan OE: Spinal injuries inthe pediatric age group: A review of 82 cases of spinal cord andvertebral column injuries. Eur Spine J 5:148–152, 1996.

62. Viccellio P, Simon H, Pressman BD, Shah MN, Mower WR,Hoffman JR: A prospective multicenter study of cervical spineinjury in children. Pediatrics 108:E20, 2001.

63. Vogel LC: Unique management needs of pediatric spinal cordinjury patients: Etiology and pathophysiology. J Spinal CordMed 20:10–13, 1997.

64. Wang J, Vokshoor A, Kim S, Elton S, Kosnik E, Bartkowski H:Pediatric atlantoaxial instability: Management with screw fixa-tion. Pediatr Neurosurg 30:70–78, 1999.

Plate from Gautier D, Duverney M: Essai D’anatomie, en Tab-leaux Imprimés. . . . Paris, 1745. Courtesy, Dr. Irwin J. Pincus,Los Angeles, California.


CHAPTER 13

Spinal Cord Injury without Radiographic Abnormality

RECOMMENDATIONSDIAGNOSIS:Standards: There is insufficient evidence to support diagnostic standards.Guidelines: There is insufficient evidence to support diagnostic guidelines.Options:• Plain spinal x-rays of the region of injury and computed tomographic scanning with attention to the

suspected level of neurological injury to exclude occult fractures are recommended.• Magnetic resonance imaging of the region of suspected neurological injury may provide useful diagnostic

information.• Plain x-rays of the entire spinal column may be considered.• Neither spinal angiography nor myelography is recommended in the evaluation of patients with spinal cord

injury without radiographic abnormality.

TREATMENT:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options:• External immobilization is recommended until spinal stability is confirmed by flexion/extension x-rays.• External immobilization of the spinal segment of injury for up to 12 weeks may be considered.• Avoidance of “high-risk” activities for up to 6 months after spinal cord injury without radiographic

abnormality may be considered.

PROGNOSIS:Standards: There is insufficient evidence to support prognostic standards.Guidelines: There is insufficient evidence to support prognostic guidelines.Options: Magnetic resonance imaging of the region of neurological injury may provide useful prognostic

information about neurological outcome after spinal cord injury without radiographic abnormality.

RATIONALE

Diagnosis

Pang and Wilberger defined the term SCIWORA (spinalcord injury without radiographic abnormality) in 1982 asobjective signs of myelopathy as a result of trauma with no

evidence of fracture or ligamentous instability on plain spinex-rays and tomography (12). In their 1982 article, the authorscautioned that if the early warning signs of transient symptomscould be recognized and promptly acted upon before the onsetof neurological signs, the tragic fate of some of these childrenmight be duly averted (12). Hamilton and Myles (8), Osenbachand Menezes (9), and Pang and Wilberger (12) documented thedelayed onset of SCIWORA in children as late as 4 days afterinjury. Therefore, a concern is whether a child with a normalneurological examination, but with a history of transient neuro-logical symptoms or persisting subjective neurological symp-toms referable to traumatic myelopathy, should be assigned the

diagnosis of SCIWORA and managed accordingly, despite theabsence of “objective signs of myelopathy.”

Pang and Pollack (11) recommended obtaining a computedtomographic scan focused at the neurological level of injury toexclude an occult fracture in a child with a neurological deficitreferable to the spinal cord, but without abnormalities onplain x-rays of the spine. In addition, dynamic flexion/extension x-rays or fluoroscopy has been advocated to ex-clude pathological intersegmental motion consistent with lig-amentous injury without fracture. If paraspinous musclespasm, pain, or uncooperation prevents dynamic studies, theyrecommended external immobilization until the child can flexand extend the spine for dynamic x-ray assessment. The find-ing of fracture, subluxation, or abnormal intersegmental mo-tion at the level of neurological injury excludes SCIWORA asa diagnosis. In the initial report by Pang and Wilberger (12), 1of 24 children exhibited pathological motion on initial dy-namic x-rays. By their own definition of SCIWORA, this one


child would not be diagnosed with SCIWORA because theinitial flexion and extension x-rays were abnormal. Althoughconcern exists for the development of pathological interseg-mental motion in children with SCIWORA after normalflexion/extension studies, there has not been documentationof such instability ever developing.

Magnetic resonance imaging (MRI) findings in childrenwith SCIWORA have spanned the spectrum from normal tocomplete cord disruption, along with evidence of ligamentousand disc injury in some (3, 6). Possible roles for MRI ofchildren with SCIWORA include 1) excluding compressivelesions of the cord or roots or ligamentous disruption thatmight warrant surgical intervention, 2) guiding treatmentregarding length of external immobilization, and/or 3) deter-mining when to allow patients to return to full activity.

Treatment

Because there exists no subluxation or malalignment inSCIWORA, the mainstay of treatment has been immobiliza-tion and avoidance of activity that may either lead to exacer-bation of the present injury or increase the potential for re-current injury. Medical management issues, such as bloodpressure support and pharmacological therapy, are of concernto this population as well and have been addressed in otherguidelines. (Of note, the often-cited prospective studies ofpharmacological therapy in the treatment of acute spinal cordinjury did not include children younger than 13 yr [1].)

Pang and Pollack (11) have recommended 12 weeks of exter-nal immobilization to allow adequate time for healing of thepresumed ligamentous strain/injury and to prevent exacerba-tion of the myelopathy. It is unclear, however, what role immo-bilization plays in this population once dynamic x-rays havedisplayed no instability. The length of and even the need forimmobilization remain debatable, given the current literature. Ifthe incidence of delayed pathological intersegmental motion inchildren with SCIWORA, who have been proven to have normaldynamic x-rays, approaches zero, then the role of spinal immo-bilization for SCIWORA patients needs to be considered in lightof the available literature. If physiological motion (normal) of thespinal column can potentiate SCIWORA in these patients whenthere is no malalignment, subluxation, or lesion causing cordcompression, then immobilization is warranted.

Prognosis

SCIWORA has been shown to be associated with a highincidence of complete neurological injuries, particularly inchildren younger than 9 years. Hadley et al. (7) reported fourcomplete injuries in six children younger than 10 years withSCIWORA. The regions of complete injury tend to be cervicaland upper thoracic. Pang and Wilberger (12) found the pre-senting neurological examination to relate strongly to out-come. There are some data to suggest that MRI abnormalities(or lack of abnormalities) of the cord may be more predictive

of outcome than presenting neurological status (3, 6). Becauseno child has been documented to develop spinal instabilityafter the diagnosis of SCIWORA and the patient has, bydefinition, normal flexion/extension x-rays, there has beenlittle impetus to define predictors of instability. On the otherhand, children have been documented to have recurrentSCIWORA (13), and predictors of a “high-risk” subgroup ofchildren with SCIWORA for recurrent injury may exist.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 was un-dertaken. The medical subject headings “pediatric” and “SCI-WORA” combined with “spinal cord injury” yielded approxi-mately 145 citations. Non-English language citations weredeleted. The articles were reviewed for those that identifiedchildren who had incurred a SCIWORA. Those articles thatdescribed the clinical aspects and management of children withSCIWORA were used to generate these guidelines. Case reportswere excluded from review. Of the 15 articles meeting the selec-tion criteria, none were Class I or Class II studies. All were caseseries representing Class III data. The articles are summarized inTable 13.1. In addition, articles germane to the topic but notmeeting criteria for inclusion in the Evidentiary Tables, such asgeneral review articles, are referenced in the Scientific Founda-tion section and included in the references.


One concern is whether the child with a normal neurologicalexamination and either a history of transient neurological deficit(e.g., paraparesis or quadriparesis) or persisting subjectivesymptoms (e.g., numbness or dysesthesias) would be a candi-date for the diagnosis of SCIWORA. Pang and Wilberger (12)described 13 of their 24 children to have a “latent” period from30 minutes to 4 days (mean, 1.2 d) before the onset of objectivesensorimotor deficits. All 13 of these children had transient sub-jective complaints at the time of their initial trauma that clearedwithin 1 hour before their neurological decline. Those who de-veloped mild neurological deficits often improved to normal,whereas those who developed severe neurological deficits wereoften left with permanent neurological dysfunction. Hamiltonand Myles (8), Osenbach and Menezes (9), and Pang and Pollack(11) also reported a 22, 23, and 27% incidence, respectively, ofdelayed onset of myelopathy within their series of children withSCIWORA. Dickman et al. (4), Eleraky et al. (5), and Hadley et al.(7) described no child having a latent period of neurologicalnormalcy after injury. The observations of delayed deteriorationby different investigators, however, raise the concern that anychild presenting with a history of transient neurological deficit orsymptoms after an appropriate mechanism of injury may beconsidered for the diagnosis of SCIWORA, despite the absenceof objective evidence of myelopathy on the initial neurologicalexamination.

Spinal Cord Injury without Radiographic Abnormality


TABLE 13.1. Summary of Reports on Spinal Cord Injury without Radiographic Abnormalitya


ClassResults

Eleraky et al., 2000 (5) Retrospective review of 102 children withcervical spinal injuries. Young (0–9 yr) childrencompared with older children. MRI performedin 12/18 children with SCIWORA.

III SCIWORA in 18%. MRI findings did not altermanagement (external immobilization).

Turgut et al., 1996 (18) Retrospective review of 11/82 children withspinal injuries with SCIWORA.

III SCIWORA represented 13% of spinal injuries inchildren.

Grabb and Pang, 1994 (6) Retrospective review of 7 children withSCIWORA underwent MRI. Neurological statusat presentation and follow-up was correlated toMRI findings.

III No compressive lesions found. Prognosis correlated withMRI findings. Hematomyelia involving �50% of corddiameter was associated with permanent severe deficits.Lesser degrees of hematomyelia and edema only wereassociated with incomplete recovery, and normal MRIpredicted full recovery.

Davis et al., 1993 (3) Retrospective review of 15 children with spinalcord injury underwent MRI 12 h to 2 mo afterinjury. 7 children with SCIWORA.

III MRI correlated with prognosis. Hemorrhagic cordcontusions and cord “infarction” were associated withpermanent deficits. No compressive lesions in SCIWORAcases. Normal MRI was associated with no myelopathy.

Hamilton and Myles, 1992 (8) Retrospective review of 174 pediatric spinalinjuries during 14-yr period.

III SCIWORA represented 13% of spinal injuries. Ofchildren aged 0–9 yr with spinal injuries, 42% hadSCIWORA, whereas of children aged 10–14 yr, only14% had SCIWORA.

Osenbach and Menezes, 1992 (10) Retrospective review of 34/179 children withspinal injuries with SCIWORA.

III SCIWORA represented 19% of spinal injuries inchildren. Younger children (�9 yr) had higher incidenceof SCIWORA.

Rathbone et al., 1992 (15) Retrospective review of 12 children withpresumed spinal cord concussion duringathletics was investigated for the presence ofcervical stenosis.

III 3 had a Torg ratio �0.8, and 4 had a canalanteroposterior diameter �13.4 mm. MRI was not usedto evaluate for stenosis.

Rossitch and Oakes, 1992 (16) Retrospective review of 5 neonates withperinatal spinal cord injury. 4 of the 5 had noabnormality on static spinal x-rays. No flexion/extension views reported. Myelograms wereunrevealing.

III Children with perinatal spinal cord injury often havenormal x-rays. The neonates are often initiallymisdiagnosed. Respiratory insufficiency and hypotoniaare common signs.

Dickman et al., 1991 (4) Retrospective review of 26 children withSCIWORA over 19-yr period. Clinical andepidemiological features were analyzed.

III SCIWORA represents 16% of spinal injuries in children.Motor vehicle accident was most common mechanism. 7children had MRI. 5 were normal studies, 2 showed cordsignal abnormalities. Younger children tended to havemore severe injuries.

Osenbach and Menezes, 1989 (9) Retrospective review of 31 children withSCIWORA.

III 26 cervical and 5 thoracic injuries. Complete cord injuryin 12. Delayed onset of deficits in 7. No surgical lesionsfound by MRI or CT-myelography. Spinal angiogramsdone in 4 thoracic cases were normal. No delayedinstability at follow-up.

Pang and Pollack, 1989 (11) Retrospective review of 55 children withSCIWORA (43 cervical, 12 thoracic). Clinicalprofiles reported to illustrate syndrome.

III 22 “severe” injuries; 33 “mild” injuries. Age �8 yrassociated with more severe injuries. 8 cases of recurrentinjury from 3 d to 10 wk after initial injury. No recurrentinjuries with 12 wk of Guilford brace.

Hadley et al., 1988 (7) Retrospective review of 122 children withspinal injuries. Young (0–9 yr) compared witholder children.

III 17% with SCIWORA. Higher incidence of SCIWORA in0–9 yr olds versus 10–16 yr olds. 5 studied with MRI, noabnormalities detected.

Pollack et al., 1988 (13) Retrospective review of 8 children withrecurrent SCIWORA compared with 12children treated with longer immobilization.

III Recurrent SCIWORA occurred from 3 to 10 wk afterinitial injury. Recurrent injuries were more severe. Norecurrent injuries with 12 wk of Guilford brace.

Ruge et al., 1988 (17) Retrospective review comparing 0–3 yr olds to4–12 yr olds with spinal injury.

III n � 47, 21% with SCIWORA.

Pang and Wilberger, 1982 (12) Retrospective review of 24 children withSCIWORA.

III 1 child with instability on flexion/extension at 1 wk.

a SCIWORA, spinal cord injury without radiographic abnormality; MRI, magnetic resonance imaging; CT, computed tomography.



Pang and Wilberger (12) had 1 child of 24 demonstrate whatwas considered to be pathological intersegmental motion onflexion/extension x-rays 1 week after injury, after resolutionof neck pain and paraspinous muscle spasm. By definition,this child would not be considered to have had SCIWORAbecause the initial flexion/extension x-rays were abnormal.This child was treated successfully with external immobiliza-tion alone for 8 weeks. No child with SCIWORA has beendocumented in the literature to have had normal dynamicx-rays and subsequently develop intersegmental instability.

In 1994, a series of seven children with SCIWORA weredemonstrated to have ligamentous, disc, and intramedullaryabnormalities identified by MRI (6). Soft tissue findings con-sisted of anterior longitudinal ligament disruption in associ-ation with a hyperextension injury, posterior longitudinalligament disruption, and a noncompressive C2–C3 disc her-niation associated with lateral flexion, and one case of C6–C7disc abnormality associated with hyperflexion. Intramedul-lary findings included cord transection and rostral stumphemorrhage, severe hematomyelia, a minor intramedullaryhemorrhage, and edema without hemorrhage. Davis et al. (3)described seven children with SCIWORA who were exam-ined with MRI. They described no abnormalities of muscles,ligaments, or discs but did correlate the presence of intramed-ullary hemorrhage or cord “infarction” with permanent neu-rological deficit. The lack of intramedullary findings corre-lated with a normal neurological outcome. Dickman et al. (4)commented on seven children with SCIWORA who wereexamined with MRI. Five of the seven studies revealed noabnormality, and two studies documented intramedullarysignal changes. Osenbach and Menezes (9) commented intheir series of childhood SCIWORA that MRI and computedtomography-myelography performed on their patients didnot demonstrate a single compressive lesion. In addition, theyperformed spinal arteriograms in four of five children withthoracic SCIWORA and found no angiographic abnormalities.Rossitch and Oakes (16) performed myelograms on neonateswith SCIWORA from birth injury and found no abnormalitiesthat changed their management. Hadley et al. (7), before 1988,obtained MRI studies on five children with SCIWORA andidentified no abnormalities. These results need to be viewedin the context of the technology available at the time of study.

There has been no report of any situation in which the careof a child with SCIWORA has been altered by the results ofMRI and/or myelography imaging studies. No child withMRI-documented ligamentous injury and SCIWORA has de-veloped spinal instability, early or delayed. To date, there hasbeen no correlation between the ligamentous findings on MRIin SCIWORA patients and subsequent spinal instability. Theappearance of the spinal cord on MRI does provide prognos-tic information regarding ultimate neurological outcome.

Hadley et al. (7) noted a 16% incidence of multiple noncon-tiguous injuries of the spine or spinal cord in children withany type of spinal column or spinal cord injury. Ruge et al.(17) had a similar incidence (17%) of multiple levels of spinalinjury in children. Although neither of these studies dealtwith an isolated population of children with SCIWORA, theydo provide consistent observations that one in six children

with spinal trauma will have multiple levels of injury. Pangand Wilberger (12) reported 1 of 24 children with a second-level injury (L2 Chance fracture) who had a T6 neural injury(SCIWORA), but they did not obtain complete spine x-rays onevery child. Because of these observations, one should con-sider x-rays of the entire spinal column when any traumaticspinal injury, SCIWORA or otherwise, is identified in a child.

In the initial series of children with SCIWORA reported byPang and Wilberger (12), treatment routinely consisted of 4weeks of external immobilization with a “cervical collar” forcervical injuries. In cases of thoracic injury, if subsequent plainx-rays showed no abnormality after 1 week of bed rest, the childwas mobilized without a brace. In a later report, in 1989, Pangand Pollack (11) recommended 12 weeks of external immobili-zation for SCIWORA patients to allow for healing of the pre-sumed ligamentous strain/injury and to prevent exacerbation ofthe myelopathy. They also advocated external immobilizationfor this timeframe to prevent recurrent injury during the healingphase. They reported seven children who sustained recurrentSCIWORA of greater severity with lesser degrees of force whenexternal immobilization was removed before 12 weeks or theywere allowed to participate in activities against physician in-structions within 6 months of the initial injury. For these reasons,the authors recommend 12 weeks of external immobilizationand 12 additional weeks of activity restriction after SCIWORA.

Dickman et al. (4), Eleraky et al. (5), and Hadley et al. (7)reported no neurological deterioration in any patient withSCIWORA after admission or discharge. None of these re-ports described the length of time children with SCIWORAwere immobilized. It has not been routine among treatingphysicians to prescribe 12 weeks of immobilization for chil-dren with SCIWORA (2). Although a single report by Pollacket al. (13) describes recurrent SCIWORA within 12 weeks ofthe original injury, this has not been validated by other ob-servations. Because MRI evaluation was not available forthose with recurrent injury, it is not known whether certainMRI characteristics (e.g., ligamentous disruption) could pre-dict an “at risk” group for recurrent SCIWORA.

Although Pang and Wilberger (12) reported that, in theirseries, neurological outcome correlated with the presentingneurological status, others have shown that the MRI appear-ance of the spinal cord is predictive of neurological outcomein children with SCIWORA (3, 6). Absence of signal changewithin the cord is associated with an excellent outcome. Sig-nal change consistent with edema or microhemorrhages, butnot frank hematomyelia, is associated with significant im-provement of neurological function over time. The presenceof frank hematomyelia or cord disruption is associated with asevere, permanent neurological injury (3, 6). The correlationof neurological outcome with spinal cord MRI findings inSCIWORA remains consistent with the findings in muchlarger numbers of patients with spinal cord injury (non-SCIWORA) who have been studied with MRI (14).

SUMMARY

Children presenting with a history of transient neurologicalsigns or symptoms referable to traumatic myelopathy, despite

Spinal Cord Injury without Radiographic Abnormality S103


having normal x-rays and the absence of objective evidence ofmyelopathy, may develop SCIWORA in a delayed fashion.No child with SCIWORA has developed pathological inter-segmental motion with instability after demonstrating normalflexion/extension x-rays. MRI has not identified any lesion ina child with SCIWORA for which the management schemewould be changed by the results of the MRI examination.Similarly, no child with MRI-documented ligamentous injuryand SCIWORA has developed evidence of spinal instability.Hard collar immobilization for patients with cervical levelSCIWORA for 12 weeks and avoidance of activities that en-courage flexion/extension of the neck for an additional 12weeks has not been associated with recurrent injury. Thespinal cord findings on MRI provide prognostic informationregarding long-term neurological outcome in patients withSCIWORA. Myelography and angiography have no definedrole in the evaluation of children with SCIWORA.


The treatment end points of spinal immobilization and activ-ity restriction for patients with SCIWORA have been arbitrarilychosen. MRI may be helpful to guide the length of time a childis immobilized and has activities restricted. The absence of evi-dence of ligamentous injury on MRI may indicate that there is noneed for external immobilization or activity restriction. It hasbeen observed that recurrent SCIWORA can occur despite a lackof evidence of spinal instability. An investigation that obtainedMRI on all children with SCIWORA and followed their clinicalstatus longitudinally might highlight the usefulness of MRI inthe management of children who go on to develop recurrentSCIWORA. The literature provides little guidance as to the like-lihood for subsequent catastrophic injury in children presentingwith SCIWORA of any severity who are found to have a preex-isting spinal or neurological abnormality, such as congenitalcervical stenosis or a Chiari malformation (15). Longitudinalclinical follow-up of SCIWORA patients of this type may pro-vide information to appropriately counsel these children. Thereare no data to elucidate the role of age in the success or failure ofvarious treatments for this condition. A longitudinal study of apatient population of reasonable size could be undertaken.

Serious attempts to address these topics cannot be forthcom-ing from a single institution or investigator because of the rela-tively small numbers of children who sustain SCIWORA annu-ally (10, 18). A multi-institutional, protocol-directed study ofSCIWORA patients may provide answers to some of the ques-tions that accompany this unique spinal cord injury subtype.


REFERENCES


2. Bruce D: Spinal cord injury without radiographic abnormality inchildren. Pediatr Neurosci 15:175, 1989 (comment).

3. Davis PC, Reisner A, Hudgins PA, Davis WE, O’Brien MS: Spinalinjuries in children: Role of MR. AJNR Am J Neuroradiol 14:607–617, 1993.

4. Dickman CA, Zabramski JM, Hadley MN, Rekate HL, SonntagVKH: Pediatric spinal cord injury without radiographic abnor-malities. J Spinal Disord 4:296–305, 1991.

5. Eleraky MA, Theodore N, Adams M, Rekate HL, Sonntag VKH:Pediatric cervical spine injuries: Report of 102 cases and review ofthe literature. J Neurosurg 92[Suppl 1]:12–17, 2000.

6. Grabb PA, Pang D: Magnetic resonance imaging in the evaluationof spinal cord injury without radiographic abnormality in chil-dren. Neurosurgery 35:406–414, 1994.

7. Hadley MN, Zabramski JM, Browner CM, Rekate H, SonntagVKH: Pediatric spinal trauma. J Neurosurg 68:18–24, 1988.


9. Osenbach RK, Menezes AH: Spinal cord injury without radiographicabnormality in children. Pediatr Neurosci 15:168–175, 1989.


11. Pang D, Pollack IF: Spinal cord injury without radiographic ab-normality in children: The SCIWORA syndrome. J Trauma 29:654–664, 1989.

12. Pang D, Wilberger JE: Spinal cord injury without radiographicabnormalities in children. J Neurosurg 57:114–129, 1982.

13. Pollack IF, Pang D, Sclabassi R: Recurrent spinal cord injurywithout radiographic abnormalities in children. J Neurosurg 69:177–182, 1988.

14. Ramon S, Dominguez R, Ramirez L, Paraira M, Olona M, CastelloT, Garcia-Fernandez L: Clinical and magnetic resonance imagingcorrelation in acute spinal cord injury. Spinal Cord 35:664–673,1997.

15. Rathbone D, Johnson G, Letts M: Spinal cord concussion in pedi-atric athletes. J Pediatr Orthop 12:616–620, 1992.

16. Rossitch E Jr, Oakes WJ: Perinatal spinal cord injury: Clinical,radiographic, and pathological features. Pediatr Neurosurg 18:149–152, 1992.


18. Turgut M, Akpinar G, Akalan N, Ozcan OE: Spinal injuries in thepediatric age group: A review of 82 cases of spinal cord andvertebral column injuries. Eur Spine J 5:148–152, 1996.



CHAPTER 14

Diagnosis and Management of Traumatic Atlanto-occipitalDislocation Injuries

RECOMMENDATIONSDIAGNOSTIC:Standards: There is insufficient evidence to support diagnostic standards.Guidelines: There is insufficient evidence to support diagnostic guidelines.Options:• A lateral cervical x-ray is recommended for the diagnosis of atlanto-occipital dislocation. If a radiological

method for measurement is used, the basion-axial interval-basion-dental interval method is recommended.• The presence of upper cervical prevertebral soft tissue swelling on an otherwise nondiagnostic plain x-ray

should prompt additional imaging.• If there is clinical suspicion of atlanto-occipital dislocation, and plain x-rays are nondiagnostic, computed

tomography or magnetic resonance imaging is recommended, particularly for the diagnosis of non-Type IIdislocations.

TREATMENT:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options: Treatment with internal fixation and arthrodesis using one of a variety of methods is recommended.

Traction may be used in the management of patients with atlanto-occipital dislocation, but it is associatedwith a 10% risk of neurological deterioration.

RATIONALE

Although traumatic atlanto-occipital dislocation (AOD)is perceived to be an uncommon injury frequentlyresulting in death, improvements in emergency man-

agement of the patient in the field, rapid transport, and betterrecognition have resulted in more survivors of AOD in thepast 2 decades. Infrequent observation of patients with AODand missed diagnoses may impair outcomes of patients withthis unusual injury (44). An assimilation of the reported ex-periences of clinicians evaluating and managing AOD mayfacilitate development of diagnostic and treatment options forthis traumatic disorder. Specific questions that were evalu-ated include the sensitivity of plain x-rays, computed tomog-raphy (CT), and magnetic resonance imaging (MRI) in thediagnosis of AOD, as well as the safety and efficacy of varioustreatment modalities for AOD, including no treatment, trac-tion, external immobilization, and internal fixation withfusion.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject headings “atlanto-occipital

joint” and “dislocation” yielded 690 and 86,205 citations, re-spectively. A subset of 233 citations contained both headings.The reference lists of the articles were reviewed to identifyadditional case reports. Because fewer than 100 cases of sur-vivors of AOD were identified, even single case reports wereconsidered, provided that basic inclusion criteria were met.The articles were reviewed using the following criteria forinclusion in diagnosis: human survivors, type of traumaticatlanto-occipital dislocation, and plain radiographic findings.The articles were also reviewed using the following criteriafor inclusion in treatment: human survivors, type of traumaticAOD, management, and outcome. The observations from thereports were combined because the usual methods for analy-sis were precluded by the infrequent observation of this in-jury. The type of dislocation was classified according toTraynelis et al. (51) into Type I (anterior), Type II (longitudi-nal), and Type III (posterior) dislocations. Lateral, rotational,and multidirectional dislocations that could not be classifiedinto one of these types were considered separately and arenoted as “Other Type.” The duration of follow-up rangedfrom several weeks to 4 years. Of the articles meeting thediagnostic selection criteria, 48 articles with 79 patients pro-vided data on 29 Type I, 32 Type II, 4 Type III, and 14 othertypes of AOD. Two of these articles (10, 44) included one


patient each from two previously published individual casereports (41, 42). Of the articles meeting the treatment selectioncriteria, 43 articles with 62 patients provided data on 24 TypeI, 23 Type II, 3 Type III, and 12 other types of AOD. Two ofthese articles (10, 44) included one patient each from twopreviously published individual case reports (41, 42). All ar-ticles contained Class III medical evidence consisting of eithersingle case studies or small case series with no report contain-ing more than six patients. The information provided by thesereports was compiled and scrutinized, and it constitutes thebasis for this guideline. Summaries of these reports are pro-vided in Tables 14. 1 and 14.2.


Diagnosis

A variety of radiographic measurements (17) have beenproposed for the diagnosis of AOD on a lateral cervical x-ray(Fig. 14.1). A displacement of more than 10 mm between thebasion and dens is considered abnormal by Wholey et al. (53).A ratio of the basion-posterior atlas arch distance to theopisthion-anterior atlas arch distance of more than 1 is con-sidered abnormal by Powers et al. (43). A distance of morethan 13 mm between the posterior mandible and anterior atlasor 20 mm between the posterior mandible and dens is con-sidered abnormal by Dublin et al. (12). Failure of a line fromthe basion to the axis spinolaminar junction to intersect C2 ora line from the opisthion to the posterior inferior corner of thebody of the axis to intersect C1 are considered abnormal byLee et al. (32). Finally, a displacement of more than �12 mmor more than �4 mm between the basion and posterior C2line, or a displacement of more than 12 mm from the basion todens (2 mm more than the Wholey recommendation) is con-sidered abnormal by Harris et al. (24, 25). A comparativestudy by Lee et al. (32) found a 50% sensitivity of the Wholeymethod, 33% sensitivity of the Powers ratio, and a 25% sen-sitivity of the Dublin method. The authors applied their X-linemethod with a 75% sensitivity (32). Although neither thePowers ratio nor X-line method could be applied in nearlyhalf their patients, a comparative study by Harris et al. (25)found a 60% sensitivity of the Powers ratio, a 20% sensitivityof the Lee method, and 100% sensitivity of the basion-axialinterval-basion-dental interval (BAI-BDI) method among pa-tients in whom the required landmarks could be identified.Przybylski et al. (44) reported failure to diagnose AOD in twoof five patients with the Powers ratio, one of five patients withthe X-line method, and two of five patients with the BAI-BDImethod. No radiographic method reviewed has complete sen-sitivity. The BAI-BDI method proposed by Harris et al. (whichincorporates the basion-dens distance described by Wholey) isat present the most reliable means to diagnose AOD on alateral cervical spine x-ray.

Many of the case reports and case series in the literature donot describe the method(s) used for diagnosis of AOD. Be-cause the most sensitive method was proposed by Harris et al.(25) in 1994, this method was probably not used for many ofthe evaluations. Although, retrospectively, a diagnosis waspossible on the first lateral x-ray in 60 of 79 patients (sensi-

tivity, 0.76), the diagnosis was actually made in only 45 of the79 patients (sensitivity, 0.57) on the first lateral x-ray. Of the 15patients whose diagnosis could have been made on the firstlateral x-ray, 3 were not stratified by type, whereas 11 of theremaining 12 were not Type II dislocations (1, 7, 9, 13, 27, 29,41, 44, 46, 49, 52, 54). A second lateral x-ray (nine cases),tomography (one case), fluoroscopy (two cases), CT (twocases), and MRI (five cases) were required for diagnosis in 19of 79 patients (1, 3, 4, 8, 10, 19–21, 26, 28, 36, 40, 42–44). Thesensitivities stratified by type of dislocation are: Type I, 0.83(24 of 29 patients); Type II, 0.72 (23 of 32 patients); Type III,0.75 (3 of 4 patients), other type, 0.71 (10 of 14 patients).Because these data were obtained from case reports and smallcase series, comparison with the accuracy of plain x-rays inpatients without AOD could not be performed. As a result,specificity, predictive values, and likelihood ratios cannot bediscerned from the available literature.

Of the 15 patients in whom the diagnosis was missed on theinitial plain x-rays, the initial neurological condition of 3patients was unknown (1). Of the remaining 12 patients, 4were neurologically normal (one Type I, one Type III, twoother type) (13, 29). Two of those four patients originallyreported as normal developed a monoparesis (one Type I, oneother type) (7, 49). Neither recovered completely. Eight of theremaining 12 patients had neurological abnormalities fromthe outset, five of whom worsened. Four of the five transientlyworsened, including one Type I injury patient with quadri-paresis and Cranial Nerve IX, X, and XII palsies (9) who wasonly spastic at last follow-up. One patient with a Type I injurydeveloped a hemiparesis but recovered (27). One Type I in-jury patient developed quadriparesis and was hemiparetic atfollow-up (46). One lateral dislocation patient with parapare-sis and torticollis had recovered at last follow-up (52). Onepatient (Type I) with initial monoparesis experienced perma-nent worsening and was quadriplegic at follow-up (54).

Although plain x-rays do not reliably identify AOD, theindex of suspicion may be increased with the identification ofprevertebral soft tissue swelling. Although plain x-rays wereobtained in all cases considered, the presence or absence ofsoft tissue swelling was described in only half (1, 3, 4, 6, 9–11,15, 19, 21, 22, 26, 27, 31, 33–36, 38, 39, 42, 45, 48, 52, 54). Thesensitivity of soft tissue swelling is 0.90 (37 of 41 cases). Acutecraniocervical CT was performed in 40 of 79 patients withAOD (1–4, 6–11, 20, 22, 26–30, 33, 34, 38, 41, 42, 44, 45, 48, 52,55). However, for 15 of 40 patients, the authors did not reportwhether AOD was diagnosed by CT. The diagnosis of AODwas made by CT in 21 of 25 patients (sensitivity, 0.84). Al-though no other computed tomographic findings were re-ported in 11 of 40 patients, 24 of the remaining 29 patientswith AOD studied with CT had hemorrhages (19 subarach-noid hemorrhages, 1 subdural hemorrhage, 4 contusions).Five patients had no computed tomographic evidence of as-sociated hemorrhage. Nine of 15 patients in whom the diag-nosis of AOD was missed on the first plain x-ray had subse-quent acute computed tomographic scanning; 8 hadsubarachnoid or other associated hemorrhage (1, 9, 8, 44).Craniocervical MRI was performed in 18 of 79 patients withAOD (6–8, 10, 15, 21, 23, 26, 27, 30, 34, 37, 40, 42, 49, 55). The

S106 Guidelines for Management of Cervical Spinal Injuries


TABLE 14.1. Summary of Reports on Imaging Diagnosis of Atlanto-occipital Dislocation Injuriesa

Series (Ref. No.) AOD Type Diagnosis Made by X-ray Findings CT Findings MRI Findings

Grabb et al., 1999 (21) I Plain x-ray STS, Powers � Unreported Partial tear tectorial membrane

II Plain x-ray STS, Powers � None performed Tear posterior AOL

II MRI STS, Powers � None performed Partial tear tectorial membrane

Naso et al., 1997 (37) I/II Plain x-ray No mention STS Unreported Delayed study

Sponseller and Cass, 1997 (49) I Plain x-ray (missed) No mention STS None performed None performed

II Plain x-ray No mention STS Unreported BS contusion

Przybylski et al., 1996 (44) I MRI Powers/BDI/X-line � SAH, � Dx BS contusion, � Dx

II Plain x-ray (missed) Powers/BDI/X-line � SAH, � Dx BS contusion, � Dx

Pang and Wilberger, 1980 (41) II 2nd plain x-ray Powers/BDI �/X-line � SAH, � Dx None performed

I/lateral Plain x-ray (missed) Powers/BDI/X-line � Normal, head only None performed

I/lateral Plain x-ray (missed) Powers/BDI/X-line � SAH, � Dx None performed

Yamaguchi et al., 1996 (55) I Plain x-ray No mention STS SAH, � tomography BS contusion, � Dx

Guigui et al., 1995 (22) I Plain x-ray STS � Dx None performed

Ahuja et al., 1994 (1) I Fluoroscopy STS, Powers � SAH, unknown None performed

II 5 plain x-ray (3

missed)

STS, Powers � None performed None performed

II STS, Powers � SAH, � Dx None performed

II STS, Powers � SAH, unknown None performed

I/II STS, Powers � None performed None performed

I/II STS, Powers � SAH, � Dx None performed

Donahue et al., 1994 (11) I Plain x-ray STS None performed None performed

II Plain x-ray STS, 5 mm distraction None performed None performed

II Plain x-ray STS None performed None performed

II Plain x-ray 6 mm distraction Intracerebral bleed None performed

Palmer and Turney, 1994 (40) II CT No mention STS Unreported Cord contusion, � Dx

Dickman et al., 1993 (10) II Plain x-ray 15 mm distraction None performed None performed

Papadopoulos et al., 1991 (42) Rotatory CT STS � Dx None performed

Rotatory MRI STS No blood, � Dx Epidural, � Dx

II/rotatory 2nd plain x-ray STS � Dx Epidural, � Dx

Harmanli and Koyfman, 1993 (23) II Plain x-ray No mention STS None performed � Dx

Hosono et al., 1993 (27) I Plain x-ray (missed) STS Edema, head only Delayed study

Matava et al., 1993 (34) II Plain x-ray STS Delayed study None performed

II Plain x-ray No mention STS None, � Dx None performed

II Plain x-ray No mention STS SAH, � Dx BS contusion

Nischal et al., 1993 (38) II Plain x-ray STS BS contusion, � Dx None performed

II Plain x-ray STS � Dx None performed

Bundshuh et al., 1992 (6) I Plain x-ray STS SAH, � Dx SAH, � Dx

I Plain x-ray STS, Powers/X-line � SAH � Dx

Farley et al., 1992 (15) I Plain x-ray STS, Power � None performed Cord contusion

Belzberg and Tranmer, 1991 (3) II 2nd plain x-ray STS SAH, � Dx None performed

Hladky et al., 1991 (26) II MRI No mention STS Contusion, head only � Dx

II MRI No STS Normal, head only � Dx

Lee et al., 1991 (33) II Plain x-ray STS SAH, � Dx None performed

I/rotatory Plain x-ray STS � Dx None performed

Maves et al., 1991 (35) II Plain x-ray No mention STS None performed None performed

II Plain x-ray No mention STS None performed None performed

III Plain x-ray No mention STS None performed None performed

Montane et al., 1991 (36) I Plain x-ray STS None performed None performed

II 2nd plain x-ray STS None performed None performed

II 2nd plain x-ray No STS None performed None performed

Dibenedetto and Lee, 1990 (9) I Plain x-ray (missed) STS ICH, � Dx None performed

Jones et al., 1990 (30) I Plain x-ray No mention STS � Dx Premedullary edema

Colnet et al., 1989 (8) Lateral rotatory Tomography Late study SAH, � Dx Delayed study

Jevtich, 1989 (29) Lateral Plain x-ray (missed) No mention STS Delayed study None performed

Hummel and Plave, 1988 (28) I 2nd plain x-ray No mention STS Subdural, head only None performed

Zampella et al., 1988 (56) II Plain x-ray No mention STS SAH, head only Delayed study

Georgopoulos et al., 1987 (20) I Cineradiography No mention STS Delayed study None performed

Traumatic Atlanto-occipital Dislocation Injuries S107


MRI findings were not reported for 4 of the 18 patients stud-ied. The diagnosis of AOD could be made in 12 of 14 casesstudied with MRI (sensitivity, 0.86).

In summary, physicians often miss the diagnosis of AOD onplain x-rays (sensitivity, 0.57), particularly in the circumstanceof nonlongitudinal dislocations (non-Type II). Although im-proved interpretation may increase the sensitivity of plainx-rays to 0.76, additional imaging of the craniovertebral junc-tion with CT or MRI is recommended in patients suspected ofhaving AOD, given their superior sensitivity over plainx-rays. Other methods, such as fluoroscopy, tomography, andmyelography, have also been used to confirm the diagnosis ofAOD. Neurological abnormalities, including lower cranialnerve paresis (particularly Cranial Nerves VI, X, and XII),monoparesis, hemiparesis, quadriparesis, respiratory dys-function including apnea, and complete high cervical cordmotor deficits in the setting of normal plain spinal x-raysshould prompt additional imaging with CT or MRI. The pres-ence of prevertebral soft tissue swelling on plain x-rays andsubarachnoid hemorrhage at the craniovertebral junction oncomputed tomographic scans should prompt consideration ofthe diagnosis of AOD (5, 44).

Treatment

Ten patients in the literature did not receive initial treat-ment for AOD, nine of whom were not correctly diagnosed

until neurological worsening occurred (7, 8, 10, 20, 46, 48, 49,52, 54). Five of the nine patients had Type I injuries, and fourof the nine had other injury types. Four of nine had persistentdeficits at last follow-up that were worse in comparison withtheir examinations on presentation (7, 10, 49, 54). Two of thesepatients were normal initially. At last follow-up, one had aCranial Nerve X deficit with spasticity (Type I) (49), and onehad monoparesis (7). The other two patients had mild initialdeficits. One patient with an initial Cranial Nerve VI palsyhad hemiparesis at last follow-up (10), and another withinitial monoparesis was quadriplegic at follow-up (54). Fivepatients who worsened initially without treatment eventuallyimproved from their initial neurological condition. Finally,one quadriplegic patient with Type II AOD (56) who was nottreated improved to quadriparesis at last follow-up. In sum-mary, failure to treat AOD resulted in worsening of all pa-tients with incomplete injuries. Nearly half of these patientsfailed to improve to their initial examination baselineconditions.

Of 21 patients with AOD initially treated with traction, 2worsened transiently and developed worsening quadriparesisand Cranial Nerve VI deficits. Both patients had resolution oftheir Cranial Nerve VI deficits but not of their quadriparesiswith discontinuation of traction. One patient had a Type IIinjury (40), and one patient had a rotational dislocation (10).Four patients were initially normal and remained normal at

TABLE 14.1. Continued

Series (Ref. No.) AOD Type Diagnosis Made by X-ray Findings CT Findings MRI Findings

Bools and Rose, 1986 (4) I Plain x-ray STS SAH, � Dx None performed

III 2nd plain x-ray No mention STS None performed None performed

Collalto et al., 1986 (7) I/lateral Plain x-ray (missed) No STS SAH, head only Delayed study

Putnam et al., 1986 (45) I Plain x-ray STS, Powers � SAH, � Dx None performed

Ramsay et al., 1986 (46) I Plain x-ray (missed) No mention STS None performed None performed

Roy-Camille et al., 1986 (48) I Late plain x-ray No mention STS Delayed study None performed

I Plain x-ray STS None performed None performed

Zigler et al., 1986 (57) I Plain x-ray No mention STS None performed None performed

Watridge et al., 1985 (52) Lateral Plain x-ray (missed) No STS Delayed study None performed

Banna et al., 1983 (2) Rotatory Plain x-ray No mention STS � Dx None performed

Kaufman et al., 1982 (31) II Plain x-ray STS None performed None performed

II Plain x-ray STS None performed None performed

Woodring et al., 1981 (54) I Plain x-ray No mention STS None performed None performed

I Plain x-ray (missed) STS None performed None performed

Powers et al., 1979 (43) I Plain x-ray Late study None performed None performed

II 2nd plain x-ray No mention STS None performed None performed

Rockswold and Seljeskog, 1979 (47) II Plain x-ray No mention STS None performed None performed

Eismont and Bohlman, 1978 (13) III Plain x-ray (missed) No mention STS None performed None performed

Fruin and Pirotte, 1977 (18) I Plain x-ray No mention STS None performed None performed

Page et al., 1973 (39) I Plain x-ray STS None performed None performed

Evarts, 1970 (14) I Plain x-ray No mention STS None performed None performed

Gabrielsen and Maxwell, 1966 (19) I 2nd plain x-ray STS None performed None performed

Farthing, 1948 (16) III Plain x-ray No mention STS None performed None performed

a One patient was eliminated because the plain x-ray interpretation was not reported: Ferrara and Bartfield (17) (1 patient). Three articles wereeliminated because the type of dislocation was not reported: Georgopoulos et al. (20) (2 of 3 patients); Hladky et al. (26) (1 of 3 patients); Nasoet al. (37) (1 of 2 patients). One article (5 patients) was eliminated because individual patient data was not reported: Bulas et al. (5) (5 of 5patients). AOD, atlanto-occipital dislocation; MRI, magnetic resonance imaging; CT, computed tomography; STS, soft tissue swelling; BDI,basion-dental interval; SAH, subarachnoid hemorrhage; Dx, diagnosis. �, done or positive; �, not done or negative; BS, brainstem; ICH,intracerebral hemorrhage; AOL, atlanto-occipital ligament.



TABLE 14.2. Summary of Reports on Treatment of Atlanto-occipital Dislocation Injuriesa

Series (Ref. No.) Type Initial Examination Treatment Outcome

Naso et al., 1997 (37) Mixed I/II Quadriplegia Supportive Death 5 wk

Sponseller and Cass, 1997 (49) I Normal None (neuro worse), traction, fusion � brace Spastic, CN 10

II Normal Brace failed (6 wk), fusion Normal

Przybylski et al., 1996 (44) I Quadriplegia Collar � fusion Quadriplegia

Pang and Wilberger, 1980 (41) II Quadriplegia Halo failed (22 wk), fusion Quadriplegia

II Normal Fusion � collar CN 10

Mixed I/lateral Hemiplegia Collar � fusion Monoparesis

Mixed I/lateral Quadriparesis, CN 6/7/12 Fusion � collar CN 12

Yamaguchi et al., 1996 (55) I Quadriplegia, CN 10, 11, 12 Brace failed (10 wk), fusion Quadriplegia, CN 10, 11, 12

Guigui et al., 1995 (22) I Normal Fusion � brace Normal

Donahue et al., 1994 (11) I Hemiparesis Halo distracted (temporarily neuro worse),

fusion

Hyperreflexic

II CN 6 Halo � fusion Normal

II Quadriplegia, CN 7/10 Collar/traction � fusion Quadriparesis, CN 7/10

II Quadriparesis, CN 3/7 Fusion Quadriparesis

Palmer and Turney, 1994 (40) II Quadriparesis, CN 6 Traction (neuro worse), brace � fusion Quadriparesis

Dickman et al., 1993 (10) II Quadriplegia, CN 9/10 Brace Unchanged (sepsis, death at 3 mo)

Papadopoulos et al., 1991 (42) Rotatory Quadriparesis, CN 6 Traction (neuro worse), fusion � halo Quadreparesis

Rotatory CN 6 None (neuro worse), fusion � halo Hemiparesis

Mixed II/rotatory Hemiparesis, CN 3/6 Halo � fusion Normal

Harmanli and Koyfman, 1993 (23) II Hemiparesis, CN 12 Fusion � brace Normal

Hosono et al., 1993 (27) I Hemiparesis Brace (neuro worse), fusion � brace Normal

Matava et al., 1993 (34) II Hemiplegia, CN 6/12 Fusion � brace Spastic, CN 6

II Hemiparesis, CN 6 Fusion � brace Normal

II CN 6/9/10 Fusion � brace Spastic

Nischal et al., 1993 (38) II Quadriparesis, CN 3, 6, 9, 10 Brace � fusion Hemiparesis, CN 3, 6, 9, 10

II Quadriplegia, CN 9, 10 Brace � fusion Hemiparesis

Bundshuh et al., 1992 (6) I Quadriparesis, CN 6, 9, 10, 12 Traction � fusion CN 6, 12

Farley et al., 1992 (15) I Quadriplegia, CN 10 Traction � brace Quadriplegia

Belzberg and Tranmer, 1991 (3) II Quadriparesis, CN 6, 9, 10 Traction � brace � fusion Monoparesis, CN 6

Lee et al., 1991 (33) II Normal Traction � fusion Normal

Mixed I/rotatory CN 6 Brace � fusion CN 6

Montane et al., 1991 (36) I Hemiparesis Fusion � brace Spastic

II Quadriparesis Traction, fusion � brace Normal

II Quadriplegia Fusion � brace Quadriplegia

Dibenedetto and Lee, 1990 (9) I Quadriparesis, CN 9, 10, 12 Collar (neuro worse, 6 wk), fusion � brace Spastic

Colnet et al., 1989 (8) Mixed lateral/rotatory Hemiplegia, CN 6, 9, 10 None (neuro worse), traction � shunt �

decompression

Hemiparesis

Jevtich, 1989 (29) Lateral Normal Traction � brace Normal

Hummel and Plaue, 1988 (28) I Hemiparesis Fusion � brace Normal

Zampella et al., 1988 (56) II Quadriplegia, CN 5–12 None Quadriplegia, CN 6

Georgopoulos et al., 1987 (20) I Normal None (neuro worse), fusion � brace Normal

Bools and Rose, 1986 (4) III Normal Traction, fusion � brace Normal

Collalto et al., 1986 (7) Mixed I/lateral Normal None (neuro worse), fusion � brace Monoparesis

Putnam et al., 1986 (45) I Quadriplegia, CN 6 Brace Death (sepsis 8 mo)

Ramsay et al., 1986 (46) I Quadriparesis None (neuro worse), traction � brace Hemiplegia

Roy-Camille et al., 1986 (48) I CN 6, 11 None, brace failed (3 mo), traction � fusion CN 6

I Quadriplegia, CN 6, 9–12 Traction � fusion Quadriplegia

Zigler et al., 1986 (57) I Quadriplegia, CN 11 Traction � brace � fusion Quadriplegia

Watridge et al., 1985 (52) Lateral Paraparesis None (neuro worse), traction � fusion �

decompression � brace

Normal

Banna et al., 1983 (2) Rotatory Normal Traction (2 wk) Normal

Kaufman et al., 1982 (31) II Quadriplegia Brace � fusion Quadriparesis, CN 9, 10

II Monoparesis Brace Normal



follow-up (2, 4, 16, 33). The remaining 15 patients had im-proved neurological function compared with their initial find-ings at last follow-up (3, 6, 14, 15, 18, 19, 36, 39, 43, 47, 48, 54,57). Ten had Type I injuries, five had Type II injuries, two hadType III injuries, and two had other dislocations. In total, oneof six patients with Type II injuries and one of three patientswith other translational injuries had transient worsening withthe use of craniocervical traction. In summary, traction forAOD has been reported to cause occasional neurologicalworsening. In both circumstances, the worsening did notpersist after discontinuation of traction. Because the fre-quency of neurological worsening with traction for AOD isapproximately 10%, 10 times higher than for subaxial injuries,the use of traction should be considered with caution inpatients with AOD.

Of 19 patients initially treated with external immobilizationexcluding traction, 8 patients were immobilized in anticipa-tion of internal fixation and fusion and none worsened duringthe presurgical interval (one Type I, four Type II, three othertype) (10, 11, 31, 33, 38, 44). Of the remaining 11 patientstreated with external immobilization alone excluding traction,4 worsened transiently (three Type I, one Type II) (9, 11, 13,27). All subsequently underwent craniocervical fixation andfusion. Two of these patients were normal at follow-up (oneinitially normal, one initially hemiparetic), and two werespastic (one initially quadriparetic and one hemiparetic). Ofthe remaining seven patients managed with external immo-bilization alone who did not worsen while in external immo-bilization, two patients managed in collars and one patienttreated in a halo were unstable after 6 to 22 weeks of immo-bilization (one Type I, two Type II). Two were quadriplegic,and one was normal. All three underwent internal fixationand fusion without change in their initial neurological condi-tion at last follow-up. Only four patients with AOD weresuccessfully treated with external immobilization alone (oneType I, two Type II, one other dislocation). Of the 21 patientsinitially treated with traction, 6 were subsequently managedwith external immobilization and none developed neurolog-

ical worsening. Two of the six (both Type I) remained unsta-ble after 3 to 5 months of bracing and were subsequentlytreated with craniocervical fixation and fusion. Five of thosesix patients had improvement in their neurological conditionat follow-up. The sixth patient remained normal.

In summary, 5 of 13 patients with AOD who did not worsenneurologically while treated with external immobilization(with or without traction) failed to achieve bony union withstability without internal fixation and fusion. In addition, sixpatients transiently worsened with external immobilization(with or without initial traction). Factors affecting fusion orpersistent nonunion (e.g., degree and type of displacement,patient age, and association with occipital condyle fractures)could not be identified. Because 11 (28%) of 40 patients man-aged with external immobilization either deteriorated neuro-logically or failed to achieve craniocervical stability withoutsurgical internal fixation and fusion, treatment of AOD withexternal immobilization alone should be considered withcaution.

Finally, 19 patients in the literature were treated withplanned early craniocervical fusion with internal fixation.Only one patient worsened neurologically after surgery. Thispatient with a Type II injury was normal initially and devel-oped a Cranial Nerve X deficit that persisted at follow-up (44).All but 3 of the remaining 18 patients improved neurologi-cally at follow-up. Four had Type I, 10 had Type II, and 4 hadother types of dislocation. None of the patients treated withcraniocervical fusion and internal fixation had late instabilityrequiring reoperation or further treatment.

SUMMARY

AOD is an uncommon traumatic injury that is difficult todiagnose and is frequently missed on initial lateral cervicalx-rays. Patients who survive often have neurological impair-ment, including lower cranial neuropathies, unilateral or bi-lateral weakness, or quadriplegia. But nearly 20% of patients

TABLE 14.2. Continued

Series (Ref. No.) Type Initial Examination Treatment Outcome

Woodring et al., 1981 (54) I Hemiparesis, CN 6 Traction CN 6

I Monoparesis None (neuro worse), traction � fusion Quadriplegia

Powers et al., 1979 (43) I Hemiparesis, CN 6 Traction � brace Hemiparesis

II Hemiparesis, CN 7 Traction � brace Normal

Rockswold and Seljeskog, 1979 (47) II Hemiparesis, CN 6 Traction, brace � fusion Ambulates

Eismont and Bohlman, 1978 (13) III Normal Collar (neuro worse), fusion � brace Normal

Fruin and Pirotte, 1977 (18) I Hemiparesis, CN 6, 9–12 Traction � fusion CN 6, 11

Page et al., 1973 (39) I Quadriplegia, CN 10, 12 Traction, brace failed (5 mo), fusion Quadriparesis, CN 10

Evarts, 1970 (14) I Hemiparesis, CN 6, 9, 10, 12 Traction, brace � fusion CN 6

Gabrielsen and Maxwell, 1966 (19) I Hyperreflexic, CN 6 Traction, brace failed (3 mo), fusion Numb scalp

Farthing, 1948 (16) III Normal Traction � brace Normal

a Three articles were eliminated because the type of dislocation was not reported: Georgopoulos et al. (20) (2 of 3 patients); Bulas et al. (5)(5 of 5 patients); Naso et al. (37) (1 of 2 patients). Two articles (8 patients) were eliminated because the initial examination was not reported:Grabb et al. (21) (3 patients); Ahuja et al. (1) (5 patients). Two articles (6 patients) were eliminated because the treatment was not reported: Maveset al. (35) (3 patients); Hladky et al. (26) (3 patients). Two articles were eliminated because the outcome was not reported: Jones et al. (30) (1patient); Bools and Rose (4) (1 of 2 patients). CN, cranial nerve; neuro, neurological examination.



with acute traumatic AOD will have a normal neurologicalexamination on presentation. The lack of localizing featuresmay impede diagnosis in the patient with a normal cervicalx-ray. A high index of suspicion must be maintained to diag-nose AOD. Prevertebral soft tissue swelling on a lateral cer-vical x-ray or craniocervical subarachnoid hemorrhage onaxial CT has been associated with AOD and may promptconsideration of the diagnosis. Additional imaging, includingCT and MRI, may be required to confirm the diagnosis ofAOD if plain x-rays are inadequate. All patients with AODshould be treated. Without treatment, nearly all patients de-veloped neurological worsening, and some did not recover.Although treatment with traction and external immobiliza-tion has been used successfully in some patients, transient orpermanent neurological worsening and late instability havebeen reported more often in association with these treatmentsthan with surgical treatment. Consequently, craniocervicalfusion with internal fixation is recommended for the treat-ment of patients with acute traumatic AOD.


Although the use of external immobilization for AOD wasoften associated with late instability, several patients achievedstability without operative management. CT with three-dimensional reconstruction for more precise measurement of

the magnitude of displacement and MRI for differentiation ofpartial and complete ligament tears from stretch injuries maybe useful in identifying a subgroup of patients in whomstability might be achieved with external immobilizationalone. Because AOD remains relatively infrequent, coopera-tive prospective collection of plain radiographic, computedtomographic, and MRI data in patients with AOD is recom-mended to determine whether a subgroup of patients withAOD can be treated with external immobilization alone withfewer occurrences of late instability.

Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, Division of Neuro-logical Surgery, 516 Medical Education Building, 1813 6th AvenueSouth, Birmingham, AL 35294-3295.

REFERENCES

1. Ahuja A, Glasauer FE, Alker GJ Jr, Klein DM: Radiology insurvivors of traumatic atlanto-occipital dislocation. Surg Neurol41:112–118, 1994.

2. Banna M, Stevenson GW, Tumiel A: Unilateral atlanto-occipitaldislocation complicating an anomaly of the atlas. J Bone JointSurg Am 65A:685–687, 1983.

3. Belzberg AJ, Tranmer BI: Stabilization of traumatic atlanto-occipital dislocation. J Neurosurg 75:478–482, 1991.

FIGURE 14.1. Midsagittal diagrams of thecraniocervical junction show the variousmethods for identifying AOD on a lateralcervical x-ray. A, the Wholey measure; B,the Powers ratio; C, the Dublin measure; D,the X-line method; E, the BAI-BDI method.



4. Bools JC, Rose BS: Traumatic atlantooccipital dislocation: Twocases with survival. AJNR Am J Neuroradiol 7:901–904, 1986.

5. Bulas DI, Fitz CR, Johnson DL: Traumatic atlanto-occipital dislo-cation in children. Radiology 188:155–158, 1993.

6. Bundschuh CV, Alley JB, Ross M, Porter IS, Gudeman SK: Mag-netic resonance imaging of suspected atlanto-occipital disloca-tion: Two case reports. Spine 17:245–248, 1992.

7. Collalto PM, DeMuth WW, Schwentker EP, Boal DK: Traumaticatlanto-occipital dislocation: Case report. J Bone Joint Surg Am68A:1106–1109, 1986.

8. Colnet G, Chabannes J, Commun C, Rigal MC, Alassaf M:Atlanto-occipital luxation and syringomyelia: 2 rare complica-tions of cervical injury—Diagnosis and therapeutic effects: Apro-pos of a case [in French]. Neurochirurgie 35:58–63, 1989.

9. Dibenedetto T, Lee CK: Traumatic atlanto-occipital instability: Acase report with follow-up and a new diagnostic technique. Spine595–597, 1990.

10. Dickman CA, Papadopoulos SM, Sonntag VKH, Spetzler RF,Rekate HL, Drabier J: Traumatic occipitoatlantal dislocations.J Spinal Disord 6:300–313, 1993.

11. Donahue DJ, Muhlbauer MS, Kaufman RA, Warner WC, SanfordRA: Childhood survival of atlantooccipital dislocation:Underdiagnosis, recognition, treatment and review of the litera-ture. Pediatr Neurosurg 21:105–111, 1994.

12. Dublin AB, Marks WM, Weinstock D, Newton TH: Traumaticdislocation of the atlanto-occipital articulation (AOA) with short-term survival: With a radiographic method of measuring theAOA. J Neurosurg 52:541–546, 1980.

13. Eismont FJ, Bohlman HH: Posterior atlanto-occipital dislocationwith fractures of the atlas and odontoid process: Report of a casewith survival. J Bone Joint Surg Am 60A:397–399, 1978.

14. Evarts CM: Traumatic occipito-atlantal dislocation. J Bone JointSurg Am 52A:1653–1660, 1970.

15. Farley FA, Graziano GP, Hensinger RN: Traumatic atlanto-occipital dislocation in a child. Spine 17:1539–1541, 1992.

16. Farthing JW: Atlantocranial dislocation with survival: A casereport. N C Med J 9:34–36, 1948.

17. Ferrera PC, Bartfield JM: Traumatic atlanto-occipital dislocation:A potentially survivable injury. Am J Emerg Med 14:291–296,1996.

18. Fruin AH, Pirotte TP: Traumatic atlantooccipital dislocation: Casereport. J Neurosurg 46:663–665, 1977.

19. Gabrielsen TO, Maxwell JA: Traumatic atlanto-occipital disloca-tion: With case report of a patient who survived. Am JRoentgenol Radium Ther Nucl Med 97:624–629, 1966.

20. Georgopoulos G, Pizzutillo PD, Lee MS: Occipito-atlantal insta-bility in children: A report of five cases and review of the litera-ture. J Bone Joint Surg Am 69A:429–436, 1987.

21. Grabb BC, Frye TA, Hedlund GL, Vaid YN, Grabb PA, Royal SA:MRI diagnosis of suspected atlanto-occipital dissociation in child-hood. Pediatr Radiol 29:275–281, 1999.

22. Guigui P, Milaire M, Morvan G, Lassale B, Deburge A: Traumaticatlantooccipital dislocation with survival: Case report and reviewof the literature. Eur Spine J 4:242–247, 1995.

23. Harmanli O, Koyfman Y: Traumatic atlanto-occipital dislocationwith survival: A case report and review of the literature. SurgNeurol 39:324–330, 1993.

24. Harris JH, Carson GC, Wagner LK: Radiologic diagnosis of trau-matic occipitovertebral dissociation: Part 1—Normal oc-cipitovertebral relationships on lateral radiographs of supine sub-jects. Am J Radiol 162:881–886, 1994.

25. Harris JH Jr, Carson GC, Wagner LK, Kerr N: Radiologic diagno-sis of traumatic occipitovertebral dissociation: Part 2—Compari-son of three methods of detecting occipitovertebral relationshipson lateral radiographs of supine subjects. Am J Radiol 162:887–892, 1994.

26. Hladky JP, Lejeune JP, Leclercq F, Dhellemmes P, Christiaens JL: Trau-matic atlanto-occipital dislocation [in French]. Neurochirurgie 37:312–317, 1991.

27. Hosono N, Yonenobu K, Kawagoe K, Hirayama N, Ono K: Trau-matic anterior atlanto-occipital dislocation: A case report withsurvival. Spine 18:786–790, 1993.

28. Hummel A, Plaue R: Diagnosis and treatment of atlanto-occipitalruptures [in German]. Unfallchirurgie 14:311–319, 1988.

29. Jevtich V: Traumatic lateral atlanto-occipital dislocation withspontaneous bony fusion: A case report. Spine 14:123–124, 1989.

30. Jones DN, Knox AM, Sage MR: Traumatic avulsion fracture of theoccipital condyles and clivus with associated unilateral atlantooc-cipital distraction. AJNR Am J Neuroradiol 11:1181–1183, 1990.

31. Kaufman RA, Dunbar JS, Botsford JA, McLaurin RL: Traumaticlongitudinal atlanto-occipital distraction injuries in children.AJNR Am J Neuroradiol 3:415–419, 1982.

32. Lee C, Woodring JH, Goldstein SJ, Daniel TL, Young AB, TibbsPA: Evaluation of traumatic atlantooccipital dislocations. AJNRAm J Neuroradiol 8:19–26, 1987.

33. Lee C, Woodring JH, Walsh JW: Carotid and vertebral injury insurvivors of atlanto-occipital dislocation: Case reports and liter-ature review. J Trauma 31:401–407, 1991.

34. Matava MJ, Whitesides TE Jr, Davis PC: Traumatic atlanto-occipital dislocation with survival: Serial computerized tomogra-phy as an aid to diagnosis and reduction—A report of three cases.Spine 18:1897–1903, 1993.

35. Maves CK, Souza A, Prenger EC, Kirks DR: Traumatic atlanto-occipital disruption in children. Pediatr Radiol 21:504–507, 1991.

36. Montane I, Eismont FJ, Green BA: Traumatic occipitoatlantal dis-location. Spine 16:112–116, 1991.

37. Naso WB, Cure J, Cuddy BG: Retropharyngealpseudomeningocele after atlanto-occipital dislocation: Report oftwo cases. Neurosurgery 40:1288–1291, 1997.

38. Nischal K, Chumas P, Sparrow O: Prolonged survival afteratlanto-occipital dislocation: Two case reports and review. Br JNeurosurg 7:677–682, 1993.

39. Page CP, Story JL, Wissinger JP, Branch CL: Traumatic atlantooc-cipital dislocation: Case report. J Neurosurg 39:394–397, 1973.

40. Palmer MT, Turney SZ: Tracheal rupture and atlanto-occipitaldislocation: Case report. J Trauma 37:314–317, 1994.

41. Pang D, Wilberger JE Jr: Traumatic atlanto-occipital dislocationwith survival: Case report and review. Neurosurgery 7:503–508,1980.

42. Papadopoulos SM, Dickman CA, Sonntag VKH, Rekate HL,Spetzler RF: Traumatic atlantooccipital dislocation with survival.Neurosurgery 28:574–579, 1991.

43. Powers B, Miller MD, Kramer RS, Martinez S, Gehweiler JA Jr:Traumatic anterior atlanto-occipital dislocation. Neurosurgery4:12–17, 1979.

44. Przybylski GJ, Clyde BL, Fitz CR: Craniocervical junction sub-arachnoid hemorrhage associated with atlanto-occipital disloca-tion. Spine 21:1761–1768, 1996.

45. Putnam WE, Stratton FT, Rohr RJ, Stitzell W, Roat G: Traumaticatlanto-occipital dislocations: Value of the Powers ratio in diag-nosis. J Am Osteopath Assoc 86:798–804, 1986.

46. Ramsay AH, Waxman BP, O’Brien JF: A case of traumatic atlanto-occipital dislocation with survival. Injury 17:412–413, 1986.



47. Rockswold GL, Seljeskog EL: Traumatic atlantocranial dislocationwith survival. Minn Med 62:151–152, 1979.

48. Roy-Camille R, Benazet JP, Saillant G, Henry P, Mamoudy P,Leonard P: Traumatic atlanto-occipital luxation: Value of newradiologic signs (apropos of 2 cases) [in French]. Rev Chir OrthopReparatrice Appar Mot 72:303–309, 1986.

49. Sponseller PD, Cass JR: Atlanto-occipital fusion for dislocation inchildren with neurologic preservation: A case report. Spine 22:344–347, 1997.

50. Deleted in proof.51. Traynelis VC, Marano GD, Dunker RO, Kaufman HH: Traumatic

atlanto-occipital dislocation: Case report. J Neurosurg 65:863–870, 1986.52. Watridge CB, Orrison WW, Arnold H, Woods GA: Lateral atlanto-

occipital dislocation: Case report. Neurosurgery 17:345–347, 1985.

53. Wholey MH, Bruwer AJ, Baker HL: The lateral roentgenogram ofthe neck (with comments on the atlanto-odontoid-basion relation-ship). Radiology 71:350–356, 1958.

54. Woodring JH, Selke AC Jr, Duff DE: Traumatic atlantooccipitaldislocation with survival. AJR Am J Radiol 137:21–24, 1981.

55. Yamaguchi N, Ikeda K, Ishise J, Yamashita J: Traumatic atlanto-occipital dislocation with long-term survival. Neurol Med Chir(Tokyo) 36:36–39, 1996.

56. Zampella EJ, Duvall ER, Langford KH: Computed tomographyand magnetic resonance imaging in traumatic locked-in syn-drome. Neurosurgery 22:591–593, 1988.

57. Zigler JE, Waters RL, Nelson RW, Capen DA, Perry J: Occipito-cervico-thoracic spine fusion in a patient with occipito-cervicaldislocation and survival. Spine 11:645–646, 1986.

Drawings by Leonardo da Vinci. Courtesy, Dr. Edwin Todd, Pasadena, California.


CHAPTER 15

Occipital Condyle Fractures

RECOMMENDATIONSDIAGNOSTIC:Standards: There is insufficient evidence to support diagnostic standards.Guidelines: Computed tomographic imaging is recommended for establishing the diagnosis of occipital

condyle fractures. Clinical suspicion should be raised by the presence of one or more of the followingcriteria: blunt trauma patients sustaining high-energy craniocervical injuries, altered consciousness, occip-ital pain or tenderness, impaired cervical motion, lower cranial nerve paresis, or retropharyngeal soft tissueswelling.

Options: Magnetic resonance imaging is recommended to assess the integrity of the craniocervical ligaments.

TREATMENT:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options: Treatment with external cervical immobilization is recommended.

RATIONALE

Although the traumatic occipital condyle fracture (OCF)was first described by Bell (2) in 1817, more frequentobservation of this injury has only been reported dur-

ing the past 2 decades. Improvements in computed tomo-graphic imaging technology and use of computed tomo-graphic imaging of the head-injured patient that includes thecraniovertebral junction have resulted in more frequent rec-ognition of this injury. However, the overall infrequent occur-rence of OCF and missed diagnoses in patients with OCF mayresult in late neurological deficits in these patients. An anal-ysis of the reported cases of OCF may facilitate developmentof diagnostic and treatment recommendations for this disor-der and is undertaken in this chapter. Specific questions thatwere addressed include the accuracy of plain x-rays andcomputed tomography (CT) in the diagnosis of OCF and thesafety and efficacy of various treatment modalities, includingno treatment, traction, external immobilization, decompres-sion, and internal fixation with fusion.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject headings “occipital bone”and “fracture” (spinal, cranial, or fracture alone) yielded 1,830and 33,537 citations, respectively. A subset of 218 citationscontained both headings. The references of the articles werereviewed to identify additional case reports. The articleswere reviewed using the following criteria for inclusion indiagnosis: human survivors, type of fracture, and tomo-graphic or plain radiographic findings. The articles were also

reviewed using the following criteria for inclusion in treat-ment: human survivors, type of fracture, management, andoutcome. Because fewer than 100 cases of survivors wereidentified, even single case reports were considered, providedthat basic inclusion criteria were met. The observations fromthe reports were combined because the usual methods foranalysis were precluded by the infrequent occurrence of thisinjury. Forty-seven articles met the selection criteria, provid-ing data on a total of 91 patients. All but two articles containedClass III data of either single case studies or small case series;none contained more than 15 patients. The two exceptionswere prospective studies to evaluate the use of clinical criteriain blunt trauma patients to prompt computed tomographicimaging of the cranial base (5, 26). The duration of follow-upin all articles ranged from several weeks to 5 years. The dataprovided by these reports was compiled and constitute thebasis for this guideline. Summaries of the 43 articles mostgermane to this topic are provided in Table 15.1.


Diagnosis

Plain x-rays of the cervical spine were obtained in nearly all91 patients culled from the literature review. Normal imagingwas reported in 42 patients. Eight patients had prevertebralsoft tissue swelling, only four of whom did not have associ-ated cervical fractures (5, 17, 19, 28, 30, 41). Ten patients withcervical fractures or displacements were described withoutmention of the presence or absence of soft tissue swelling (3,4, 16, 25, 31, 35, 40, 41). Three patients had multiple cervicalfractures (5, 31). Associated fractures included four atlas, two


Type II odontoid, three axis, two C5 fractures, one each of C3,C6, and C7 fractures, and one unspecified cervical fracture.One patient had atlantoaxial widening, and one had C5–C6subluxation (16, 41). Only two patients were identified withOCF on plain x-rays of the cranium or cervical spine (21, 44).The results of plain x-rays were not reported in 28 patients,and plain x-rays were not obtained in one patient with an oldhealed fracture identified by CT (12). The calculated sensitiv-ity of plain x-rays from these reports in the diagnosis of OCFis 3.2% (2 of 62 patients). Because the data were obtained fromcase reports and small case series of patients known to haveOCF, comparison with the findings of plain x-rays in patientswithout OCF could not be performed. As a result, specificity,positive predictive value, and negative predictive value couldnot be determined.

The type of fracture was classified according to Andersonand Montesano (1) into Type I (comminuted from impact),Type II (extension of a linear basilar cranial fracture) (20), andType III (avulsion of a fragment) fractures (Fig. 15.1). The 91patients in this review population provided data on 85 uni-lateral fractures (12 Type I, 24 Type II, and 49 Type III), 4bilateral fractures (one Type I, two Type III, and one mixedType I and Type III), and 2 old fractures.

All but one patient (44) underwent tomographic imaging (6polytomography alone, 83 computed tomographic imagingalone, and 1 both). One OCF was missed with polytomogra-phy and subsequently identified on CT (33). Two patients hadOCF diagnosed from retrospective review of computed tomo-graphic images that were initially interpreted as normal (12).The diagnosis of OCF could be made in every patient withOCF. Bloom et al. (5) performed a prospective study during a1-year period to identify the frequency of OCF in patientsmeeting certain clinical criteria. Fifty-five consecutive patientswith high-energy blunt craniocervical trauma underwentthin-section craniocervical junction computed tomographicimaging. Supplemental criteria included reduced GlasgowComa Score at admission, occipitocervical tenderness, re-duced craniocervical motion, lower cranial nerve abnormal-ity, and retropharyngeal soft tissue swelling. Nine (16.4%) of55 patients were identified with OCF. Other reports haveestimated a 1 to 3% frequency of OCF in patients sustainingblunt craniocervical trauma (24, 31). Similarly, Link et al. (26)reported the results of craniocervical CT on 202 patients witha Glasgow Coma Score between 3 and 6. OCF was identifiedin 9 (4.4%) of 202 patients.

Loss of consciousness was observed in 36 of 44 patients.Among 64 patients who had a sufficiently detailed neurolog-ical examination reported, 25 were normal, 24 had acute ordelayed cranial nerve deficits alone, 6 had cranial nerve def-icits with limb weakness, 6 had mild to severe limb weaknesswithout cranial nerve deficits, 1 had a delayed onset of ver-tigo, 1 had hyperreflexia, and 1 had diplopia. Only 4 patientswere found who did not have occipitocervical pain in theabsence of significantly impaired consciousness (28, 32, 41).

One patient was intoxicated, one had severe extremity pain,and two had severe facial trauma.

Only 11 patients were investigated with magnetic reso-nance imaging (MRI) (3, 12, 13, 15, 19, 21–23, 41, 42). Earlycraniocervical MRI was performed in eight patients, and lateMRI studies were obtained in three patients. Cervicomedul-lary hemorrhages were seen in three patients, two had normalimaging, one had a retrodental hemorrhage, one had a torntectorial membrane, and one had demonstration of the frac-ture. Displaced fracture fragments were observed in all threepatients with delayed MRI. Although early MRI has beeninfrequently reported after OCF, Tuli et al. (41) proposed anew classification scheme using MRI to differentiate stablefrom unstable OCF. However, the case example they gavedemonstrated concurrent atlantoaxial instability thatprompted occipitocervical fusion (rather than atlanto-occipital instability and OCF fracture).

In summary, the diagnosis of OCF is rarely made on plainx-rays. Imaging of the craniovertebral junction with CT orother tomographic methods is recommended in patients sus-pected of having OCF. Blunt trauma patients sustaining high-energy craniocervical injuries may be more likely to sustainOCF. Consequently, cranial imaging should include evalua-tion of the craniocervical junction. Other clinical criteria, in-cluding altered consciousness, occipital pain or tenderness,impaired craniocervical motion, lower cranial nerve paresis,or retropharyngeal soft tissue swelling, should prompt com-puted tomographic imaging of the craniocervical junction.

Treatment

Twenty-three patients (2 Type I, 14 Type II, 5 Type III, 2unknown type) did not receive treatment (6, 9, 12, 13, 18, 31,33, 34, 41, 42, 44, 45). Nine of these patients (one Type I, fourType II, four Type III) developed cranial nerve deficits withindays to weeks after injury (6, 9, 12, 13, 31, 33, 34, 42, 45). Onehypoglossal nerve palsy resolved, two hypoglossal nerve def-icits improved, three other cranial nerve deficits persisted(two hypoglossal, one glossopharyngeal, and one vagal), andthree outcomes were not reported. Six additional patientsdeveloped delayed deficits or symptoms. Two initially un-treated patients (one Type II, one Type III) developed multi-ple lower cranial nerve deficits that improved after 6 weeks ofcervical immobilization (23). Another initially untreated pa-tient (Type III) developed vertigo after 3 months that resolvedafter 8 weeks of collar immobilization (7). One patient (TypeIII) developed nystagmus and a lateral rectus palsy afterprecautionary collar immobilization was discontinued. Thedeficit resolved after cervical immobilization was resumed(14). One patient (Type III) developed double vision duringcervical traction that resolved with surgical decompression(45). Finally, one patient (Type III) developed delayed vagal,spinal accessory, and hypoglossal nerve palsies during cervi-cal immobilization in a cervical collar (8). The Cranial Nerve

Occipital Condyle Fractures


TAB

LE15

.1.

Sum

mar

yof

Rep

orts

onO

ccip

ital

Con

dyle

Frac

ture

sa

Seri

es(R

ef.

No.

)A

ge(y

r)/S

exTy

peLo

catio

nPa

inPl

ain

X-r

ayC

TM

RI

Exam

inat

ion

Trea

tmen

tO

utco

me

Legr

oset

al.,

2000

(23)

71/F

III�

Unr

epU

nrep

L,�

Epid

ural

Del

CN

6,7,

106

wk

colla

r18

mo

CN

10

44/M

II�

Unr

epU

nrep

R,

�N

orm

alD

elC

N6,

9–12

6w

kco

llar

3.5

mo

CN

10

Ide

etal

.,19

98(1

9)25

/MIII

��

STS,

C1

FxR

,�

Tect

oria

lm

embr

ane

tear

Nor

mal

10w

kco

llar

10w

kno

rmal

Dem

isch

etal

.,19

98(1

3)45

/FII

Unr

epU

nrep

Unr

epR

,�

FxD

elC

N12

Non

e1

yrim

pC

N12

Blo

omet

al.,

1997

(5)

21/M

IIIU

nrep

Unr

epST

S,C

6,7

FxR

,�

Unr

epN

orm

al�

8w

kco

llar

Nor

mal

36/F

IIIU

nrep

Unr

epU

nrep

L,�

Unr

epN

orm

al�

8w

kco

llar

Pain

15/F

I/IU

nrep

Unr

epU

nrep

B,

�U

nrep

Qpa

resi

s�

8w

kco

llar

Imp

Qpa

resi

s

45/F

III/I

Unr

epU

nrep

Unr

epB

,�

Unr

epC

N12

�8

wk

colla

rPa

in,

CN

12

22/F

IIU

nrep

Unr

epU

nrep

R,

�U

nrep

Nor

mal

�8

wk

colla

rU

nrep

21/M

IU

nrep

Unr

epST

S,C

1,2,

5Fx

R,

�U

nrep

Nor

mal

�8

wk

colla

rU

nrep

41/M

IU

nrep

Unr

epU

nrep

R,

�U

nrep

Nor

mal

�8

wk

colla

rN

orm

al

6/F

IIU

nrep

Unr

epU

nrep

L,�

Unr

epN

orm

al�

8w

kco

llar

Nor

mal

25/F

IU

nrep

Unr

epST

S,C

2Fx

L,�

Unr

epN

orm

al�

8w

kco

llar

Unr

ep

20/M

IU

nrep

Unr

epC

FxR

,�

Unr

epPp

legi

a�

8w

kco

llar

Unr

ep

Tuli

etal

.,19

97(4

1)64

/FIII

Unr

ep�

STS

R,

�N

one

Nor

mal

12w

kco

llar

3m

ono

rmal

69/F

IIIU

nrep

�A

Aw

ide

L,�

FxM

pare

sis,

CN

OC

fusi

onIm

p

27/M

Old

Unr

ep�

Nor

mal

L,�

Non

eN

orm

alN

one

3yr

norm

al

Cot

talo

rda

etal

.,19

96(1

0)15

/FI

Unr

ep�

Nor

mal

R,

�N

one

Nor

mal

7w

kM

iner

vaTr

colla

r4

mo

norm

al

Lam

and

Stra

tford

,19

96(2

2)20

/FIII

Unr

epU

nrep

Nor

mal

R,

�C

ontu

sion

Hpa

,C

N12

3m

oha

lo5

yrim

pC

N12

Urc

ulo

etal

.,19

96(4

2)62

/MIII

Unr

epU

nrep

Nor

mal

R,

�Fx

Del

CN

9,10

Non

e6

mo

sam

e

Nob

lean

dSm

oker

,19

96(3

1)33

/MI

Unr

epU

nrep

Unr

ep?,

�N

one

Del

CN

12N

one

Unr

ep

26/M

IU

nrep

Unr

epU

nrep

?,�

Non

eG

CS

15N

one

Unr

ep

16/M

IIU

nrep

Unr

epU

nrep

?,�

Non

eG

CS

13N

one

Unr

ep

32/M

IIU

nrep

Unr

epC

2Fx

?,�

Non

eC

N7,

12N

one

Unr

ep

53/F

IIU

nrep

Unr

epU

nrep

?,�

Non

eG

CS

8N

one

Unr

ep

47/F

IIU

nrep

Unr

epU

nrep

?,�

Non

eG

CS

15N

one

Unr

ep

37/M

IIU

nrep

Unr

epU

nrep

?,�

Non

eG

CS

8N

one

Unr

ep

11/M

IIU

nrep

Unr

epU

nrep

?,�

Non

eG

CS

13N

one

Unr

ep

33/M

IIU

nrep

Unr

epU

nrep

?,�

Non

eG

CS

15N

one

Unr

ep

23/M

IIU

nrep

Unr

epU

nrep

?,�

Non

eU

nrep

Unr

epU

nrep

39/M

IIIU

nrep

Unr

epII

Od

Fx?,

�N

one

CN

7H

alo

Unr

ep

88/M

IIIU

nrep

Unr

epC

1,II

Od

Fx?,

�N

one

GC

S15

Hal

oU

nrep

29/M

IIIU

nrep

Unr

epU

nrep

?,�

Non

eU

nrep

Unr

epU

nrep

14/F

IIIU

nrep

Unr

epU

nrep

?,�

Non

eG

CS

11C

olla

rU

nrep

17/F

IIIU

nrep

Unr

epU

nrep

?,�

Non

eG

CS

7N

one

Unr

ep

Cas

tling

and

Hic

ks,

1995

(9)

21/M

II�

�N

orm

alR

,�

Non

eD

elC

N12

Non

e2

yrno

rmal

Emer

yet

al.,

1995

(15)

26/M

IIIU

nrep

�N

orm

alL,

�Fr

actu

reH

yper

refle

xic

Col

lar

4m

ono

rmal

Pale

yan

dW

ood,

1995

(34)

21/M

IIIU

nrep

�N

orm

alL,

�N

orm

alD

elC

N12

Non

e6

mo

imp

CN

12

Stro

oban

tset

al.,

1994

(39)

27/M

III�

�N

orm

alR

,�

Non

eN

orm

al10

wk

colla

r21

mo

norm

al

12/F

III�

�C

1Fx

L,�

Non

eN

orm

al4

wk

Min

erva

Nor

mal

Was

serb

erg

and

Bar

tlett,

1995

(45)

39/M

III�

Unr

epN

orm

alL,

�N

one

Del

CN

12N

one

12

24/M

III�

�N

orm

alL,

�N

one

Del

dipl

opia

Dec

ompr

essi

onN

orm

al

16/M

III�

Unr

epN

orm

alR

,�

Non

eB

rain

inju

ryC

olla

r3

mo

CN

12

34/M

IIIU

nrep

Unr

epN

orm

alR

,�

Non

eU

nrep

Hal

oU

nrep



TAB

LE15

.1.

Con

tinu

edSe

ries

(Ref

.N

o.)

Age

(yr)

/Sex

Type

Loca

tion

Pain

Plai

nX

-ray

CT

MR

IEx

amin

atio

nTr

eatm

ent

Out

com

e

You

nget

al.,

1994

(47)

26/F

III�

Unr

epN

orm

alL,

�N

one

Hpa

,C

N9–

1212

wk

halo

14m

oim

pC

N9–

12

20/M

III�

Unr

epN

orm

alR

,�

Non

eG

CS

7C

olla

r1

yrH

pa

Man

nan

dC

ohen

,19

94(2

7)23

/MIII

��

Nor

mal

R,

�N

one

Nor

mal

6w

kco

llar

Nor

mal

Ols

son

and

Kun

z,19

94(3

2)43

/MIII

Unr

ep�

Nor

mal

L,�

Non

eN

orm

alC

olla

rN

orm

al

Shar

ma

etal

.,19

94(3

7)35

/MII

Unr

epU

nrep

Nor

mal

L,�

Non

eC

N9,

10D

ecom

pres

sion

3m

oim

pC

N9,

10

Mas

saro

and

Lano

tte,

1993

(29)

21/M

IIIU

nrep

Unr

epN

orm

alL,

�N

one

Hse

nsor

y,C

N12

8w

kM

iner

va2

yrC

N12

Rai

laet

al.,

1993

(35)

25/M

III�

�N

orm

alL,

�N

one

Nor

mal

6w

kco

llar

Nor

mal

67/M

III�

�C

1ab

norm

alL,

�N

one

Nor

mal

Col

lar

Nor

mal

Bet

tini

etal

.,19

93(3

)39

/FI

Unr

ep�

C3

FxL,

�N

one

Nor

mal

Unr

epU

nrep

24/M

II�

Unr

epN

orm

alR

,�

Non

eC

oma

Unr

epU

nrep

21/F

III�

Unr

epU

nrep

?,�

Con

tusi

onC

oma

Unr

epU

nrep

21/M

III/II

IU

nrep

�N

orm

alB

,�

Non

eN

orm

alU

nrep

Unr

ep

Leve

ntha

let

al.,

1992

(25)

42/F

II�

Unr

epN

orm

alL,

�N

one

CN

6,7

3m

oco

llar

Unr

ep

19/F

III�

�N

orm

alL,

�N

one

Nor

mal

Col

lar

Unk

now

n

43/M

IIIU

nrep

�C

5Fx

R,

�N

one

Nor

mal

3m

oco

llar

Nor

mal

17/F

II�

Unr

epL1

FxR

,�

Non

eG

CS

103

mo

colla

rN

orm

al

36/M

I�

GC

S8

T1Fx

R,

�N

one

GC

S8

3m

oha

loN

orm

al

17/M

I�

GC

S4

Nor

mal

R,

�N

one

GC

S4

3m

oco

llar

Nor

mal

Mod

yan

dM

orri

s,19

92(3

0)21

/MIII

�U

nrep

STS

L,�

Non

eU

nrep

6w

kco

llar

18m

ono

sym

ptom

s

Boz

boga

etal

.,19

92(7

)34

/FIII

��

Nor

mal

L,�

Non

eL

Hpa

,di

plop

iaLa

tede

com

pres

sion

4yr

norm

al

37/M

III�

Unr

epU

nrep

L,�

Non

eD

elve

rtig

oD

el8

wk

colla

r3

yrno

rmal

Bri

dgm

anan

dM

cNab

,19

92(8

)32

/MIII

��

Nor

mal

L,�

Non

eD

elC

N10

–12

Col

lar

1yr

imp

CN

10–1

2

Wan

iet

al.,

1991

(44)

67/M

II�

Unr

ep�

Con

dfx

L,no

neN

one

CN

9–12

Non

eC

N10

,12

Wes

sels

,19

90(4

6)26

/MIII

��

Unr

epR

,�

Non

eC

N7–

12C

olla

r6

wk

imp

CN

7–12

7m

o/M

II�

Unr

epU

nrep

L,�

Non

eC

N5,

7–12

Col

lar

4m

oC

N7–

12

27/M

II�

Unr

epU

nrep

R,

�N

one

CN

7–12

Col

lar

6w

kim

p

Mar

iani

,19

90(2

8)30

/MIII

��

STS

R,

�N

one

Nor

mal

8w

kco

llar

Nor

mal

Jone

set

al.,

1990

(21)

43/M

III/II

I�

Unr

ep�

Con

dFx

B,

�C

ontu

sion

Qpl

egia

OC

Fx4

wk

Qpl

egia

Des

aiet

al.,

1990

(14)

33/M

III�

�N

orm

alL,

�N

one

CN

6C

olla

r4

mo

norm

al

Val

aska

tzis

and

Ham

mer

,19

90(4

3)19

/MIII

��

Nor

mal

R,

�N

one

Nor

mal

6w

kco

llar

Nor

mal

Orb

ayet

al.,

1989

(33)

37/M

IIIU

nrep

�N

orm

alL,

�(to

mo

�)

Non

eD

elC

N12

Non

e15

mo

CN

12

Savo

lain

eet

al.,

1989

(36)

71/F

III�

�N

orm

alR

,�

Non

eH

pleg

ia,

CN

6Tr

,ha

loLM

pare

sis

And

erso

nan

dM

onte

sano

,19

88(1

)3/

MI

�U

nrep

Nor

mal

R,

�N

one

Unc

onSo

ft24

mo

norm

al

18/F

III�

Unr

epN

orm

al?,

�N

one

Unr

epM

iner

va36

mo

22/M

III�

Unr

epN

orm

alR

,to

mo

�N

one

Unc

onH

alo

12m

ono

rmal

23/M

III�

Unr

epN

orm

alL,

�N

one

Unc

onC

olla

rD

eath

25/M

III�

Unr

epN

orm

al?,

tom

o�

Non

eU

nrep

Min

erva

17m

o

37/M

II�

Unr

epN

orm

alL,

�N

one

Unc

onC

olla

r12

mo

norm

al

Cur

riet

al.,

1988

(11)

16/F

III�

Unr

epN

orm

alR

,�

Non

eD

ecer

ebra

teC

olla

r6

mo

unre

p

Has

him

oto

etal

.,19

88(1

8)71

/MII

�U

nrep

Nor

mal

L,�

Non

eC

N9–

12N

one

6m

oC

N9–

12

Dee

bet

al.,

1988

(12)

25/F

IIU

nrep

Unr

epN

orm

alD

elL,

�N

one

CN

12N

one

Unr

ep

66/F

Old

Unr

ep�

Non

eD

elL,

�Fx

Nor

mal

Non

eU

nrep

Spen

cer

etal

.,19

84(3

8)19

/MI

�G

CS

8N

orm

alL,

�N

one

GC

S8

Col

lar,

halo

BC

N9,

10

Gol

dste

inet

al.,

1982

(16)

24/F

IIIU

nrep

�C

5,6

slx

L,to

mo

�N

one

Nor

mal

2m

oco

llar

Nor

mal

Har

ding

-Sm

ithet

al.,

1981

(17)

18/M

III�

Unr

epST

SR

,to

mo

�N

one

Unc

onC

olla

r16

mo

norm

al

Bol

ende

ret

al.,

1978

(6)

23/M

IIIU

nrep

Unr

epN

orm

alR

,to

mo

�N

one

CN

9–12

Non

eU

nrep

22/M

IIU

nrep

Unr

epN

orm

alR

,to

mo

�N

one

Del

CN

6,9,

10N

one

Unr

ep

aC

T,co

mpu

ted

tom

ogra

phy;

MR

I,m

agne

tic

reso

nanc

eim

agin

g;U

nrep

,unr

epor

ted;

Del

,del

ayed

;�,d

one

orpo

siti

ve;�

,not

done

orne

gati

ve;F

x,fr

actu

re;I

mp,

impr

ovem

ent;

STS,

soft

tiss

uesw

ellin

g;C

,cer

vica

l;A

A,a

tlan

toax

ial;

Q,q

uadr

i-;P

,par

a-;M

,mon

o-;O

C,o

ccip

ital

cond

yle;

CN

,cra

nial

nerv

e;G

CS,

Gla

sgow

Com

aSc

ale;

Od,

odon

toid

;L,l

eft;

R,

righ

t;B

,bi

late

ral;

tom

o,to

mog

raph

y;Tr

,tr

acti

on;

Unc

on,

unco

nfir

med

;H

pa,

hem

ipar

esis

;C

ond,

cond

ylar

;sl

x,su

blux

atio

n.

Occipital Condyle Fractures S117


X and XI palsies improved. However, the hypoglossal palsypersisted at 1 year.

Forty-four patients were treated with cervical collar immo-bilization (8 Type I, 8 Type II, 28 Type III) (1, 4, 5, 7, 8, 10, 11,14–17, 19, 23, 25, 27, 28, 30–32, 35, 39, 41, 43, 45, 46). Thirteenpatients were treated with halo/Minerva immobilization (2Type I, 11 Type III) (1, 22, 25, 29, 31, 36, 38, 40, 45, 47).Treatment was unreported in six patients (3, 31). Five patients(one Type II, four Type III) underwent surgery. Two patientswith Type III OCF were treated with occipitocervical fusion(one with concurrent atlanto-occipital dislocation and onewith atlantoaxial instability) (21, 41). One patient (Type III)with delayed diplopia had symptom resolution after removalof the fracture fragment (45), whereas one patient (Type II)with lower cranial nerve deficits (37) and one (Type III) withdiplopia and hemiparesis (7) remained unchanged severaldays after surgery. The latter patient subsequently recoverednormal function.

In summary, 12 of 15 patients who developed delayedsymptoms or deficits were not initially treated. Only 3 of these12 patients were subsequently treated with cervical immobi-lization. All three improved. In comparison, only three of sixpatients demonstrated improvement in deficits without treat-ment. Only one patient (Type III) developed a deficit duringtreatment that persisted (hypoglossal nerve palsy) despitecollar use. Only three patients underwent surgery for decom-pression of the brainstem, one of whom had immediate andlasting improvement in symptoms postoperatively. Because12 of 23 patients developed delayed deficits without treat-ment and another developed a deficit after premature discon-tinuation of treatment, the literature suggests that patientswith Type III OCF should be treated with external immobili-zation. Treatment of patients with OCF Types I and II mayinclude external immobilization.

SUMMARY

OCF is an uncommon injury requiring computed tomo-graphic imaging for diagnosis. Patients sustaining high-energy blunt craniocervical trauma, particularly in the settingof loss of consciousness, impaired consciousness, occipitocer-vical pain or motion impairment, and lower cranial nervedeficits, should undergo computed tomographic imaging ofthe craniocervical junction. Untreated patients with OCF often

develop lower cranial nerve deficits that usually recover orimprove with external immobilization. Identification of TypeIII OCF should prompt external immobilization. Additionaltreatment may be dictated by the presence of associated cer-vical fractures or instability.


Although Type III OCF is considered by many authors tobe unstable, not all patients, treated or not, developed neuro-logical deficits. Computed tomographic imaging with three-dimensional reconstruction for more precise measurement ofthe magnitude of fracture displacement and MRI for differ-entiation of partial and complete ligamentous injuries may beuseful in identifying subgroups of patients who do not re-quire treatment or, conversely, require more rigid halo immo-bilization, rather than collar immobilization. Because OCFinjuries remain relatively infrequent, cooperative retrospec-tive collection of plain x-ray, computed tomographic, andMRI data in patients with OCF is recommended.


REFERENCES

1. Anderson PA, Montesano PX: Morphology and treatment of oc-cipital condyle fractures. Spine 13:731–736, 1988.

2. Bell C: Surgical observations. Middlesex Hosp J 4:469, 1817.3. Bettini N, Malaguti MC, Sintini M, Monti C: Fractures of the

occipital condyles: Report of four cases and review of the litera-ture. Skeletal Radiol 2:187–190, 1993.

4. Bloom AI, Neeman Z, Floman Y, Gomori J, Bar-Ziv J: Occipitalcondyle fracture and ligament injury: Imaging by CT. PediatrRadiol 26:786–790, 1996.

5. Bloom AI, Neeman Z, Slasky BS, Floman Y, Milgrom M, RivkindA, Bar-Ziv J: Fracture of the occipital condyles and associatedcraniocervical ligament injury: Incidence, CT imaging and impli-cations. Clin Radiol 52:198–202, 1997.

6. Bolender N, Cromwell LD, Wendling L: Fracture of the occipitalcondyle. AJR Am J Radiol 131:729–731, 1978.

7. Bozboga M, Unal F, Hepgul K, Izgi N, Turantan MI, Turker K:Fracture of the occipital condyle: Case report. Spine 17:1119–1121,1992.

FIGURE 15.1. Classification of occipital condyle fractures according to Anderson and Montesano (1). A, Type I fractures mayoccur with axial loading. B, Type II fractures are extensions of a basilar cranial fracture. C, Type III fractures may result froman avulsion of the condyle during rotation, lateral bending, or a combination of mechanisms.



8. Bridgman SA, McNab W: Traumatic occipital condyle fracture,multiple cranial nerve palsies, and torticollis: A case report andreview of the literature. Surg Neurol 38:152–156, 1992.

9. Castling B, Hicks K: Traumatic isolated unilateral hypoglossalnerve palsy: Case report and review of the literature. Br J OralMaxillofac Surg 33:171–173, 1995.

10. Cottalorda J, Allard D, Dutour N: Fracture of the occipital con-dyle. J Pediatr Orthop 5:161–163, 1996.

11. Curri D, Cervellini P, Zanusso M, Benedetti A: Isolated fracture ofoccipital condyle: Case report. J Neurosurg Sci 32:157–159, 1988.

12. Deeb ZL, Rothfus WE, Goldberg AL, Daffner RH: Occult occipitalcondyle fractures presenting as tumors. J Comput Tomogr 12:261–263, 1988.

13. Demisch S, Lindner A, Beck R, Zierz S: The forgotten condyle:Delayed hypoglossal nerve palsy caused by fracture of the occip-ital condyle. Clin Neurol Neurosurg 100:44–45, 1998.

14. Desai SS, Coumas JM, Danylevich A, Hayes E, Dunn EJ: Fractureof the occipital condyle: Case report and review of the literature.J Trauma 30:240–241, 1990.

15. Emery E, Saillant G, Ismail M, Fohanno D, Roy-Camille R: Frac-ture of the occipital condyle: Case report and review of theliterature. Eur Spine J 4:191–193, 1995.

16. Goldstein SJ, Woodring JH, Young AB: Occipital condyle fractureassociated with cervical spine injury. Surg Neurol 17:350–352,1982.

17. Harding-Smith J, MacIntosh PK, Sherbon KJ: Fracture of theoccipital condyle: A case report and review of the literature.J Bone Joint Surg Am 63A:1170–1171, 1981.

18. Hashimoto T, Watanabe O, Takase M, Koniyama J, Kobota M:Collet-Sicard syndrome after minor head trauma. Neurosurgery23:367–370, 1988.

19. Ide C, Nisolle JF, Misson N, Trigaux JP, Gustin T, De Coene B,Gilliard C: Unusual occipitoatlantal fracture dissociation with noneurological impairment: Case report. J Neurosurg 88:773–776,1998.

20. Jacoby CG: Fracture of the occipital condyle. AJR Am JRoentgenol 132:500, 1979 (letter).

21. Jones DN, Knox AM, Sage MR: Traumatic avulsion fracture of theoccipital condyles and clivus with associated unilateral atlantooc-cipital distraction. AJNR Am J Neuroradiol 11:1181–1183, 1990.

22. Lam CH, Stratford J: Bilateral hypoglossal nerve injury withoccipital condylar fracture. Can J Neurol Sci 23:145–148, 1996.

23. Legros B, Fournier P, Chiaroni P, Ritz O, Fusciardi J: Basal frac-ture of the skull and lower (IX, X, XI, XII) cranial nerves palsy:Four case reports including two fractures of the occipital con-dyle—A literature review. J Trauma 48:342–348, 2000.

24. Leone A, Cerase A, Colosimo C, Lauro L, Puca A, Marano P:Occipital condylar fractures: A review. Radiology 216:635–644,2000.

25. Leventhal MR, Boydston WR, Sebes JI, Pinstein ML, Watridge CB,Lowrey R: The diagnosis and treatment of fractures of the occip-ital condyle. Orthopedics 15:944–947, 1992.

26. Link TM, Schuierer G, Hufendiek A, Horch C, Peters PE: Sub-stantial head trauma: Value of routine CT examination of thecervicocranium. Radiology 196:741–745, 1995.

27. Mann FA, Cohen W: Occipital condyle fracture: Significance inthe assessment of occipitoatlantal stability. AJR Am J Radiol163:193–194, 1994.

28. Mariani PJ: Occipital condyle fracture presenting as aretropharyngeal hematoma. Ann Emerg Med 19:1447–1449, 1990.

29. Massaro F, Lanotte M: Fracture of the occipital condyle. Injury24:419–420, 1993.

30. Mody BS, Morris EW: Fracture of the occipital condyle: Casereport and review of the world literature. Injury 23:350–352, 1992.

31. Noble ER, Smoker WR: The forgotten condyle: The appearance,morphology, and classification of occipital condyle fractures.AJNR Am J Neuroradiol 17:507–513, 1996.

32. Olsson R, Kunz R: Fracture of the occipital condyle as an inciden-tal finding during CT-evaluation of a maxillary fracture. ActaRadiol 35:90–91, 1994.

33. Orbay T, Aykol S, Seckin Z, Ergun R: Late hypoglossal nervepalsy following fracture of the occipital condyle. Surg Neurol31:402–404, 1989.

34. Paley MD, Wood GA: Traumatic bilateral hypoglossal nervepalsy. Br J Oral Maxillofac Surg 33:239–241, 1995.

35. Raila FA, Aitken AT, Vickers GN: Computed tomography andthree-dimensional reconstruction in the evaluation of occipitalcondyle fracture. Skeletal Radiol 22:269–271, 1993.

36. Savolaine ER, Ebraheim NA, Jackson WT, Rusin JJ: Three-dimensional computed tomography in evaluation of occipitalcondyle fracture. J Orthop Trauma 3:71–75, 1989.

37. Sharma BS, Mahajan RK, Bhatia S, Khosla VK: Collet-Sicard syn-drome after closed head injury. Clin Neurol Neurosurg 96:197–198, 1994.

38. Spencer JA, Yeakley JW, Kaufman HH: Fracture of the occipitalcondyle. Neurosurgery 15:101–103, 1984.

39. Stroobants J, Fidlers L, Storms JL, Klaes R, Dua G, Van Hoye M:High cervical pain and impairment of skull mobility as the onlysymptoms of an occipital condyle fracture: Case report.J Neurosurg 81:137–138, 1994.

40. Stroobants J, Seynaeve P, Fidlers L, Klaes R, Brabants K, VanHoye M: Occipital condyle fracture must be considered in thepediatric population: Case report. J Trauma 36:440–441, 1994.

41. Tuli S, Tator CH, Fehlings MG, Mackay M: Occipital condylefractures. Neurosurgery 41:368–377, 1997.

42. Urculo E, Arrazola M, Arrazola M Jr, Riu I, Moyua A: Delayedglossopharyngeal and vagus nerve paralysis following occipitalcondyle fracture: Case report. J Neurosurg 84:522–525, 1996.

43. Valaskatzis EP, Hammer AJ: Fracture of the occipital condyle: Acase report. S Afr Med J 77:47–48, 1990.

44. Wani MA, Tandon PN, Banerji AK, Bhatia R: Collet-Sicard syn-drome resulting from closed head injury: Case report. J Trauma31:1437–1439, 1991.

45. Wasserberg J, Bartlett RJV: Occipital condyle fractures diagnosedby high-definition CT and coronal reconstructions. Neuroradiol-ogy 37:370–373, 1995.

46. Wessels LS: Fracture of the occipital condyle: A report of 3 cases.S Afr J Surg 28:155–156, 1990.

47. Young WF, Rosenwasser RH, Getch C, Jallo J: Diagnosis andmanagement of occipital condyle fractures. Neurosurgery 34:257–261, 1994.

Occipital Condyle Fractures S119


CHAPTER 16

Isolated Fractures of the Atlas in Adults

RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS: Treatment options in the management of isolated fractures of the atlas are based on the specific

atlas fracture type. It is recommended that isolated fractures of the atlas with an intact transverse atlantalligament be treated with cervical immobilization alone. It is recommended that isolated fractures of theatlas with disruption of the transverse atlantal ligament be treated with either cervical immobilization aloneor surgical fixation and fusion.

RATIONALE

The atlas vertebra is subject to a variety of acute fractureinjuries and may be associated with other cervical frac-tures and ligamentous traumatic injuries (4, 8, 25, 26,

31). Although the treatment of atlas fractures in combinationwith other cervical fracture injuries is most commonly linkedto the treatment of the associated injury (8), isolated fracturesof the atlas occur with sufficient frequency to warrant review.

The medical literature addressing the management of frac-tures of the atlas was examined using evidence-based medi-cine techniques to determine the optimal treatment for iso-lated atlas fractures, including isolated anterior or posteriorarch fractures, anterior and posterior arch fractures (burstfractures), lateral mass fractures, comminuted fractures, andtransverse process fractures.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject heading “vertebral fracture”combined with “atlas” and “human” yielded 360 references.The abstracts were reviewed, and articles addressing clinicalmanagement and follow-up of atlas fractures were selectedfor inclusion. The relative infrequency of these fractures, thesmall number of collected case series, and the numerous casereports with pertinent information required rather broad in-clusion and exclusion criteria. Several papers addressing rel-evant background information such as biomechanics and ra-diology were included. The bibliographies of the selectedarticles were also reviewed to provide additional referencesand to assess completeness of the literature review. Theseefforts resulted in 32 articles describing acute traumatic atlasfractures. Ten Class III articles (eight case series and two casereports) documenting treatment of patients with atlas frac-tures are summarized in Table 16.1. The remaining 22 articlesare included in the reference list and contribute to the scien-

tific foundation. Treatment options are summarized in Table16.2.


Atlas fractures account for approximately 1 to 2% of allfractures of the human spinal column and roughly 2 to 13% ofall acute cervical spine fractures (8, 21, 30). The first knownfracture of the atlas was demonstrated at autopsy by Cooperin 1822 and has been the subject of a series of historicalpublications (28). In 1920, Jefferson (15) reviewed 42 previ-ously described cases of atlas fracture and added 4 new cases.Although his article documents a variety of atlas fracturepatterns, it is best known for the characterization of the “Jef-ferson fracture,” a burst fracture injury of the atlas ring (10). In1945, Hinchey and Bickel (13a) added 112 cases of atlas frac-ture to the literature. Sherk and Nicholson summarized anadditional 73 cases in 1970 (30).

Spence et al. (31), in 1970, reported their findings of a studyof the mechanism of atlas fracture and potential rupture of thetransverse atlantal ligament. Using 10 cadaveric specimens,the authors studied the application of force required to frac-ture C1 and to rupture the transverse ligament (range, 38–104kg; mean, 58 kg). The sum of the excursion of the C1 lateralmasses over the C2 lateral masses after traumatic injuryranged from 4.8 to 7.6 mm (mean, 6.3 mm). The authorsconcluded that if the sum of lateral mass displacement (LMD)of C1 over C2 on the anteroposterior radiographic image ismore than 6.9 mm, then the transverse atlantal ligament is“probably torn.” In a follow-up clinical and biomechanicalstudy, Fielding et al. (5) confirmed these findings. These twostudies, completed before the era of magnetic resonance im-aging (MRI), are the basis for the widely quoted “rules ofSpence” (i.e., �6.9 mm LMD � transverse atlantal ligamentdisruption) offered to assist in the management of patientswith isolated atlas fractures. Subsequently, Heller et al. (12)reported their observations on 35 open-mouth odontoid filmsusing calibration markings to assess radiographic magnifica-


tion. They found an 18% magnification factor on open-mouthodontoid-view x-rays. Applying this information to the eval-uation of atlas burst fractures by means of the rules of Spencesuggests that the sum of the LMD measurements indicatingatlantal transverse ligament disruption should be increasedfrom 6.9 to 8.1 mm. This study pointed out the difficulty inusing plain radiographic measurements to assess the integrityof the transverse atlantal ligament after acute traumatic atlasfracture.

Hadley et al. (8) reported a series of 57 patients with atlasfractures, representing 6.6% of their series of cervical fracturesin 1988. The authors managed 32 patients with isolated atlasfractures, none of whom sustained neurological injury. Theauthors based their treatment recommendations on the degreeof LMD in each patient. Twelve patients had nondisplacedatlas fractures. Of these, 10 were treated with a rigid collar,one with a soft collar, and one with a suboccipital-mandibularimmobilizer (SOMI). The duration of treatment for these pa-tients was 8 to 12 weeks. Fifteen patients had an LMD of lessthan 7 mm. Eight were treated with a rigid collar, three witha SOMI, and four with a halo immobilization brace for 10 to 12weeks. The remaining five patients had an LMD of more than7 mm and were managed with a halo orthosis. These fivepatients were treated for 12 to 16 weeks. Of the 32 patients, 29were available for long-term follow-up (median, 40 mo).Three complained of neck pain. All were successfully treated.No patient required subsequent surgical fixation. The authorsconcluded that isolated fractures of the atlas are effectivelymanaged with external immobilization alone for 12 weeks(median duration). Atlas fractures with an LMD of more than6.9 mm required more rigid immobilization (halo orthosis)than those with an LMD of less than 6.9 mm (cervical collar).Levine and Edwards (21) described their experience with 34patients with isolated atlas fractures in 1991. They followed asimilar treatment algorithm with similar success.

Fowler et al. (6) reported a series of 48 consecutive atlasfracture patients, representing 5.5% of all cervical fractures intheir experience. In their series, 33% of their patients had otherassociated cervical spine fractures. Atlas fractures were di-vided into burst (n � 30), posterior arch (n � 17), and anteriorarch fractures (n � 1). None of the patients with an isolatedatlas fracture presented with neurological deficit. These au-thors suggested treatment with closed traction reduction ofisolated atlas fractures if the LMD is more than 7.0 mm, andthen immobilization in a rigid collar. No patient in this seriesunderwent surgical fixation. All were effectively treated withthis management scheme at last follow-up, although the du-ration of treatment was not specified. In 1991, Kesterson et al.(16) reported their series of 17 cases of atlas fractures. Thirteenwere isolated atlas fractures and were considered stable. Allwere successfully managed with rigid cervical immobilization(nine collar, one SOMI, one halo, two Minerva). Again, theduration of treatment was not specified. Several other authorshave described the successful treatment of isolated atlas frac-

tures with rigid cervical immobilization, using similar man-agement principles (9, 13, 20, 29, 32).

Landells and Van Peteghem (18) described a series of 35patients with atlas fractures, representing 4.7% of their insti-tutional experience with acute cervical fracture injuries. Theauthors categorized atlas fractures into three types. Type Ifractures involved a single arch and occurred in 16 of their 35patients. Type II fractures were burst fractures and repre-sented 13 of the 35 isolated fractures they treated. Type IIIfractures were atlas lateral mass fractures identified in 6 of the35 patients. The authors used the original rules of Spence toassist with the identification of stability and noted four pa-tients with an LMD of more than 6.9 mm. Regardless of thefracture type or stability, all fractures except one were initiallytreated with external immobilization for an unreported lengthof time. The one exception was a patient with a Type I fracturewho underwent early surgery with C1–C2 wiring and fusion.The reason for the exception is not made clear in the text. Only1 of 34 patients treated with external immobilization requiredsurgery for late instability. The authors observed no relation-ship between successful treatment and the amount of initialLMD. They recommended that atlas fractures be initiallytreated with rigid external immobilization. They noted thatlate instability can occur and recommended clinical follow-upof these patients.

Clinically observed atlas fracture patterns can be repro-duced in cadaveric experimental models of axial loading (11).In a series of biomechanical studies, Oda et al. (24, 25) andPanjabi et al. (26) reproduced these atlas fracture patternswith axial loading and found that the burst fracture wasassociated with postinjury hypermobility at C1–C2. Theseauthors described a 42% increase in flexion/extension motion,a 24% increase in lateral bending, and a 5% increase in axialrotation. They found that in all instances of transverse atlantalligament disruption, the atlantodens interval was more than 3mm. The authors concluded that the atlantodens interval wasthe most reliable predictor of transverse ligament disruptionin adults after acute C1 fracture.

McGuire and Harkey (22), in 1995, described two cases ofunstable atlas burst fractures treated with surgical fixation andfusion. The fractures were thought to be unstable based on apredental space more than 5 mm and an LMD more than 9 mm,respectively. Both were treated with posterior C1–C2 transartic-ular screw fixation and fusion with good results. The authorsreported that transarticular screw fixation obviated the need forhalo immobilization postoperatively. Several other authors havereported successful surgical fixation and fusion for atlas frac-tures when they are associated with disruption of the transverseligament with resultant instability (17, 18, 27). These few patientswere treated with posterior C1–C2 wiring and fusion proceduresand were managed in rigid orthoses (halo, Guilford brace) for 12to 16 weeks postoperatively.

More recently, it has been proposed that MRI is a more sen-sitive indicator of transverse atlantal ligament disruption than

Isolated Fractures of the Atlas in Adults


TAB

LE16

.1.

Sum

mar

yof

Rep

orts

onTr

eatm

ent

ofFr

actu

res

ofth

eA

tlas

inA

dult

sa

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ceC

lass

Con

clus

ions

Lee

etal

.,19

98(2

0)R

etro

spec

tive

revi

ewin

clud

ing

12ca

ses

ofis

olat

edat

las

frac

ture

.III

All

trea

ted

succ

essf

ully

with

exte

rnal

imm

obili

zatio

n.Tr

eatm

ent

algo

rith

mpr

opos

ed:

Stab

le,

trea

tw

ithco

llar

12w

k.U

nsta

ble,

prop

oses

surg

ical

fixat

ion.

(Inst

abili

tyde

fined

asla

tera

lm

ass

disp

lace

men

t�

7m

mor

MR

Iev

iden

ceof

tran

sver

selig

amen

tdi

srup

tion)

.

McG

uire

and

Har

key,

1995

(22)

2ca

ses

ofun

stab

leat

las

burs

tfr

actu

retr

eate

dw

ithpo

ster

ior

tran

sart

icul

arsc

rew

fixat

ion

and

fusi

on.

IIIC

onsi

dere

dun

stab

leba

sed

onpr

eden

tal

spac

e�

5m

man

dLM

D�

9m

m.

Bot

htr

eate

dsu

cces

sful

ly.

Ace

rvic

alco

llar

was

used

post

oper

ativ

ely.

Levi

nean

dEd

war

ds,

1991

(21)

Ret

rosp

ectiv

ere

view

of34

patie

nts

with

atla

sfr

actu

res.

Med

ian

follo

w-u

p,4.

5yr

.III

Ifla

tera

lm

ass

disp

lace

men

t�

7m

m,

patie

nts

trea

ted

with

colla

r,an

dif

�7

mm

,pa

tient

str

eate

dw

ithei

ther

halo

alon

eor

redu

ced

intr

actio

nan

dm

aint

aine

dun

tilhe

aled

(6w

kin

trac

tion

and

6w

kin

halo

).N

opa

tient

str

eate

dsu

rgic

ally

.

Kes

ters

onet

al.,

1991

(16)

Ret

rosp

ectiv

ere

view

incl

udin

g13

patie

nts

with

isol

ated

atla

sbu

rst

(Jeffe

rson

)fr

actu

re.

IIIA

llsu

cces

sful

lytr

eate

dw

ithim

mob

iliza

tion.

Fow

ler

etal

.,19

90(6

)R

etro

spec

tive

revi

ewof

48co

nsec

utiv

eat

las

frac

ture

sdi

vide

din

tobu

rst

(30)

,po

ster

ior

arch

(17)

,an

dan

teri

orar

chfr

actu

res

(1).

IIIA

utho

rssu

gges

tre

duct

ion

bytr

actio

nif

LMD

�7.

0,fo

llow

edby

colla

r.N

opa

tient

sun

derw

ent

surg

ical

fixat

ion.

Had

ley

etal

.,19

88(8

)R

etro

spec

tive

revi

ewin

clud

ing

32is

olat

edfr

actu

res

ofth

eat

las.

Med

ian

follo

w-u

p,40

mo

on29

/32

frac

ture

s.

IIITh

efo

llow

ing

trea

tmen

tpa

ttern

sw

ere

reco

gniz

ed:

LMD

�7

mm

(5pa

tient

s):

trea

ted

with

halo

.LM

D0–

7m

m(1

5pa

tient

s):

4tr

eate

din

halo

,11

trea

ted

inSO

MI.

LMD

none

(12

patie

nts)

:tr

eate

din

rigi

dco

llar.

Non

eof

thes

eis

olat

edC

1fr

actu

res

sust

aine

dne

urol

ogic

alin

jury

orre

quir

edsu

rger

y.3

com

plai

ned

ofne

ckpa

in;

othe

rwis

e,al

lw

ere

succ

essf

ully

trea

ted.

Aut

hors

’re

com

men

datio

n:is

olat

edC

1fr

actu

res

can

bem

anag

edw

ithou

tea

rly

surg

ical

fixat

ion.

Ifth

eLM

D�

6.9

mm

,th

enha

loim

mob

iliza

tion

isin

dica

ted.

Land

ells

and

Van

Pete

ghem

,19

88(1

8)R

etro

spec

tive

revi

ewof

35pa

tient

sw

ithfr

actu

reof

the

atla

s.III

The

auth

ors

outli

nea

clas

sific

atio

nsc

hem

e:Ty

peI:

sing

lear

ch(1

6),

mos

tpr

eval

ent

and

mos

tof

ten

asso

ciat

edw

ithot

her

frac

ture

s.Ty

peII:

burs

tfr

actu

re(1

3),

mos

tof

ten

inis

olat

ion,

only

1/13

with

defic

it.Ty

peIII

:la

tera

lm

ass

frac

ture

(6).

Trea

tmen

tno

tst

anda

rdbu

t34

patie

nts

wer

etr

eate

dw

ithri

gid

exte

rnal

imm

obili

zatio

n.O

nly

1pa

tient

trea

ted

with

earl

ysu

rger

y(T

ype

Ifr

actu

retr

eate

dw

ithC

1–C

2fu

sion

).O

nepa

tient

requ

ired

surg

ery

info

llow

-up.

56%

ofpa

tient

sre

port

edsi

gnifi

cant

sym

ptom

sat

1yr

(nec

kpa

in,

scal

pdy

sest

hesi

as).

Aut

hors

argu

efo

rco

nser

vativ

em

easu

res

with

trac

tion

and

imm

obili

zatio

nw

ithca

refu

lfo

llow

-up.

Sega

let

al.,

1987

(28)

Ret

rosp

ectiv

ere

view

incl

udin

g8

isol

ated

atla

sfr

actu

res.

Med

ian

follo

w-u

p,46

mo.

III2/

4pa

tient

sw

itha

com

min

uted

frac

ture

,de

scri

bed

asa

unila

tera

lav

ulsi

onof

the

tran

sver

selig

amen

tat

tach

men

tan

dad

jace

ntar

chfr

actu

re,

deve

lope

da

nonu

nion

and

rem

aine

dsy

mpt

omat

icat

follo

w-u

p.Th

eau

thor

sre

com

men

dth

atth

ese

patie

nts

beco

nsid

ered

for

the

“mos

tef

fect

ive

imm

obili

zatio

n.”

Non

eof

the

patie

nts

unde

rwen

tsu

rgic

alfix

atio

n.

Kor

nber

g,19

86(1

7)R

epor

tof

asi

ngle

case

ofun

stab

leat

las

burs

tfr

actu

re.

IIIA

utho

rfe

els

that

fusi

onis

appr

opri

ate

for

unst

able

burs

tfr

actu

res

ofth

eat

las

(LM

D�

6.9

mm

)an

dde

scri

bes

aca

seof

post

erio

rar

chdi

srup

tion

inw

hich

they

wer

est

illab

leto

perf

orm

C1–

C2

post

erio

rfu

sion

beca

use

one

arch

rem

aine

dco

nnec

ted

tola

tera

lm

ass.

Schl

icke

and

Cal

laha

n,19

81(2

7)R

epor

tof

asi

ngle

case

ofun

stab

leat

las

burs

tfr

actu

re.

IIIC

ase

ofun

stab

lebu

rst

frac

ture

ofth

eat

las

with

LMD

of12

mm

.Th

eau

thor

spr

opos

ea

trea

tmen

tal

gori

thm

ofco

nsid

erin

gsu

rger

yfo

rdi

srup

tion

ofth

etr

ansv

erse

ligam

ent

(LM

D�

6.9

mm

).

aM

RI,

mag

neti

cre

sona

nce

imag

ing;

LMD

,la

tera

lm

ass

disp

lace

men

t;SO

MI,

subo

ccip

ital

-man

dibu

lar

imm

obili

zer.



the rules of Spence (3, 4, 7). Dickman et al. (4) described twotypes of isolated transverse atlantal ligament injuries they iden-tified on MRI. In Type I, the substance of the ligament is injuredwithout associated fracture of the atlas. Type II involves anavulsion fracture of the atlas at the insertion of the transverseatlantal ligament. The authors concluded that patients with TypeI injuries should be treated with early surgical fixation because ofthe inherent instability at C1–C2 after ligamentous disruption.They favor rigid external immobilization for patients with TypeII ligament fracture injures. Applying MRI to their series of 39patients with atlas and/or axis fractures, the authors reportedthat the use of standard cervical x-rays and the rules of Spencewould have failed to identify 60% of the fractures they foundwith associated disruption of the transverse atlantal ligament (asdetermined by MRI) (3).

Lee et al. (20) described 16 patients with atlas fractures.These included six isolated anterior or posterior arch fractures(Landell’s Type I), four burst fractures (Landell’s Type II), andsix lateral mass fractures (Landell’s Type III). Twelve of the 16fracture injuries were isolated atlas fractures and were judgedto be stable as determined by integrity of the transverseligament either by MRI or by LMD criteria. All 12 weresuccessfully treated with rigid collar immobilization for 10 to12 weeks. The authors recommended a treatment algorithm ofcervical immobilization for stable atlas fractures and surgicalfixation and fusion for unstable atlas fractures and unstableC1–C2 combination fracture injuries. Their series, however,did not include any patient with an unstable isolated atlasfracture, nor any patient with an isolated atlas fracture whorequired surgical management.

Unusual isolated atlas fractures have been described in theliterature, often as radiographic curiosities (1, 11, 14, 19, 23).None of the cases for which clinical information was providedneeded surgical treatment. Fractures of the transverse processof the atlas have been described, including one of the casesdescribed by Jefferson in 1920 (2, 15). Although injury to thevertebral artery has been associated with fractures throughthe C1 transverse foramen, the bony C1 injury has not re-quired surgical fixation and has been treated with immobili-zation alone.

SUMMARY

There are no Class I or Class II studies that address themanagement of patients with isolated atlas fractures. All ofthe articles reviewed described case series or case reports thatprovide Class III evidence supporting several treatment strat-

egies for patients with acute C1 fracture injuries. Isolatedanterior or posterior atlas arch fractures and fractures of theatlas lateral mass have been effectively treated with externalcervical immobilization devices. Rigid collars, SOMI braces,and halo ring-vest orthoses have all been used for a durationof treatment of 8 to 12 weeks with good results. No study hasprovided evidence for using one of these devices over theother. Combined anterior and posterior arch fractures of theatlas (burst fractures) with an intact transverse atlantal liga-ment (implying C1–C2 stability) have been effectively man-aged with the use of a rigid collar, a SOMI brace, or a haloorthosis for a duration of 10 to 12 weeks. Combined anteriorand posterior arch fractures of the atlas (burst fractures) withevidence of transverse atlantal ligament disruption have beeneffectively treated with either rigid immobilization alone(halo orthosis) for a period of 12 weeks, or surgical stabiliza-tion and fusion. The type of C1–C2 internal fixation andfusion procedure performed may influence the need for andduration of postoperative immobilization.

Criteria proposed to determine transverse atlantal ligamentinjury with associated C1–C2 instability include the sum ofthe displacement of the lateral masses of C1 on C2 of morethan 8.1 mm on plain x-rays (rules of Spence corrected formagnification), a predental space of more than 3.0 mm inadults, and MRI evidence of ligamentous disruption oravulsion.


The ability to identify the atlas fracture types at greatestrisk of nonunion and subsequent instability is a key issue indetermining appropriate management. Prospective data col-lection generating case-control studies at multiple institutionswould be feasible and useful in examining this issue. Therelative infrequency of isolated atlas fractures would make arandomized study less likely to be implemented. A uniformand clinically useful definition of instability in associationwith isolated atlas fractures would be of benefit. The sub-group of patients with isolated atlas fractures with transverseligament disruption that can be managed either by externalimmobilization alone or surgical fixation and fusion should beexamined in terms of long-term success, economic benefit,patient satisfaction, and return to preinjury activities. Therelative paucity of patients with atlas fractures treated with

TABLE 16.2. Treatment Options for Atlas Fractures

Atlas Fracture Type Treatment Options

Anterior or posterior arch fractures CollarAnterior and posterior arch (burst):

Stable (transverse atlantal ligament intact) Collar, haloUnstable (transverse atlantal ligament disrupted) Halo, C1–C2 stabilization and fusion

Lateral mass fractures:Comminuted fracture Collar, haloTransverse process fractures Collar

Isolated Fractures of the Atlas in Adults S123


surgical stabilization and fusion described in the literaturelimits the ability to address these issues at present.

Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tional Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.

REFERENCES

1. Barker EG Jr, Krumpelman J, Long JM: Isolated fracture of themedial portion of the lateral mass of the atlas: A previouslyundescribed entity. AJR Am J Roentgenol 126:1053–1058, 1976.

2. Contostavlos DL: Massive subarachnoid hemorrhage due to lac-eration of the vertebral artery associated with fracture of thetransverse process of the atlas. J Forensic Sci 16:40–56, 1971.

3. Dickman CA, Sonntag VKH: Injuries involving the transverseatlantal ligament: Classification and treatment guidelines basedupon experience with 39 injuries. Neurosurgery 40:886–887,1997 (letter).

4. Dickman CA, Greene KA, Sonntag VKH: Injuries involving thetransverse atlantal ligament: Classification and treatment guide-lines based upon experience with 39 injuries. Neurosurgery38:44–50, 1996.

5. Fielding JW, Cochran GV, Lawsing JF III, Hohl M: Tears of thetransverse ligament of the atlas: A clinical and biomechanicalstudy. J Bone Joint Surg Am 56A:1683–1691, 1974.

6. Fowler JL, Sandhu A, Fraser RD: A review of fractures of theatlas vertebra. J Spinal Disord 3:19–24, 1990.

7. Greene KA, Dickman CA, Marciano FF, Drabier J, Drayer BP,Sonntag VKH: Transverse atlantal ligament disruption associ-ated with odontoid fractures. Spine 19:2307–2314, 1994.

8. Hadley MN, Dickman CA, Browner CM, Sonntag VKH: Acutetraumatic atlas fractures: Management and long term outcome.Neurosurgery 23:31–35, 1988.

9. Han SY, Witten DM, Mussleman JP: Jefferson fracture of theatlas: Report of six cases. J Neurosurg 44:368–371, 1976.

10. Hays MB, Alker GJ Jr: Fractures of the atlas vertebra: The two-part burst fracture of Jefferson. Spine 13:601–603, 1988.

11. Hays MB, Bernhang AM: Fractures of the atlas vertebra: Athree-part fracture not previously classified. Spine 17:240–242,1992.

12. Heller JG, Viroslav S, Hudson T: Jefferson fractures: The role ofmagnification artifact in assessing transverse ligament integrity.J Spinal Disord 6:392–396, 1993.

13. Highland TR, Salciccioli GG: Is immobilization adequate treat-ment of unstable burst fractures of the atlas? A case report withlong-term follow-up evaluation. Clin Orthop 201:196–200, 1985.

13a. Hinchey JJ, Bickel WH: Fracture of the atlas: Review and pre-sentation of data on eight cases. Ann Surg 121:826–830, 1945.

14. Jakim I, Sweet MB, Wisniewski T, Gantz ED: Isolated avulsionfracture of the anterior tubercle of the atlas. Arch OrthopTrauma Surg 108:377–379, 1989.

15. Jefferson G: Fractures of the atlas vertebra: Report of four cases and areview of those previously reported. Br J Surg 7:407–422, 1920.

16. Kesterson L, Benzel EC, Orrison W, Coleman J: Evaluation andtreatment of atlas burst fractures (Jefferson fractures).J Neurosurg 75:213–220, 1991.

17. Kornberg M: Atypical unstable burst fracture of the atlas:Treated by primary atlantoaxial fusion. Orthop Rev 15:727–729,1986.

18. Landells CD, Van Peteghem PK: Fractures of the atlas: Classifi-cation, treatment and morbidity. Spine 13:450–452, 1988.

19. Lee C, Woodring JH: Unstable Jefferson variant atlas fractures:An unrecognized cervical injury. AJNR Am J Neuroradiol 12:1105–1110, 1991.

20. Lee TT, Green BA, Petrin DR: Treatment of stable burst fractureof the atlas (Jefferson fracture) with rigid cervical collar. Spine23:1963–1967, 1998.

21. Levine AM, Edwards CC: Fractures of the atlas. J Bone JointSurg Am 73A:680–691, 1991.

22. McGuire RA Jr, Harkey HL: Primary treatment of unstable Jef-ferson’s fractures. J Spinal Disord 8:233–236, 1995.

23. Merianos P, Tsekouras G, Koskinas A: An unusual fracture ofthe atlas. Injury 22:489–490, 1991.

24. Oda T, Panjabi MM, Crisco JJ III, Oxland TR: Multidirectionalinstabilities of experimental burst fractures of the atlas. Spine17:1285–1290, 1992.

25. Oda T, Panjabi MM, Crisco JJ III, Oxland TR, Katz L, Nolte LP:Experimental study of atlas injuries: Part II—Relevance to clin-ical diagnosis and treatment. Spine 16[Suppl 10]:S466–S673,1991.

26. Panjabi MM, Oda T, Crisco JJ III, Oxland TR, Katz L, Nolte LP:Experimental study of atlas injuries: Part I—Biomechanical anal-ysis of their mechanisms and fracture patterns. Spine 16[Suppl10]:S460–S465, 1991.

27. Schlicke LH, Callahan RA: A rational approach to burst fracturesof the atlas. Clin Orthop 154:18–21, 1981.

28. Segal LS, Grimm JO, Stauffer ES: Non-union of fractures of theatlas. J Bone Joint Surg Am 69A:1423–1434, 1987.

29. Seljeskog EL: Non-operative management of acute upper cervi-cal injuries. Acta Neurochir (Wien) 41:87–100, 1978.

30. Sherk HH, Nicholson JT: Fractures of the atlas. J Bone Joint SurgAm 52A:1017–1024, 1970.

31. Spence KF Jr, Decker S, Sell KW: Bursting atlantal fracture asso-ciated with rupture of the transverse ligament. J Bone Joint SurgAm 52A:543–549, 1970.

32. Zimmerman E, Grant J, Vise WM, Yashon D, Hunt WE: Treat-ment of Jefferson fracture with a halo apparatus: Report of twocases. J Neurosurg 44:372–375, 1976.


CHAPTER 17

Isolated Fractures of the Axis in Adults

RECOMMENDATIONSFRACTURES OF THE ODONTOID:Standards: There is insufficient evidence to support treatment standards.Guidelines: Type II odontoid fractures in patients 50 years and older should be considered for surgical

stabilization and fusion.Options: Type I, Type II, and Type III fractures may be managed initially with external cervical immobilization.

Type II and Type III odontoid fractures should be considered for surgical fixation in cases of densdisplacement of 5 mm or more, comminution of the odontoid fracture (Type IIA), and/or inability toachieve or maintain fracture alignment with external immobilization.

TRAUMATIC SPONDYLOLISTHESIS OF THE AXIS (HANGMAN’S FRACTURE):Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options: Traumatic spondylolisthesis of the axis may be managed initially with external immobilization in

most cases. Surgical stabilization should be considered in cases of severe angulation of C2 on C3 (FrancisGrade II and IV, Effendi Type II), disruption of the C2–C3 disc space (Francis Grade V, Effendi Type III),or inability to establish or maintain alignment with external immobilization.

FRACTURES OF THE AXIS BODY (MISCELLANEOUS FRACTURES):Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options: External immobilization is recommended for treatment of isolated fractures of the axis body.

RATIONALE

Fractures of the axis represent unique cervical vertebralinjuries owing to the unique anatomy and biomechanicsof the C2 vertebra and the stresses applied to the dy-

namic atlantoaxial complex during trauma. Fractures of theaxis may be associated with other cervical fractures or liga-mentous injuries. Isolated fractures of the axis are commonand warrant independent consideration. Fractures of the axishave been divided into three general subtypes: fractures of theodontoid process, traumatic spondylolisthesis of the axis(hangman’s fractures), and miscellaneous non-odontoid non-hangman’s fractures of the C2 vertebra. Each of these fracturesubtypes has been further subdivided on the basis of theanatomic features and the functional significance of the indi-vidual fracture injury. The purpose of this review is to iden-tify evidence-based management strategies for each injurysubtype of traumatic fractures of the second cervical vertebra.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject heading “spinal cord injury”

in combination with “axis,” “vertebrae,” “fracture,” and “hu-man” yielded 711 articles. Those manuscripts focusing on theclinical management of acute traumatic axis fractures wereselected for review. The bibliographies of these papers werescanned for additional references to confirm completeness ofthe literature review. Relevant papers addressing the mecha-nism of injury or the biomechanics and radiology of the C2vertebra were included. The articles were reviewed and classi-fied using established methodology. Thirty-eight articles forodontoid fracture, 17 for traumatic spondylolisthesis, and 8 formiscellaneous axis fractures provided the basis for the scientificfoundation of this guideline. Data from articles describing axisfractures and/or their management were categorized and areprovided in Tables 17.1 to 17.4). Fifteen additional articles arereferenced in the reference list as supporting information.


Odontoid fractures

Overview

The most common traumatic axis injury is fracture throughthe odontoid process, either through the tip of the dens (Type


I), through its base (Type II), or involving the odontoid butextending into the C2 body (Type III) (1, 6, 40). The anatomyand biomechanics of the C1–C2 complex provide for weight-bearing support for the head on the spine and for the mostmotion of any intervertebral unit in the cervical spine. Motionat C1–C2 is primarily rotational, accounting for one-half of theaxial rotation of the head on the neck (76). Translationalmotion of C1 on C2 is restricted by the transverse atlantalligament that approximates and secures the odontoid processto the anterior arch of the ring of C1. With a fracture of theodontoid process, restriction of translational movement of C1on C2 may be lost. Anterolisthesis or retrolisthesis of theC1-odontoid complex may occur relative to the body of C2. Ifsubstantial subluxation of C1 on C2 occurs, spinal cord injurymay result. The atlantoaxial complex is one of the most com-mon sites of dislocation in fatal cervical spinal injuries (21).

Earlier publications using evidence-based methodology forevaluating the literature on odontoid fracture managementhave focused on fusion as the primary outcome criterion witha minimum follow-up of 18 months (44, 72). Articles on odon-toid fractures containing this information were included inthis survey. Although it has been argued that the radiographicdetermination of fusion may be difficult and subject to ob-server variability, it seems to be the most appropriate out-come measure and is described in most of the clinical articlesaddressing odontoid fractures. It is recognized that outcomemeasures that incorporate patient satisfaction, quality of lifemeasures, and function would perhaps be superior, but thisinformation is sparse and less objective than the fusion criteriadescribed in the literature.

Classification of odontoid fractures

In 1974, Anderson and D’Alonzo (1) classified fractures ofthe odontoid into three types. This categorization has metwith general acceptance and remains in use with minor mod-ification. On the basis of a series of 49 patients managed from1954 through 1972 with an average follow-up of 22 months,

the authors defined three odontoid fracture types. Type Ifractures are oblique fractures through the upper portion ofthe odontoid process. Type II fractures cross the base of theodontoid process at the junction with the axis body. Type IIIfractures are fractures through the odontoid that extend intothe C2 body. The authors considered the Type III odontoidfracture to be more accurately described as a fracture of thebody of the axis. Using this scheme, the authors identified andtreated 2 Type I fractures, 32 Type II fractures, and 15 Type IIIfractures.

In 1988, Hadley et al. (38) added another fracture subtype tothis classification scheme. They described the Type IIA odon-toid fracture as a comminuted fracture involving the base ofthe dens with associated free fracture fragments. The inci-dence of a Type IIA fracture was estimated as 5% of all TypeII fractures (3 of 62 Type II fractures in their series) and wasassociated with severe instability and inability to obtain andmaintain fracture reduction and realignment. The authorsproposed that Type IIA odontoid fractures be managed withearly posterior surgical fixation and fusion of C1–C2.

Treatment

A variety of treatment strategies have been proposed forodontoid fractures based on the fracture type, the degree ofinitial dens displacement, the extent of angulation of the denswith respect to the body of C2, and the age of the patient.These include nonoperative and operative methods (1, 15, 16,30, 34, 38–40, 44, 50, 58, 72). Patients with odontoid fractureinjuries have been treated with external immobilization usinga variety of orthoses with varying results (1, 15, 16, 30, 34,38–40, 44, 50, 58, 72). Surgical options include posterior cer-vical fusion with or without transarticular screw fixation oranterior odontoid screw fixation techniques.

No treatment

In 1985, the Cervical Spine Research Society (16) publisheda multicenter review addressing the management of odontoidfractures. This report includes 18 patients with Type II odon-toid fractures and 3 patients with Type III odontoid fractureswho received no treatment. None of these cases achievedsubsequent bony fusion. The authors concluded that nontreat-ment of odontoid fractures should be eliminated as a man-agement choice.

Traction

Reviews by Traynelis (72) and Julien et al. (44) includeevidentiary tables describing seven articles containing ClassIII medical evidence addressing the treatment of odontoidfractures with traction and subsequent immobilization in acervical collar (1, 15, 16, 30, 34, 50, 58). All patients with TypeI odontoid fractures achieved radiographic fusion (3 of 3patients). Of patients with Type III fractures, 87% (55 of63 patients) achieved fusion. The failure rate for patients withType II fractures treated in this fashion was 43% (42 of 97patients); 57% achieved bony union. It seems that traction andthen cervical collar immobilization may be considered a man-agement option for patients with odontoid fractures, particu-

TABLE 17.1. Initial Management of Isolated Axis Fracture inthe Adult

Fracture Type Treatment Options

Odontoid fractureType I Collar, haloType II Consider for early

surgery or halo, collarType IIA Consider for early

surgery or haloType III Collar, halo, surgical

fusion

Traumatic spondylolisthesis of theaxis (hangman’s fracture)

Effendi Type I, Francis Type I, II Halo, collarEffendi Type II, III, Francis Type III,IV, V

Halo, consider surgicalstabilization

Miscellaneous axis fractures Collar or halo



TAB

LE17

.2.

Sum

mar

yof

Rep

orts

onO

dont

oid

Frac

ture

s

Seri

es(R

ef.

No.

)St

udy

Des

ign

Evid

ence

Cla

ssC

omm

ents

And

erss

onet

al.,

2000

(2)

Ret

rosp

ectiv

eno

nran

dom

ized

repo

rtof

29pa

tient

sag

e�

65w

ithod

onto

id

frac

ture

sm

anag

edw

ithpo

ster

ior

fusi

on,

ante

rior

odon

toid

fixat

ion,

or

imm

obili

zatio

n.

IIIPo

ster

ior

fusi

onre

sulte

din

7/7

fuse

d(1

00%

),an

teri

orod

onto

idsc

rew

resu

lted

in3/

11fu

sed

(27%

),

and

halo

imm

obili

zatio

nre

sulte

din

3/10

fuse

d(3

0%).

The

auth

ors

argu

efo

rpo

ster

ior

fusi

onin

the

elde

rly

patie

ntw

ithan

odon

toid

frac

ture

.

Apf

elba

umet

al.,

2000

(3)

Ret

rosp

ectiv

ere

view

of2-

inst

itutio

nex

peri

ence

with

ante

rior

odon

toid

scre

w

fixat

ion.

147

odon

toid

frac

ture

s(T

ype

II[n

�13

8]an

dTy

peIII

[n�

9])

divi

ded

into

rece

nt(�

6m

o,12

9pa

tient

s)an

dre

mot

e(�

6m

ofr

omin

jury

,18

patie

nts)

grou

ps.

IIITh

efu

sion

rate

was

sign

ifica

ntly

high

erin

the

rece

ntgr

oup

com

pari

ngfu

sion

rate

sof

88%

vers

us25

%

(P�

0.05

)w

itha

mea

nfo

llow

-up

of18

mo.

Apo

sitiv

eco

rrel

atio

nw

asse

enbe

twee

nfu

sion

and

frac

ture

sor

ient

edin

the

hori

zont

alor

post

erio

rob

lique

plan

es.

No

effe

ctof

age,

sex,

num

ber

ofsc

rew

s

plac

ed,

ordi

spla

cem

ent

was

dem

onst

rate

d.

Dai

etal

.,20

00(2

0)R

evie

wof

57ca

ses

offa

iled

man

agem

ent

for

odon

toid

frac

ture

.III

50tr

eate

dw

ithoc

cipi

toce

rvic

alfu

sion

and

7w

ithat

lant

oaxi

alfu

sion

.2

case

sof

nonu

nion

afte

r

atla

ntoa

xial

fusi

onal

one.

38ac

hiev

edan

exce

llent

resu

lt.

Lenn

arso

net

al.,

2000

(46)

33pa

tient

sw

ithis

olat

edTy

peII

odon

toid

frac

ture

str

eate

dw

ithha

love

st

imm

obili

zatio

n.C

ases

defin

edas

nonu

nion

sin

halo

and

cont

rols

defin

edas

unio

ns.

IIC

ase-

cont

rol

Patie

nts

age

�50

yrha

da

risk

for

failu

re21

times

high

erth

anag

e�

50yr

.N

osi

gnifi

cant

diffe

renc

ein

med

ical

cond

ition

s,se

x,am

ount

offr

actu

redi

spla

cem

ent,

dire

ctio

nof

frac

ture

disp

lace

men

t,le

ngth

of

hosp

ital

stay

,or

leng

thof

follo

w-u

pbe

twee

ngr

oups

.

Julie

net

al.,

2000

(44)

Evid

ence

-bas

edre

view

ofm

anag

emen

tof

odon

toid

frac

ture

s.III

Aut

hors

conc

lude

that

ther

eis

insu

ffici

ent

evid

ence

for

stan

dard

sor

guid

elin

es.

Type

Ian

dIII

odon

toid

frac

ture

sca

nbe

man

aged

with

exte

rnal

imm

obili

zatio

n(1

00%

and

84%

fusi

on,

resp

ectiv

ely)

.A

nter

ior

fixat

ion

for

Type

IIIod

onto

idfr

actu

res

appe

ars

toim

prov

eth

efu

sion

rate

tone

arly

100%

.Ty

peII

frac

ture

sca

nbe

man

aged

with

exte

rnal

imm

obili

zatio

nw

ithan

expe

cted

fusi

onra

teof

appr

oxim

atel

y

65%

.Su

rgic

alin

stru

men

tatio

nan

dfu

sion

appe

arto

impr

ove

the

fusi

onra

tean

din

clud

epo

ster

ior

cerv

ical

fusi

on(7

4%)

oran

teri

orod

onto

idsc

rew

fixat

ion

(90%

)w

ithac

cept

able

mor

bidi

ty.

Cam

pane

lliet

al.,

1999

(14)

7pa

tient

sw

ithdi

spla

ced

Type

IIod

onto

idfr

actu

res

unde

rwen

tpo

ster

ior

tran

sart

icul

arsc

rew

fixat

ion.

III6/

7(8

6%)

achi

eved

rigi

dim

mob

iliza

tion.

1ve

rteb

ral

inju

ry.

The

auth

ors

conc

lude

that

this

isa

reas

onab

leop

tion.

Mul

ler

etal

.,19

99(5

6)R

etro

spec

tive

revi

ewof

23pa

tient

s�

70yr

ofag

ew

ithod

onto

idfr

actu

res.

IIIC

ompl

icat

ion

rate

sign

ifica

ntly

incr

ease

din

the

elde

rly

grou

p(5

2%ve

rsus

33%

)pr

imar

ilyow

ing

to

nonu

nion

afte

rno

nope

rativ

etr

eatm

ent.

The

auth

ors

sugg

est

that

the

elde

rly

patie

ntis

athi

ghri

skfo

r

mor

bidi

tyan

dm

orta

lity

and

sugg

est

earl

yha

lofix

atio

nor

prim

ary

stab

iliza

tion.

Mor

andi

etal

.,19

99(5

5)17

case

sof

odon

toid

scre

wfix

atio

n.III

Fusi

onin

16/1

7(9

4%).

The

auth

ors

sugg

est

patie

ntse

lect

ion

for

ante

rior

fixat

ion

beba

sed

onth

e

orie

ntat

ion

ofth

efr

actu

relin

e.

Suba

chet

al.,

1999

(69)

26pa

tient

s(m

ean

age,

35yr

)w

ithTy

peII

frac

ture

str

eate

dw

ithan

teri

orod

onto

id

scre

wfix

atio

n(s

ingl

esc

rew

)pl

usco

llar

(med

ian,

7.2

wk)

.

III25

/26

fusi

on(9

6%).

1pa

tient

requ

ired

post

erio

rfu

sion

for

inad

equa

tere

duct

ion.

Seyb

old

and

Bay

ley,

1998

(67)

Ret

rosp

ectiv

ere

view

of37

Type

IIan

d20

Type

IIIod

onto

idfr

actu

res

divi

ded

into

age

grou

ps.

Age

�60

and

�60

yr.

IIIA

ge�

60:

Type

II,7/

12fu

sed

(58%

).A

ge�

60:

Type

II,8/

11fu

sed

(73%

).Ty

peIII

:al

ltr

eate

dw

ith

halo

,95

%fu

sion

rega

rdle

ssof

type

.Fu

sion

rate

sdi

dno

tdi

ffer

sign

ifica

ntly

betw

een

the

2gr

oups

.Th

e

auth

ors

note

da

decr

ease

dto

lera

nce

inth

eel

derl

ypa

tient

for

halo

imm

obili

zatio

nan

d,as

are

sult,

favo

rsu

rger

yin

sele

cted

case

s.

Jenk

ins

etal

.,19

98(4

3)C

ompa

riso

nof

1ve

rsus

2sc

rew

sin

nons

elec

ted

patie

nts

with

Type

IIod

onto

id

frac

ture

sdi

vide

din

to2

grou

ps:

20(1

scre

w)

vers

us22

(2sc

rew

s).

Follo

w-u

p,9

mo.

IIITh

edi

ffere

nce

infu

sion

rate

,81

%(1

scre

w)

vers

us85

%(2

scre

ws)

,w

asno

tsi

gnifi

cant

.

Ber

lem

ann

and

Schw

arze

nbac

h,19

97(7

)

19pa

tient

s�

65yr

with

Type

IIod

onto

idfr

actu

res

trea

ted

with

ante

rior

odon

toid

scre

wfix

atio

nw

itha

follo

w-u

pof

4.5

yr.

IIIB

ony

fusi

onin

16/1

9(8

4%).

15/1

9(7

9%)

wer

eas

ympt

omat

ic.

The

auth

ors

favo

ran

teri

orfix

atio

nin

this

age

grou

p.

Tray

nelis

,19

97(7

2)Ev

iden

ce-b

ased

revi

ewof

Type

IIod

onto

idfr

actu

res.

IIIFi

rst

evid

ence

-bas

edre

port

onod

onto

idfr

actu

rem

anag

emen

t.In

dica

tes

that

insu

ffici

ent

data

wer

e

avai

labl

efo

rst

anda

rds

orgu

idel

ines

and

that

4tr

eatm

ent

optio

nsfo

rTy

peII

odon

toid

frac

ture

sw

ere

avai

labl

e,in

clud

ing:

trac

tion

follo

wed

byim

mob

iliza

tion,

imm

obili

zatio

nw

ithha

loor

Min

erva

,

post

erio

rce

rvic

alfu

sion

,or

ante

rior

scre

wfix

atio

n.Th

eau

thor

note

sth

atth

ehi

gher

fusi

onra

te

repo

rted

with

ante

rior

scre

wfix

atio

nm

ight

beof

fset

byits

high

erco

mpl

icat

ion

rate

and

lear

ning

curv

e.

Isolated Fractures of the Axis in Adults S127


TAB

LE17

.2.

Con

tinu

ed

Seri

es(R

ef.

No.

)St

udy

Des

ign

Evid

ence

Cla

ssC

omm

ents

Gre

ene

etal

.,19

97(3

6)R

etro

spec

tive

revi

ewof

340

case

sof

axis

frac

ture

s,in

clud

ing

199

odon

toid

frac

ture

s.

IIITy

peI:

2pa

tient

s,2/

2he

aled

with

halo

imm

obili

zatio

n(1

2w

k).

Type

II:12

0pa

tient

s,20

trea

ted

with

earl

ysu

rger

y.8

had

Type

IIw

ith�

6m

m,

4Ty

peIIA

with

inst

abili

tyde

spite

exte

rnal

imm

obili

zatio

n(1

Type

IIAtr

eate

dsu

cces

sful

lyw

itha

halo

),7

patie

nts

unde

rwen

tsu

rgic

alfu

sion

toav

oid

halo

imm

obili

zatio

n.95

wer

etr

eate

dw

ithex

tern

alim

mob

iliza

tion

(med

ian,

13w

k).

88w

ere

avai

labl

efo

r

follo

w-u

p.So

lidfu

sion

faile

din

25(2

8.4%

).7

wer

esu

cces

sful

lytr

eate

dw

ithad

ditio

nal

imm

obili

zatio

n,an

d18

wer

esu

cces

sful

lytr

eate

dw

ithpo

ster

ior

fusi

on(la

tesu

rger

y).

Sign

ifica

ntfa

ctor

was

dens

disp

lace

men

t�

6m

m(�

233

.74,

P�

0.00

01),

givi

ngan

86%

failu

rera

tein

the

halo

trea

tmen

tgr

oup;

5di

ed.

Type

III:

77pa

tient

s,69

man

aged

nono

pera

tivel

yw

ithex

tern

al

imm

obili

zatio

n.68

fuse

d(m

edia

n,12

wk)

.Th

eon

eth

atfa

iled

also

had

aC

1po

ster

ior

arch

frac

ture

and

requ

ired

post

erio

rfu

sion

.6

patie

nts

wer

etr

eate

dw

ithea

rly

surg

ery:

5be

caus

eth

eha

lofa

iled

to

mai

ntai

nal

ignm

ent

and

1be

caus

eof

aco

mbi

ned

C2–

C3

subl

uxat

ion.

2ha

dco

ncom

itant

late

ral

mas

s

frac

ture

sof

the

atla

sw

ithav

ulsi

onof

the

ligam

ento

usin

sert

ion

onth

etu

berc

le;

2di

ed.

Con

clus

ions

:

The

high

est

nonu

nion

rate

was

obse

rved

inTy

peII

odon

toid

disp

lace

d�

6m

m.

Surg

ery

was

reco

mm

ende

dfo

r1)

acut

efr

actu

rein

stab

ility

desp

iteex

tern

alim

mob

iliza

tion,

2)tr

ansv

erse

ligam

ent

disr

uptio

n,an

d3)

Type

IIod

onto

idfr

actu

rew

ith�

6m

mdi

spla

cem

ent.

Polin

etal

.,19

96(6

0)R

etro

spec

tive

revi

ewof

36Ty

peII

frac

ture

str

eate

dw

ithPh

ilade

lphi

aco

llar

(16)

or

halo

vest

imm

obili

zatio

n(2

0).

IIITy

peII:

54%

fuse

dw

ithco

llar.

74%

fuse

dw

ithha

lo.

Chi

baet

al.,

1996

(15)

104

patie

nts

with

odon

toid

frac

ture

s:Ty

peI,

2pa

tient

s.Ty

peII,

62pa

tient

s.Ty

pe

III,

32pa

tient

s.2

grou

ps:

Fres

hgr

oup,

72pa

tient

sw

hose

frac

ture

sw

ere

iden

tifie

d

with

in3

wk

oftr

aum

atic

even

t.O

ldgr

oup,

32pa

tient

sw

hoha

dan

exte

nded

peri

odbe

fore

defin

itive

trea

tmen

t:1

Type

I,21

Type

II,an

d8

Type

III.

IIITy

peI:

2pa

tient

s,co

llar

2/2,

both

fuse

d(1

00%

).Ty

peII:

62pa

tient

s,im

mob

iliza

tion

10/6

2,su

rger

y

52/6

2.In

fres

hfr

actu

regr

oup

trea

ted

with

surg

ery,

31/3

2fu

sed

(97%

).In

dela

yed

frac

ture

grou

p

trea

ted

with

surg

ery,

13/1

9fu

sed

(68%

).Ty

peIII

:32

patie

nts,

surg

ery

15/3

2(4

7%)

fuse

d,

imm

obili

zatio

n17

/32

(53%

)fu

sed,

10/1

5(6

6%)

trea

ted

with

surg

ery

fuse

d.11

/17

(65%

)tr

eate

dw

ith

imm

obili

zatio

nfu

sed.

Ever

ypa

tient

trea

ted

with

aha

lofu

sed,

5/5

(100

%).

Surg

ical

proc

edur

es:

66

patie

nts:

Post

erio

rce

rvic

alfu

sion

:10

patie

nts.

Type

II:9/

9su

cces

sful

fusi

ons

(100

%).

Type

III:

1/1

succ

essf

ulfu

sion

s(1

00%

).A

nter

ior

scre

wfix

atio

n:46

patie

nts.

36Ty

peII,

10Ty

peIII

,42

/45

patie

nts

achi

eved

bony

unio

n(9

3.3%

).Tr

anso

ral

fusi

on:

9pa

tient

s,6/

8Ty

peII.

Succ

essf

ulfu

sion

s(7

5%),

1/1

Type

IIIsu

cces

sful

fusi

on(1

00%

).Th

eau

thor

sof

this

larg

ese

ries

conc

lude

:Ty

peI

frac

ture

sca

n

gene

rally

bem

anag

edno

nope

rativ

ely.

Ant

erio

rsc

rew

fixat

ion

reco

mm

ende

dfo

rm

ost

Type

IIan

d

unst

able

Type

IIIfr

actu

res.

Con

trai

ndic

atio

nsin

clud

eol

des

tabl

ishe

dno

nuni

ons,

irre

duci

ble

frac

ture

s,

caud

aldi

spla

cem

ent,

seve

reos

teop

oros

is.

Type

IIIfr

actu

res

can

betr

eate

dw

ithha

loim

mob

iliza

tion

or

ante

rior

scre

wfix

atio

n.Es

tabl

ishe

dno

nuni

ons

and

irre

duci

ble

frac

ture

ssh

ould

betr

eate

dw

ith

post

erio

rfu

sion

.Tr

anso

ral

fusi

onre

serv

edfo

rra

reca

ses

ofan

teri

orco

rdco

mpr

essi

on.

Bed

nar

etal

.,19

95(5

)Pr

ospe

ctiv

ere

port

ofea

rly

surg

ical

stab

iliza

tion

in11

geri

atri

cpa

tient

sw

ith

odon

toid

frac

ture

s.

IIITh

eau

thor

ssu

gges

tth

atm

orta

lity

can

bere

duce

dby

surg

ical

inte

rven

tion

and

avoi

ding

the

use

of

halo

imm

obili

zatio

n.

Dic

kman

and

Sonn

tag,

1995

(22)

14pa

tient

sw

ithei

ther

acut

eor

suba

cute

Type

IIfr

actu

res

trea

ted

with

ante

rior

odon

toid

scre

wfix

atio

n.R

adio

grap

hic

crite

ria

for

fusi

on:

post

oper

ativ

ex-

rays

and

com

pute

dto

mog

raph

icsc

ans.

IIITy

peII:

14/1

4su

cces

sful

fusi

ons

(100

%).

Dic

kman

etal

.,19

95(2

3)D

escr

ibes

salv

age

proc

edur

esfo

rfa

iled

atla

ntoa

xial

nonu

nion

s.III

Rep

ort

incl

udes

2ca

ses

inw

hich

ante

rior

atla

ntoa

xial

tran

sart

icul

arsc

rew

sw

ere

used

and

8ca

ses

of

post

erio

rtr

ansa

rtic

ular

scre

ws.

Coy

neet

al.,

1995

(18)

15pa

tient

str

eate

dw

ithpo

ster

ior

wir

efu

sion

and

imm

obili

zed

post

oper

ativ

ely

in

eith

erPh

ilade

lphi

aco

llar

orha

lo.

Min

imum

follo

w-u

p,2

yr;

mea

n,4.

7yr

.

Rad

iogr

aphi

ccr

iteri

afo

rfu

sion

:ab

senc

eof

C1–

C2

mov

emen

ton

late

ral

flexi

on/

exte

nsio

nx-

rays

and

evid

ence

ofco

ntin

uity

oftr

abec

ular

bone

form

atio

nbe

twee

n

C1

and

C2

acro

ssth

egr

aft.

IIITy

peII:

13/1

4su

cces

sful

fusi

ons

(93%

).Ty

peIII

:2/

2su

cces

sful

fusi

ons

(100

%).



TAB

LE17

.2.

Con

tinu

ed

Seri

es(R

ef.

No.

)St

udy

Des

ign

Evid

ence

Cla

ssC

omm

ents

Han

igan

etal

.,19

93(4

1)19

patie

nts

�80

yrof

age

with

odon

toid

frac

ture

s(1

6Ty

peII,

3Ty

peIII

).III

5pa

tient

sw

ithdi

spla

cem

ent

�5

mm

requ

ired

post

erio

rsu

rgic

alfix

atio

nw

ithgo

odre

sults

.Th

ere

was

27%

mor

talit

yin

the

cons

erva

tive

trea

tmen

tgr

oup,

with

prol

onge

dim

mob

iliza

tion

note

das

one

ofth

e

cont

ribu

ting

fact

ors.

Wad

dell

and

Rea

rdon

,

1983

(74)

24pa

tient

sw

ithod

onto

idfr

actu

re:

20Ty

peII

and

4Ty

peIII

frac

ture

s.16

/20

Type

IIfr

actu

res

wer

etr

eate

dw

ithC

1–C

2ar

thro

desi

s(G

allie

proc

edur

e).

All

Type

III

frac

ture

sw

ere

trea

ted

nono

pera

tivel

y.

IIITy

peII:

15/1

6su

cces

sful

fusi

ons

(94%

);1

patie

ntw

aslo

stto

follo

w-u

p.Ty

peIII

:3/

4su

cces

sful

fusi

ons

(75%

),1/

4no

nuni

on(2

5%).

Rya

nan

dTa

ylor

,19

93

(63)

30pa

tient

s�

60yr

ofag

ew

ithTy

peII

frac

ture

s.III

The

fusi

onra

tein

the

patie

nts

age

�60

trea

ted

with

imm

obili

zatio

nw

ason

ly7/

29(2

3%).

Des

pite

the

low

fusi

onra

tefo

rth

isag

egr

oup,

the

auth

ors

favo

rha

loim

mob

iliza

tion

over

surg

ical

fixat

ion.

Buc

holz

,19

81(1

2)26

patie

nts;

0Ty

peI,

17Ty

peII,

9Ty

peIII

.Pa

tient

sw

ere

imm

obili

zed

inha

lofo

r

am

inim

umof

3m

oan

d,if

nom

ovem

ent

onfle

xion

/ext

ensi

onx-

rays

,pl

aced

ina

Phila

delp

hia

colla

rfo

ran

addi

tiona

l4

wk.

Rad

iogr

aphi

ccr

iteri

afo

rfu

sion

:no

mov

emen

tor

subl

uxat

ion

atth

efr

actu

resi

teon

flexi

on/e

xten

sion

x-ra

ys.

IIITy

peII:

15/1

7su

cces

sful

fusi

ons

(88%

).2/

17no

nuni

ons

(12%

).Ty

peIII

:9/

9su

cces

sful

fusi

ons

(100

%).

3de

aths

:2

patie

nts

had

Type

IIfr

actu

res

whi

lebe

ing

trea

ted

inha

lo,

and

1pa

tient

with

Type

III

frac

ture

.

Had

ley

etal

.,19

88(3

8)R

etro

spec

tive

stud

yin

clud

ing

62pa

tient

sw

ithTy

peII

odon

toid

frac

ture

s,in

clud

ing

3w

ithco

mm

inut

ion

atth

eba

se.

IIITh

esu

bgro

upof

Type

IIod

onto

idfr

actu

rew

ithco

mm

inut

ion

atth

eba

sew

asde

fined

asth

eTy

peIIA

odon

toid

frac

ture

.Th

ecl

inic

alsi

gnifi

canc

eof

this

obse

rvat

ion

was

that

the

frac

ture

fuse

dpo

orly

with

imm

obili

zatio

nan

dw

asco

nsid

ered

for

earl

ysu

rger

y.

Gov

ende

ran

d

Gro

otbo

om,

1988

(34)

Rev

iew

of41

patie

nts

with

odon

toid

frac

ture

s:26

Type

II,15

Type

III.

1m

oin

trac

tion

(2–4

kg),

then

ari

gid

colla

rfo

r6–

8w

k,an

das

sess

edat

3m

o.

Rad

iogr

aphi

ccr

iteri

afo

rfu

sion

:bo

nyco

ntin

uity

acro

ssfr

actu

resi

tean

dno

mov

emen

ton

flexi

on/e

xten

sion

tom

ogra

ms.

IIITy

peII:

19/2

6su

cces

sful

fusi

ons

(73%

).2/

26fib

rous

unio

ns(8

%).

5/26

nonu

nion

s(1

9%).

Type

III:

15/1

5su

cces

sful

fusi

ons

(100

%).

No

mor

talit

y.7

halo

pin-

site

infe

ctio

ns.

3pa

tient

sha

dsk

in

exco

riat

ion

over

chin

seco

ndar

yto

halte

rtr

actio

n.

Fujii

etal

.,19

88(3

0)R

etro

spec

tive

revi

ewof

52pa

tient

sw

ithod

onto

idfr

actu

res,

incl

udin

gda

taon

24

trea

ted

with

imm

obili

zatio

n,10

trea

ted

with

ante

rior

scre

wfix

atio

n,an

d7

trea

ted

with

post

erio

rfu

sion

.R

adio

grap

hic

crite

ria

for

fusi

on:

ante

ropo

ster

ior

and

late

ral

tom

ogra

ms.

IIIIm

mob

iliza

tion:

Type

I:1/

1su

cces

sful

fusi

on(1

00%

).Ty

peII:

3/7

succ

essf

ulfu

sion

s(4

3%).

Type

III:

10/1

4su

cces

sful

fusi

ons

(72%

).Po

ster

ior

fusi

on:

Type

II:7/

7su

cces

sful

fusi

ons

(100

%).

Ant

erio

rsc

rew

fixat

ion:

Type

II:6/

8su

cces

sful

fusi

ons

(75%

).Ty

peIII

:2/

2su

cces

sful

fusi

ons

(100

%).

Lind

etal

.,19

87(4

8)R

evie

wof

14pa

tient

sw

ithod

onto

idfr

actu

res

man

aged

with

halo

imm

obili

zatio

n

and

eval

uate

dat

12w

kw

ithfle

xion

/ext

ensi

onx-

rays

.In

clud

ed9

Type

IIan

d5

Type

IIIfr

actu

res

with

a2-

yrfo

llow

-up.

Rad

iogr

aphi

ccr

iteri

afo

rfu

sion

:fle

xion

-

exte

nsio

nx-

rays

.

III10

/11

succ

essf

ulfu

sion

s(9

1%)

com

bine

dTy

peII

and

Type

IIIfr

actu

res.

Aut

hors

supp

ort

the

use

of

halo

imm

obili

zatio

nas

the

initi

altr

eatm

ent

for

Type

IIan

dIII

odon

toid

frac

ture

s.

Dun

nan

dSe

ljesk

og,

1986

(24)

Ret

rosp

ectiv

ere

port

of80

patie

nts

with

odon

toid

frac

ture

sin

clud

ing

data

on74

patie

nts

trea

ted

prim

arily

with

rigi

dbr

acin

gfo

r3–

6m

ofo

llow

edby

addi

tiona

l

colla

rsu

ppor

tfo

r6

wk

and

41pa

tient

sun

derg

oing

post

erio

rce

rvic

alfu

sion

.

Min

imum

follo

w-u

ppe

riod

was

6m

o;80

%of

the

patie

nts

had

follo

w-u

plo

nger

than

8m

o.R

adio

grap

hic

crite

ria

for

fusi

on:

late

ral

flexi

on-e

xten

sion

x-ra

ysat

3–4

mo.

IIIR

igid

imm

obili

zatio

n:Ty

peII:

40/5

9su

cces

sful

fusi

ons

(68%

).19

/59

nonu

nion

s(3

2%).

Type

III:

15/1

5

succ

essf

ulfu

sion

s(1

00%

).Po

ster

ior

fusi

on:

40/4

1su

cces

sful

fusi

ons

(98%

)fo

rco

mbi

ned

Type

IIan

d

Type

IIIfr

actu

res.

Cla

rkan

dW

hite

,19

85

(16)

Mul

ticen

ter

revi

ewin

clud

ing

144

patie

nts

man

aged

by27

diffe

rent

surg

eons

.

Fusi

onra

tes

repo

rted

base

don

frac

ture

type

and

trea

tmen

t.R

adio

grap

hic

crite

ria

for

fusi

on:

evid

ence

oftr

abec

ulat

ion

acro

ssth

efr

actu

resi

tean

dab

senc

eof

mov

emen

t

onla

tera

lfle

xion

-ext

ensi

onx-

rays

.

IIIN

otr

eatm

ent:

Type

II:0/

18su

cces

sful

fusi

ons

(0%

).Ty

peIII

:0/

3su

cces

sful

fusi

ons

(0%

).C

olla

r:Ty

pe

II:0/

3su

cces

sful

fusi

ons

(0%

).Typ

eIII

:5/

10su

cces

sful

fusi

ons

(50%

).Tr

actio

n:Ty

peII:

2/3

succ

essf

ul

fusi

ons

(66%

).Ty

peIII

:7/

8su

cces

sful

fusi

ons

(88%

).H

alo:

Type

II:25

/38

succ

essf

ulfu

sion

s(6

6%).

Type

III:

13/1

6su

cces

sful

fusi

ons

(81%

).A

nter

ior

fusi

on:

Type

II:7/

8su

cces

sful

fusi

ons

(88%

).Ty

peIII

:

2/2

succ

essf

ulfu

sion

s(1

00%

).Po

ster

ior

fusi

on:

Type

II:25

/26

succ

essf

ulfu

sion

s(9

6%).

Type

III:

4/4

succ

essf

ulfu

sion

s(1

00%

).



TAB

LE17

.2.

Con

tinu

ed

Seri

es(R

ef.

No.

)St

udy

Des

ign

Evid

ence

Cla

ssC

omm

ents

Pepi

net

al.,

1985

(58)

Ret

rosp

ectiv

ere

view

of41

patie

nts

with

odon

toid

frac

ture

sin

clud

ing

26tr

eate

d

cons

erva

tivel

yw

ithto

ngs,

four

-pos

ter

brac

e,co

llars

,an

d/or

halo

vest

s(0

Type

I,13

Type

II,13

Type

III).

12pa

tient

sun

derw

ent

post

erio

rce

rvic

alfu

sion

(1Ty

peI,

4

Type

II,an

d7

Type

III).

Rad

iogr

aphi

ccr

iteri

afo

rfu

sion

:un

ion

onpl

ain

x-ra

ysan

d

tom

ogra

ms

asw

ell

asla

tera

lfle

xion

-ext

ensi

onvi

ews.

Non

unio

nw

asde

fined

as

mov

emen

tof

the

dens

frag

men

ton

late

ral

flexi

on-e

xten

sion

x-ra

ys.

IIIH

alo/

trac

tion:

Type

II:6/

13su

cces

sful

fusi

ons

(46%

).Ty

peIII

:11

/13

succ

essf

ulfu

sion

s(8

5%).

Post

erio

r

cerv

ical

fusi

on:

Type

I:1/

1su

cces

sful

fusi

ons

(100

%).

Type

II:4/

4su

cces

sful

fusi

ons

(100

%).

Type

III:

7/7

succ

essf

ulfu

sion

s(1

00%

).Th

eau

thor

sno

ted

that

the

halo

vest

sw

ere

poor

lyto

lera

ted

inpa

tient

s

age

�75

yr.

Wan

get

al.,

1984

(75)

Ret

rosp

ectiv

ere

view

of25

patie

nts

with

odon

toid

frac

ture

str

eate

dw

itha

vari

ety

of

cerv

ical

imm

obili

zatio

nte

chni

ques

.

IIITy

peI:

1/1

fuse

dw

ithha

lo(1

00%

).Ty

peII:

4/7

fuse

dw

ithco

llar

(57%

).4/

5fu

sed

inha

lo(8

0%).

Type

III:

2/2

fuse

dw

ithco

llar

(100

%).

10/1

2fu

sed

with

halo

(83%

).

Boh

ler,

1982

(9)

15pa

tient

s.W

ithod

onto

idfr

actu

res,

both

acut

ean

dch

roni

ctr

eate

dw

ithan

teri

or

scre

wfix

atio

nfo

llow

edby

ape

riod

ofce

rvic

alfix

atio

nin

apl

astic

colla

rfo

ra

peri

odof

4–16

wk.

Frac

ture

dist

ribu

tion:

0Ty

peI,

8Ty

peII,

and

7Ty

peIII

.

Rad

iogr

aphi

ccr

iteri

afo

rfu

sion

:no

tgi

ven.

IIITy

peII:

8/8

succ

essf

ulfu

sion

s(1

00%

).Ty

peIII

:7/

7su

cces

sful

fusi

ons

(100

%).

Mai

man

and

Lars

on,

1982

(49)

Ret

rosp

ectiv

ere

view

of49

case

sof

odon

toid

frac

ture

,in

clud

ing

34Ty

peII

frac

ture

str

eate

dw

ithea

rly

post

erio

rw

ire/

graf

tst

abili

zatio

n.Po

st-o

pera

tive

imm

obili

zatio

nw

itha

Min

erva

for

anav

erag

eof

5w

k.2

Type

IIIfr

actu

res

wer

e

incl

uded

.R

adio

grap

hic

crite

ria

for

nonu

nion

:to

mog

raph

icev

iden

ceof

avas

cula

r

necr

osis

,gr

oss

inst

abili

tyw

itha

dem

onst

rabl

ega

pat

the

frac

ture

line,

and

no

evid

ence

ofhe

alin

g.Fu

sion

resu

ltsev

alua

ted

6m

opo

stsu

rger

y.

IIITh

eau

thor

sob

serv

eda

100%

fusi

onra

teat

the

post

erio

rsu

rgic

alsi

te,

but

only

a35

%fu

sion

rate

acro

ssth

efr

actu

resi

te.

Rya

nan

dTa

ylor

,19

82

(62)

Ret

rosp

ectiv

ere

view

of23

patie

nts

with

odon

toid

frac

ture

sov

era

10-y

rpe

riod

,

incl

udin

g1

Type

I,16

Type

II,an

d6

Type

III.

Rad

iogr

aphi

ccr

iteri

afo

rfu

sion

:no

mov

emen

ton

late

ral

flexi

on-e

xten

sion

x-ra

ys.

IIIH

alo/

Min

erva

/Sub

occi

pita

l-m

andi

bula

rim

mob

ilize

r:Ty

peI:

1/1

succ

essf

ulfu

sion

.Ty

peII:

9/15

succ

essf

ulfu

sion

s(6

0%).

Type

III:

6/6

succ

essf

ulfu

sion

s(1

00%

).

Ekon

get

al.,

1981

(26)

Ret

rosp

ectiv

ere

view

of22

case

sof

odon

toid

frac

ture

trea

ted

with

halo

imm

obili

zatio

nfo

r3

mo.

Type

I,0

patie

nts.

Type

II,16

patie

nts.

Type

III,

6

patie

nts.

Incl

udes

outc

ome

on17

patie

nts

with

anav

erag

efo

llow

-up

of30

mo.

Rad

iogr

aphi

ccr

iteri

afo

rfu

sion

:la

tera

lfle

xion

-ext

ensi

onx-

rays

.

IIITy

peII:

6/12

succ

essf

ulfu

sion

s(5

0%).

Type

III:

4/5

succ

essf

ulfu

sion

s(8

0%).

Mar

aran

dTa

y,19

76(5

0)R

evie

wof

26ca

ses

ofod

onto

idfr

actu

rein

clud

ing

24Ty

peII

and

2Ty

peIII

trea

ted

with

cerv

ical

trac

tion

for

�10

wk.

Rad

iogr

aphi

ccr

iteri

afo

rfu

sion

:fib

rous

unio

nat

frac

ture

site

.

IIITy

peII:

9/24

succ

essf

ulfu

sion

s(3

7.5%

).Ty

peIII

:2/

2su

cces

sful

fusi

ons

(100

%).

And

erso

nan

dD

’Alo

nzo,

1974

(1)

Ret

rosp

ectiv

ere

view

of49

patie

nts

with

odon

toid

frac

ture

scl

assi

fied

into

Type

I,II,

and

IIIba

sed

onfr

actu

re.

IIIN

on-o

pera

tive

trea

tmen

t:37

patie

nts.

Type

I:co

llar/

brac

e,2/

2su

cces

sful

fusi

ons

(100

%).

Type

II:ha

lo,

14/2

2su

cces

sful

fusi

ons

(64%

).8/

22no

nuni

ons

(36%

).Ty

peIII

:ha

lo,

12/1

3su

cces

sful

fusi

ons

(92%

).

1/13

nonu

nion

(8%

).O

pera

tive

trea

tmen

t:12

patie

nts.

Type

II:8/

10su

cces

sful

fusi

ons

(80%

).Ty

peIII

:

2/2

(100

%).



TAB

LE17

.3.

Sum

mar

yof

Rep

orts

onTr

aum

atic

Spon

dylo

listh

esis

ofth

eA

xis

Seri

es(R

ef.

No.

)St

udy

Des

ign

Evid

ence

Cla

ssC

omm

ents

Bar

ros

etal

.,19

99(4

)C

ase

repo

rtof

surg

ical

fixat

ion

inha

ngm

an’s

frac

ture

.III

Surg

ical

trea

tmen

tfo

rha

ngm

an’s

frac

ture

isan

optio

n.

Ver

hegg

enan

dJa

nsen

,19

98(7

3)R

etro

spec

tive

stud

yof

16pa

tient

str

eate

dw

ithea

rly

post

erio

rsc

rew

fixat

ion

ofth

e

neur

alar

chaf

ter

hang

man

’sfr

actu

re.

IIITh

eau

thor

ssu

gges

tth

atth

isis

the

optim

alth

erap

yfo

rEd

war

dsan

dLe

vine

(Effe

ndi)

Type

IIan

dIII

frac

ture

s,de

scri

bing

exce

llent

resu

ltsin

thei

rse

ries

.

Gre

ene

etal

.,19

97(3

6)34

0ca

ses

ofax

isfr

actu

res,

incl

udin

g74

patie

nts

with

trau

mat

icsp

ondy

lolis

thes

isof

the

axis

.Fo

llow

-up

avai

labl

eon

72pa

tient

s.

IIIM

ost

com

mon

:Ef

fend

iTy

peI

(72%

),Fr

anci

sG

rade

I(6

5%).

65tr

eate

dsu

cces

sful

lyw

ith

imm

obili

zatio

n(1

2w

k).

7re

quir

edea

rly

surg

ery

(pos

teri

orfu

sion

)ow

ing

topo

oral

ignm

ent

inth

eha

lo

(Effe

ndi

II,6

patie

nts;

Effe

ndi

III,

1pa

tient

;Fr

anci

sG

rade

I,1

patie

nt;

II,1

patie

nt;

III,

2pa

tient

s;IV

,3

patie

nts)

.33

%of

all

Effe

ndi

Type

sII

and

IIIan

d36

%of

all

Fran

cis

Type

sIII

,IV

,V

patie

nts

requ

ired

surg

ery.

Stro

ngco

rrel

atio

nob

serv

edbe

twee

nEf

fend

iI

and

Fran

cis

Ian

dEf

fend

iIII

and

Fran

cis

IV.

Con

clus

ions

:Im

mob

iliza

tion

isge

nera

llysu

ffici

ent

trea

tmen

t.Su

rger

ym

aybe

cons

ider

edfo

rse

vere

Fran

cis

orEf

fend

ity

peha

ngm

an’s

frac

ture

s.

Cor

icet

al.,

1996

(17)

Ret

rosp

ectiv

ere

view

ofha

ngm

an’s

frac

ture

incl

udin

g39

nond

ispl

aced

(�6

mm

C2

onC

3)tr

eate

dw

ithno

nrig

idim

mob

iliza

tion

(Phi

lade

lphi

aco

llar

for

anav

erag

eof

12

wk)

and

10di

spla

ced

(�6

mm

)tr

eate

dw

ithha

lo(3

),co

llar

(6),

orsu

rger

y(1

).

IIIN

ondi

spla

ced

grou

p:39

/39

fuse

dus

ing

colla

ral

one.

Dis

plac

edgr

oup:

also

fuse

dre

gard

less

of

trea

tmen

t.C

1–C

3fu

sion

requ

ired

in1

patie

ntfo

rfa

ilure

ofcl

osed

redu

ctio

n.

Star

ran

dEi

smon

t,19

93(6

8)R

evie

wof

19ca

ses

ofax

isfr

actu

rein

clud

ing

6ca

ses

ofa

patte

rnoc

curr

ing

thro

ugh

the

post

erio

ras

pect

ofth

eve

rteb

ral

body

cont

inui

tyof

the

post

erio

rco

rtex

with

subl

uxat

ion

resu

lting

inna

rrow

ing

ofth

esp

inal

cana

l.

IIIH

angm

an’s

frac

ture

vari

atio

noc

curr

edin

6/19

patie

nts,

incl

udin

g2

with

spin

alco

rdin

jury

from

the

asso

ciat

edsu

blux

atio

n.

Tan

and

Bal

acha

ndra

n,19

92(7

0)R

etro

spec

tive

stud

yof

33pa

tient

sw

ithha

ngm

an’s

frac

ture

.C

lass

ified

byEf

fend

i:21

Type

I,11

Type

II,an

d1

Type

III.

III20

/26

had

none

urol

ogic

alde

ficit

onad

mis

sion

.28

/33

with

com

plet

ere

cove

ryaf

ter

1yr

.

Torr

eman

,19

90(7

1)Lo

ng-t

erm

stud

yof

23pa

tient

sw

ithha

ngm

an’s

frac

ture

str

eate

dw

ithim

mob

iliza

tion.

Ave

rage

follo

w-u

p,9.

6yr

.

III10

0%lo

ng-t

erm

fusi

onra

tew

ithce

rvic

alim

mob

iliza

tion.

Gov

ende

ran

dC

harl

es,

1987

(33)

Pros

pect

ive

stud

yof

39pa

tient

s.III

All

patie

nts

succ

essf

ully

man

aged

with

trac

tion

and

imm

obili

zatio

n.

Gra

dyet

al.,

1986

(35)

Ret

rosp

ectiv

ere

view

of27

patie

nts

incl

udin

g16

man

aged

with

halo

,8

with

aco

llar,

and

3w

ithbe

dres

t.

IIIA

llac

hiev

edfu

sion

with

nore

sidu

alsy

mpt

oms.

The

auth

ors

reco

mm

end

the

use

ofa

Phila

delp

hia

colla

ral

one

infr

actu

res

with

min

imal

disp

lace

men

t.

Levi

nean

dEd

war

ds,

1985

(47)

Ret

rosp

ectiv

eca

sese

ries

of52

patie

nts

with

trau

mat

icsp

ondy

lolis

thes

isof

the

axis

clas

sifie

dus

ing

the

Effe

ndi

crite

ria.

IIIIs

olat

edTy

peI,

II,an

dIIa

wer

eal

lm

anag

edno

nope

rativ

ely.

3of

5Ty

peIII

patie

nts

unde

rwen

tsu

rgic

al

stab

iliza

tion

for

failu

reto

obta

inor

mai

ntai

nre

duct

ion

ina

halo

.Th

eau

thor

sid

entif

yth

eTy

peIIa

subg

roup

ofth

eEf

fend

iTy

peII

patie

nts

who

dist

ract

sign

ifica

ntly

with

the

appl

icat

ion

oftr

actio

nan

d

note

the

mec

hani

smof

inju

ryfo

rth

isgr

oup

islik

ely

flexi

on-d

istr

actio

n.3/

3Ty

peIIa

patie

nts

wer

e

trea

ted

with

gent

leex

tens

ion

and

com

pres

sion

unde

rflu

oros

copi

cgu

idan

cefo

llow

edw

ithha

lo

imm

obili

zatio

n.

Bor

neet

al.,

1984

(10)

Ret

rosp

ectiv

ere

view

of18

case

sof

“ped

icle

”fr

actu

reof

the

axis

trea

ted

with

dire

ct

inte

rnal

fixat

ion.

IIIA

ggre

ssiv

esu

rgic

alap

proa

chfo

rfix

atio

nof

pedi

cle-

isth

mus

frac

ture

sof

the

axis

with

100%

fusi

onra

te.

Mes

tdag

het

al.,

1984

(52)

Com

bine

dcl

inic

alan

dan

atom

icst

udy

desc

ribi

ng41

frac

ture

sof

the

post

erio

rne

ural

arch

ofth

eax

is.

11ca

ses

trea

ted

with

ante

rior

C2–

C3

inte

rbod

yfu

sion

.30

trea

ted

with

trac

tion

and

imm

obili

zatio

n.Fo

llow

-up

avai

labl

eon

30pa

tient

s.

IIIC

adav

eric

stud

yde

mon

stra

ted

that

frac

ture

sw

ithdi

spla

cem

ent

of�

5m

mw

ere

stab

le.

Cer

vica

l

mob

ility

was

mai

ntai

ned

bette

rin

the

cons

erva

tive

man

agem

ent

grou

p.Th

eau

thor

sre

com

men

d

cons

erva

tive

mea

sure

sex

cept

inca

ses

ofm

arke

din

stab

ility

orno

nuni

on.

Fran

cis

etal

.,19

81(2

9)C

lass

ifica

tion

pape

rba

sed

on12

3ca

ses

offr

actu

res

ofth

epo

ster

ior

arch

ofth

eax

is.

Gra

deba

sed

ondi

spla

cem

ent

and

angu

latio

n.

IIIG

rade

I(1

5%of

tota

lse

ries

):0%

nonu

nion

with

imm

obili

zatio

n.G

rade

II(7

%):

33%

nonu

nion

.G

rade

III(3

7%):

0%no

nuni

on.

Gra

deIV

(34%

):2%

nonu

nion

.G

rade

V(6

%):

28%

nonu

nion

.

Buc

holz

,19

81(1

2)A

utop

syst

udy

of17

0ca

ses

oftr

aum

atic

deat

h.III

38ha

dce

rvic

alsp

ine

frac

ture

san

d8/

38ha

dtr

aum

atic

spon

dylo

listh

esis

ofth

eax

is.

Pepi

nan

dH

awki

ns,

1981

(57)

Def

ined

anea

rly

clas

sific

atio

nsc

hem

efo

rha

ngm

an’s

frac

ture

base

don

42ca

ses.

IIITy

peI:

ano

ndis

plac

edfr

actu

reof

the

post

erio

rel

emen

ts.

Type

II:di

spla

ced

frac

ture

invo

lvin

gpo

ster

ior

and

ante

rior

stru

ctur

es.

All

42pa

tient

sw

ere

succ

essf

ully

trea

ted

nons

urgi

cally

.

Effe

ndi

etal

.,19

81(2

5)C

lass

ifica

tion

pape

rba

sed

on13

1ca

ses

ofpa

tient

sw

ithfr

actu

res

ofth

eri

ngof

the

axis

.Fr

actu

res

divi

ded

into

3gr

oups

base

don

mec

hani

smof

inju

ry,

disp

lace

men

t,

and

stab

ility

.

IIITy

peI

(65%

ofto

tal

grou

p):

Isol

ated

hair

line

frac

ture

sof

the

ring

ofth

eax

isw

ithm

inim

aldi

spla

cem

ent

ofth

ebo

dyof

C2

caus

edby

axia

llo

adin

gan

dhy

pere

xten

sion

.Ty

peII

(28%

):D

ispl

acem

ent

ofan

teri

or

frag

men

tw

ithdi

srup

tion

ofth

edi

scsp

ace

belo

wth

eax

isca

used

byhy

pere

xten

sion

and

rebo

und

flexi

on.

Type

III(7

%).

Dis

plac

emen

tof

ante

rior

frag

men

tw

ithC

2–C

3fa

cet

disl

ocat

ion

caus

edby

prim

ary

flexi

onan

dre

boun

dex

tens

ion.

Alth

ough

5pa

tient

sun

derw

ent

surg

ery,

the

auth

ors

conc

lude

that

the

vast

maj

ority

ofth

ese

patie

nts

are

best

man

aged

with

cerv

ical

imm

obili

zatio

n.

Bra

shea

ret

al.,

1975

(11)

29pa

tient

sw

ithha

ngm

an’s

frac

ture

sfo

llow

edfo

ran

aver

age

of6

yr.

IIIN

oca

seof

neur

olog

ical

defic

it.23

/23

(100

%)

trea

ted

with

imm

obili

zatio

nac

hiev

edfu

sion

.Su

ppor

ts

nono

pera

tive

man

agem

ent.



larly those with Type I and Type III fractures. The low fusionsuccess rate reported for Type II odontoid fractures managedwith traction and collar immobilization (57%) implies thatperhaps collar immobilization is not the ideal strategy forType II fracture patients.

Cervical collar

Several authors have proposed treatment of odontoid frac-tures with cervical collars. Polin et al. (60) in 1996 describe aseries of 36 Type II fractures treated with either a Philadelphiacollar or halo vest immobilization. The fusion rate was lowerin the patients treated with collars (53%) compared with 74%for patients managed in halos. An earlier report from the sameinstitution described a similar rate of fusion (57%) in a studyincluding seven Type II fractures treated with a collar alone(75). The infrequent Type I odontoid fracture seems to have anacceptable rate of fusion with rigid cervical collar immobili-zation, approaching 100% in one study (1, 15, 16). Type IIIodontoid fractures have been treated with cervical collars aswell, but they have a less favorable union rate, with fusionrates ranging from 50 to 65% in small series (16, 75).

Halo immobilization

In a series of publications resulting in the largest institu-tional series of axis fractures published to date, 340 cases ofaxis fractures were reviewed, including 199 odontoid frac-tures (2 Type I, 116 Type II, 4 Type IIA, 77 Type III (36, 39, 40).Excellent results were obtained with rigid external immobili-zation in the Type I and Type III fracture patients (2 of 2 and68 of 69 patients with successful fusion, respectively). Of theType II patients, 95 were treated with external immobilizationfor a median of 13 weeks. The authors reported a 28% failurerate. Seven failures were successfully treated with additionalexternal immobilization, and 18 patients underwent subse-quent posterior C1–C2 fusion. The authors found that a dis-placement of the dens of 6 mm or more was associated with ahigh nonunion rate, irrespective of patient age, direction ofdisplacement, or neurological deficit (86% failure rate; �2

33.74; P � 0.001). The degree of dens displacement and anegative correlation with fusion was noted by at least fourother investigators (16, 24, 26, 48). The amount of odontoiddisplacement observed ranged from 2 to 6 mm in thesestudies.

TABLE 17.4. Summary of Reports on Miscellaneous Axis Fractures

Series (Ref. No.) Study DesignEvidence

ClassComments

Greene et al., 1997 (36) 340 cases of axis fractures, including 67non-odontoid, non-hangman’s fractures(miscellaneous), most involving the bodyof lateral masses.

III 60/61 (98%) were successfully treated withexternal mobilization in all but 1 patient (1.6%nonfusion rate). 4 patients died, and 1underwent early surgery for 5-mm luxation ofC2 on C3.

Fujimura et al., 1996 (31) 31 cases of axis body fracturescategorized into 4 types based onradiographic imaging.

III 4 types: Avulsion: 9/9 fused withimmobilization. Transverse: 2/2 healed withimmobilization. Burst: 2/3 treated with C2–C3fusion. Sagittal fractures: 15/17 healed withimmobilization. 8 sagittally oriented fracturepatients still had pain despite a bony union.

Benzel et al., 1994 (6) Retrospective report of 15 patientsdescribed with fractures of the axis body.

III The authors propose classification into: Type 1:coronal (n � 12). Type 2: sagittal (n � 3). Type3: oblique and equivalent to the Type IIIodontoid fracture.

Korres et al., 1994 (45) 14 cases of avulsion fracture of theanteroinferior portion of the axissecondary to extension-type injuries.Mean follow-up, 8.5 yr.

III 3% of the cervical spine trauma cases over a12-yr period. All patients treated successfullywith cervical immobilization.

Bohay et al., 1992 (8) Describes 3 cases of vertical fractures ofthe axis.

III Notes that this is an unusual variant fracture ofthe axis body. All treated with immobilization.

Craig and Hodgson, 1991(19)

Describes 9 cases of superior facetfracture of the axis vertebra.

III 5 treated with reduction and immobilization. 3required open reduction and posterior fusion.

Burke and Harris, 1989(13)

Review of 165 patients with axisfractures. 31 miscellaneous bodyfractures identified and classified onmechanism of injury.

III Identified 31 patients with axis body fractures.21/38 (68%) were extension teardrop; 10/31(32%) were hyperextension.

Jakim and Sweet, 1988(42)

Case report of a transverse fracture of theaxis and literature review. Aclassification scheme is proposed.

III 3 types of axis body fractures were described:the Type III odontoid fracture of Anderson andD’Alonzo, the transverse body fracture, and theavulsion fracture.



Julien et al. (44) reviewed nine articles that dealt withtreatment of odontoid fractures (total, 269 patients) usinghalo/Minerva fixation for 8 to 12 weeks (12, 15, 16, 24, 26, 30,48, 58, 62). All patients with Type I odontoid fractures werefound to have successful fusion (3 of 3 patients) (15, 30, 62).One hundred sixty-eight patients with Type II odontoid frac-tures were treated with halo immobilization; 110 (65%) hadsuccessful fusion. There was a 30% nonunion rate (50 of 168patients). Eight patients were described as having a malunion.Of patients with Type III odontoid fractures, 84% (67 of 80patients) achieved a solid fusion. There was an 8% failure rate(6 of 80 patients), and 7 cases were described as malunions.The authors of these series generally concluded that rigidexternal immobilization can be considered a viable treatmentoption for Type I, Type II, and Type III odontoid fractures.Rigid external immobilization seems to be most successful forpatients with Type I, Type III, and nondisplaced Type IIodontoid fractures, but it should be considered with cautionin elderly patients.

Posterior cervical fixation

Posterior cervical fixation and fusion has been successfullyused in the treatment of acute traumatic odontoid fractures.Although no criteria defining the indications for surgical fix-ation have been established, a number of retrospective caseseries suggest treatment options (15, 16, 18, 24, 30, 49, 58, 74).These papers describe a total of 147 patients who underwentposterior cervical fixation and fusion for Type II odontoidfractures and 29 patients treated similarly for Type III frac-tures. One patient with a Type I fracture was treated success-fully with posterior fusion. The overall fusion rates for Type IIand Type III fractures managed with surgical fixation andfusion were 87 and 100%, respectively, in these series. Thereport of Maiman and Larson (49) described a fusion rate ofonly 35% across the fracture line but a fusion rate of 100% atthe posterior operative site.

The aforementioned series typically describe an instru-mented (wire or cable) posterior C1–C2 arthrodesis and thencervical immobilization in a rigid orthosis. More recently,transarticular screw fixation and fusion of C1–C2 has beenused for traumatic odontoid fractures, particularly in cases offailed fusion after initial management (14, 43). The reportedsurgical morbidity and mortality is 2 to 4% and includesfailure of fracture reduction, vertebral artery injury, and thenew onset of neurological deficit. Loss of motion at the atlan-toaxial joint after posterior C1–C2 fusion results from of dor-sal C1–C2 arthrodesis. Despite this, several authors favorposterior C1–C2 fusion rather than anterior odontoid screwfixation as the ideal treatment of unstable odontoid fractures(2, 14, 56).

Anterior cervical fixation

Anterior single and double screw fixation of odontoid frac-tures has been accomplished with success. The technical chal-lenges associated with this procedure have limited wide-spread application. If successful, this technique has thepotential to maintain rotational motion at the atlantoaxial

joint. It has been suggested that this is an appropriate strategywhen the odontoid fracture line is either horizontal or obliqueand posterior and that it is contraindicated in situations wherethe fracture line is oblique and anterior (2, 3, 20, 55). In casesof transverse atlantal ligament disruption, anterior screw fix-ation can result in an unsatisfactory outcome despite union ofthe odontoid fracture owing to persistent transverse atlantalligament incompetence. Julien et al. (44) summarized a seriesof articles that describe retrospective experiences with ante-rior screw fixation for odontoid fractures (9, 15, 22, 30, 43). Thecombined fusion rate of Type II fractures treated in thesereports is 89% (112 of 126 patients). Patients with Type IIIodontoid fractures achieved radiographic fusion in 20 of 20patients (100%). In a recent series reported by Subach et al.(69), 26 patients with Type II fractures (mean age, 35 yr)underwent anterior odontoid fixation with a single screw andthen immobilization in a cervical collar (median, 7.2 wk).Twenty-five (96%) of 26 patients achieved successful fusion.The one failure was attributed to inadequate fracture reduc-tion. That patient required subsequent posterior C1–C2 fu-sion. Jenkins et al. (43), in 1998, described a retrospectivenonrandomized series of 42 patients undergoing anteriorscrew fixation for Type II odontoid fractures. The authorscompared single-screw with two-screw techniques. The fu-sion rate in their experience was similar for single-screwfixation (81%) compared with two-screw fixation (85%). Useof lag screws to achieve anterior odontoid fixation is recom-mended. Complications of the procedure include retropha-ryngeal wall injury, screw fracture, infection, and screw mis-placement with injury to surrounding vascular and neuralstructures (9, 22, 30). Attempts at anterior odontoid fixationusing a transoral approach was associated with multiple sig-nificant complications (15).

Apfelbaum et al. (3) compared anterior screw fixation forrecent and remote odontoid fractures at two institutions. Onehundred forty-seven patients with Type II (n � 138) and TypeIII (n � 9) odontoid fractures underwent anterior screw fixa-tion either within 6 months of injury (129 patients) or morethan 18 months after injury (18 patients). The fusion rateswere 88% in the �6-month group versus 25% in the remotefracture injury group (P � 0.05), with a mean follow-up of 18months. A positive correlation was identified between fusionand fractures oriented in the horizontal or posterior obliqueplanes. No effect of age, sex, number of screws placed, ordegree of dens displacement was identified. Their experiencesuggests that anterior odontoid screw fixation for odontoidfractures is most effective when performed early after injury,particularly within 6 months of fracture.

Odontoid fracture management in the elderly patient

One of the controversial issues in the management of odon-toid fractures is the influence of age on treatment selection. Anumber of studies have examined the circumstance of acuteodontoid fracture in the older patient. Three case series argueagainst surgical fixation in the elderly patient (36, 62, 67).Seven other case series favor surgical fixation in this agegroup. There is also one case-control study by Lennarson et al.



(46) providing Class II medical evidence for surgical treat-ment of elderly patients. Ryan and Taylor (63) described 30patients 60 years and older with Type II odontoid fractures.The fusion success rate in patients older than 60 years treatedwith external immobilization was only 23%. The authorsthought that the high fracture nonunion rate was secondary toinadequate immobilization and delays in diagnosis in mostcases. If these issues were eliminated, no significant differencein outcome between surgical and nonsurgical managementwould have been demonstrated. They concluded that surgicalfixation and fusion for elderly patients with odontoid frac-tures should be reserved for unusual circumstances. Greene etal. (36) reported the largest series (120 patients) of retrospec-tively reviewed cases of traumatic odontoid Type II axis frac-tures. Patients with dens displacement of 6 mm or more intheir experience had a nonunion rate of 86%, compared witha nonunion rate of 18% for patients with displacement of lessthan 6 mm. The authors reported no significant relationshipbetween fracture nonunion and age using �2 analysis. It mightbe argued that statistical tests of association would be moreappropriate in this circumstance, and age might have beenshown to be a factor had it been used.

Andersson et al. (2) described 29 patients 65 years and olderwith odontoid fractures managed by surgical and nonsurgicalmeans. In their series, six (86%) of seven patients achievedsuccessful fusion after posterior cervical C1–C2 arthrodesis.Worse results were observed in patients treated with anteriorodontoid screw fixation (20% fusion rate) and in patientsmanaged with external immobilization alone (20% fusionrate). These authors favored posterior cervical fusion overother management options in elderly patients with Type IIodontoid fractures. Pepin et al. (58) reported their experiencewith 41 acute odontoid fractures (1 Type I, 19 Type II, 21 TypeIII). The authors found that halo immobilization was poorlytolerated in patients 75 years and older. They suggested thatearly C1–C2 fixation and fusion was appropriate in thisgroup. Hanigan et al. (41) described 19 patients 80 years andolder with odontoid fractures (16 Type II, 3 Type III). Fivepatients with displacement of more than 5 mm were treatedwith posterior cervical fixation and fusion with good results.Three of the five had stable nonunions. The authors reporteda mortality rate of 26% in patients managed conservativelywith prolonged immobilization rather than surgical fixationand fusion. On the other hand, they noted that no patienttreated with external immobilization alone developed clini-cally significant instability. Pitzen et al. (59) described theirexperience with surgical therapy in seven patients 70 yearsand older with odontoid fractures. Two patients died of re-lated medical comorbidity. Five patients did well and weremobilized early. The authors concluded that early surgicalfixation in this age group is the preferred management strat-egy. This view is shared by several other investigators, includ-ing Seybold and Bayley (67), Campanelli et al. (14), and Mul-ler et al. (56). Bednar et al. (5) reported a prospectiveassessment of elderly patients with odontoid fractures man-aged with early operative stabilization and fusion. Elevenpatients were included in their study. The authors found a91% fusion rate (10 of 11 patients). One patient died of unre-

lated causes. The authors argued in favor of early surgicalintervention for elderly patients with odontoid fractures. In1997, Berleman and Schwarzenbach (7) offered a retrospectivereview of their experience with 19 patients 65 years and olderwith Type II odontoid fractures treated with anterior odon-toid screw fixation. Radiographic fusion with nearly 5-yearfollow-up was obtained in 16 (85%) of 19 patients. The authorsconcluded that anterior odontoid screw fixation is a successfultherapy for elderly patients with Type II odontoid fractures.

In the only case-control Class II evidence study publishedon this topic, Lennarson et al. (46) examined 33 patients withisolated Type II odontoid fractures treated with halo vestimmobilization. The authors found that age older than 50years was a significant factor for failure of fusion in a haloimmobilization device. Patients 50 years and older had a riskfor nonunion 21 times higher than that found for patientsyounger than 50 years. No significant effect on outcome wasfound attributable to other medical conditions, sex of thepatient, degree of fracture displacement, direction of fracturedisplacement, length of hospital stay, or length of follow-up.

Traumatic spondylolisthesis of the axis(hangman’s fracture)

Overview

Traumatic fractures of the posterior elements of the axis,often related to hyperextension injuries from motor vehicleaccidents, diving, and falls, are reminiscent of the injury in-duced to the axis by judicial hangings (65, 78). A distinctionhas been made between the two fracture types because themechanisms of injury are different. The mechanism of injuryassociated with judicial hanging is one of distraction andhyperextension. The more common hangman’s fracture injuryinduced by motor vehicular trauma is typically a result ofhyperextension, compression, and possible rebound flexion.The incidence of head injury is high with the latter hangman’sfracture injury type.

Wood-Jones (78) described the cervical fracture-dislocationinjury induced by hanging in 1913. Garber (32) used the term“traumatic spondylolisthesis” of the axis in 1964. He de-scribed eight patients with “pedicular” fractures of the axisafter motor vehicle accidents. The term “hangman’s fracture”has been attributed to Schneider (65), who described a seriesof eight patients and noted the similarity between the fractureof the posterior elements of the axis to the pattern of frac-ture injury induced by judicial hanging. Williams (77) docu-mented four cases of hangman’s fracture injury in 1975, not-ing that three occurred associated with motor vehicleaccidents and the fourth with a fall. A number of authors havesuggested that a more appropriate term for this axis injurytype may be “traumatic spondylolisthesis of the axis” becauseof the differences in the mechanism of injury between hangingand the deceleration injuries of falls and motor vehicle acci-dents (28, 29). Most traumatic spondylolisthesis fractures ofC2 caused by motor vehicle accidents seem to result fromhyperextension and compression, rather than the hyperexten-sion and distraction associated with hangings. These differ-ences in the mechanism of injury, along with the wide range



of neurological deficits identified with these injuries,prompted a series of investigators to attempt to better char-acterize and classify traumatic spondylolisthesis injuries ofthe axis.

Classification of hangman’s fractures

In 1981, Pepin and Hawkins (57) published a two-typeclassification scheme for hangman’s fractures. Type I wasdescribed as a nondisplaced fracture of the posterior elementsalone. Type II was a displaced fracture involving the posteriorelements and the body of C2. The authors successfully treated42 patients without surgery using their scheme, which in-volved reduction (Type II injuries) and immobilization. Theynoted a low incidence of associated spinal cord injury, but afrequent association with head injury. Although simple andeffective, Pepin and Hawkins’ scheme has not gained popularacceptance and is not widely used. In the same year, Franciset al. (29) published a collaborative experience in treating 123patients with traumatic spondylolisthesis of the axis. Injurieswere divided into five grades based on displacement andangulation of C2 on C3. Grade I was defined as displacementof less than 3.5 mm and angulation of less than 11 degrees.Grade V was defined as complete C2–C3 disc disruption.Grade IV in their scheme had more than 3.5 mm of C2–C3disruption, but less than half of C3 vertebral width with morethan 11 degrees of C2–C3 angulation. Grades II and III wereinjury types graded between Grades I and IV.

Effendi et al. (25) described three types of fractures of thering of the axis based on a series of 131 patients. Their clas-sification scheme was based on the mechanism of injury: TypeI, axial loading and hyperextension; Type II, hyperextensionand rebound flexion; Type III, primary flexion and reboundextension. Type I fractures were defined as isolated hairlinefractures of the ring of the axis with minimal displacement ofthe body of C2. Type II fractures were defined as displace-ment of the anterior fragment with disruption of the discspace below the axis. Type III fractures were defined as dis-placement of the anterior fragment with the body of the axisin a flexed position in conjunction with C2–C3 facet disloca-tion. This Type III fracture is associated with a flexed forwardposition of the axis body. The incidence of Type I, II, and IIIfracture injury in their series was 65, 28, and 7%, respectively.Levine and Edwards (47) modified Effendi’s classificationscheme in 1985. They added flexion-distraction as a mecha-nism of injury (Type IIA) and offered a tailored treatmentstrategy for each of the four injury types. In the largest seriesof axis fractures yet described, Greene et al. (36) used theclassification schemes of both Effendi et al. and Francis et al.to characterize 74 hangman’s fractures. The most commonfracture pattern identified was the Effendi Type I (72%) andthe Francis Grade I (65%). The investigators found a strongcorrelation between Effendi Types I and III and FrancisGrades I and IV, respectively.

Not all authors think that all hangman’s fractures fit intoone or both of these classification schemes. In the review byBurke and Harris (13) of 165 acute injuries of the axis vertebra,62 (38%) were traumatic spondylolisthesis of the axis, includ-

ing 13 Effendi I, 35 Effendi II, and 3 Effendi III injuries. Elevenpatients (18%) had a fracture pattern not previously describedin which one or both fractures involved a portion of theposterior cortex of the body of the axis.

Incidence of traumatic spondylolisthesis andassociated injuries

In Greene et al.’s (36) series of 1820 cervical fractures, 340(19%) were fractures of the axis and 74 (4%) were hangman’stype. In the series of acute fractures of the axis vertebradescribed by Burke and Harris (13), injuries of the axis wereassociated with other fractures of the cervical vertebra in 8%of cases. Ryan and Henderson (61) studied 657 patients withcervical spine fractures over a 13-year period. Hangman’s-type fractures occurred as isolated fractures in 74% of theirseries. Only 9% were associated with fractures of C1. Anadditional 9% were associated with subaxial cervical spinefractures. In the series of Guiot and Fessler (37) of 10 complexcombined atlantoaxial fractures, only one involved a hang-man’s injury. Although the incidence of spinal cord and nerveroot injury as a result of a hangman’s fracture is reportedlylow, unstable hangman’s injuries do occur with some fre-quency (12, 57). If the patient survives the initial injury, it hasbeen proposed that the relatively spacious intracanicular di-ameter affords some protection against spinal cord compres-sion (54). Starr and Eismont (68) described an atypical fracturepattern occurring through the posterior aspect of the vertebralbody, with continuity of the posterior cortex or pedicle andnarrowing of the spinal canal as a result of the associatedsubluxation. In their series of 19 patients, this hangman’sfracture variant occurred in six patients, including two pa-tients with resultant paralysis. In the series described by Fran-cis et al. (29), 8 (6%) of 123 patients they managed had neu-rological deficits. Tan and Balanchandran’s (70) retrospectiveseries of 33 hangman’s fractures included 14 patients withno neurological deficit at admission. The other 19 (57%) hadneurological deficits ranging from quadriparesis to urinaryretention. Twenty-eight patients (85%) had returned to em-ployment at the 1-year follow-up. Mirvis et al.’s (53) seriesof 27 patients had associated neurological findings in 26% ofpatients with hangman’s fractures. Combination fracturesof C1 and C2 in association with a hangman’s-type C2 injuryseem to have a higher incidence of associated neurologicalinjury, likely because of increased instability and a moresevere traumatic injury pattern (23, 37).

Treatment

Most patients with traumatic hangman’s fractures reportedin the reviewed literature were treated with cervical immobi-lization with good results. The three largest experiences re-ported are the multi-institutional series of Effendi et al. (25)and Francis et al. (29) and the single institutional experiencedescribed by Greene et al. (36). Management strategies andsurgical indications vary somewhat among investigators.

In the series reported by Effendi et al. (25) in 1981, therewere 85 Type I fractures, 62 of which were managed withexternal immobilization. They reported 37 Type II fractures;



17 of the patients were treated with bracing and 15 withsurgical fusion. Of the patients managed surgically, 4 weretreated with a C2–C3 anterior fusion and 11 were treated withdorsal internal fixation and fusion. Nine patients had EffendiType III fractures. Three died without definitive treatment,one was managed in a brace, and five were treated surgicallywith fusion, one anterior and four posterior. The authorsconcluded that most hangman’s fractures were best managednonoperatively. They commented that they might have over-treated patients early in their series, offering surgery whenexternal immobilization might well have been successful.They decided that surgery should be reserved for unusualType III fractures and those patients with failure of fusiondespite 3 months of halo immobilization.

In Pepin and Hawkins’ (57) series, also reported in 1981, all42 patients with hangman’s fractures they treated healed suc-cessfully with external immobilization alone. Francis et al. (29)described and classified hangman’s fracture injuries in 123patients from four institutions. Injuries were categorized intoGrades I through V on the basis of displacement and angula-tion. There were 19 Grade I, 9 Grade II, 46 Grade III, 42 GradeIV, and 7 Grade V fractures. All patients were initially man-aged with traction with conversion to a halo orthosis or weretreated in a halo vest without traction. Healing occurred in116 patients (95%) with halo immobilization alone. Sevenpatients received surgical management with fusion for non-union despite halo immobilization (four had an anteriorC2–C3 fusion, two had a posterior C1–C3 fusion, and one hada posterior C2–C4 fusion). The authors assessed the injurytype with respect to success with nonoperative management.Three (33%) of 9 Grade II injury patients and 2 (28%) of 7Grade V injury patients developed nonunion despite halomanagement and required subsequent surgical treatment.Halo treatment failed in none of the Grade I and Grade IIIinjury patients and in only one Grade IV injury patient. Theauthors concluded that primary surgical treatment for hang-man’s fracture injuries is not indicated. All patients should beprovided late follow-up to assess for nonunion, particularlyGrade II and Grade V injury patients. When surgical manage-ment is considered, the authors recommended an anteriorC2–C3 fusion.

In Levine and Edwards’ (47) series of 52 patients withhangman’s fractures, all isolated Type I, II, and IIa injurieswere managed nonoperatively. Three of five Type III patientsunderwent surgical stabilization and fusion for failure to ob-tain or maintain fracture reduction in a halo. The authorssingled out the subgroup of the Effendi Type II fracture thatsignificantly distracted with the application of craniocervicaltraction. They thought that Type II injuries were likely theresult of flexion-distraction forces. The three patients withType II fractures in their series were successfully treated withmild compression-extension in a halo vest under fluoroscopiccontrol (47).

Greene et al. (36) noted a strong correlation between EffendiType I and Francis Grade I hangman’s injury and betweenEffendi Type III and Francis Grade IV fractures in their seriesof 74 patients. Sixty-five of 74 patients were treated nonop-eratively with external immobilization for a median of 12

weeks. There were two early deaths. Seven patients requiredearly surgical fixation and fusion for inability to maintainfracture alignment in a halo brace. All seven early surgicalpatients were either Effendi Grade II or III, and five of theseven were Francis Grade III or IV. Overall, 33% of patientswith unstable Effendi Types II and III or 36% of FrancisGrades III, IV, and V injuries required early surgical treat-ment. Eventually, all seven patients achieved solid fusionwithout evidence of instability. The authors compared theirexperiences with those of Effendi et al. and Francis et al. andconcluded that conservative management (external immobi-lization) should be the initial treatment in virtually everypatient with a hangman’s fracture. Early surgical manage-ment should be reserved for unstable injuries that are ineffec-tively immobilized in a halo device.

In a combined clinical and cadaveric anatomic study, Mest-dagh et al. (52) described their experience with 41 fractures ofthe posterior neural arch of the axis. Eleven cases were treatedsurgically with anterior C2–C3 interbody fusion, and 30 pa-tients were treated with external immobilization. Thirty pa-tients were available for follow-up. Cervical mobility wasbetter in patients managed conservatively. Displacement ofup to 5 mm at the hangman’s fracture site in a cadaveric studywas compatible with stability without disruption of the liga-ments or the C2–C3 disc space. The authors concluded thatconservative management was the ideal treatment for hang-man’s fractures, except in cases of marked instability or fail-ure of union. Grady et al. (35) reported their experience with27 patients with hangman’s fractures. Sixteen were managedin a halo device, eight in a rigid collar, and three with bed restonly. All achieved fusion with no residual symptoms. Theauthors concluded that use of a Philadelphia collar alone forhangman’s fractures is a reasonable alternative to halo immo-bilization, particularly for injuries with minimal C2–C3 dis-placement. In 1987, Govender and Charles (33) prospectivelystudied 39 patients with traumatic spondylolisthesis of theaxis. Injuries were classified for stability by the criteria ofWhite and Panjabi (76). All patients were successfully treatedwith collar immobilization regardless of assessment of stabil-ity. The authors argue against basing treatment on dynamicimaging, as advocated by Effendi et al. (25) and Levine andEdwards (47). A number of other reports favor nonoperativemanagement of hangman’s fractures (4, 11, 17, 27, 33, 47, 50,52, 66, 71).

Surgical management

Surgical options for unstable hangman’s fracture injuries,particularly those that fail to heal despite external immobili-zation, include anterior C2–C3 interbody fusion and dorsalC1–C3 fusion procedures. In the series of Effendi et al. (25), 42of 131 patients with hangman’s fractures were treated surgi-cally. Ten were treated with an anterior C2–C3 fusion, and 32underwent a posterior fusion. All were successfully stabilizedat last follow-up. In the Francis et al. (29) series of 123 hang-man’s fracture patients, only 7 patients were treated surgi-cally. Four underwent anterior C2–C3 fusion, two had poste-rior C1–C3 fusion, and one underwent posterior C2–C4



fusion. The authors noted that 6 of the 7 patients requiringfusion for nonunion had C2–C3 angulation of more than 11degrees. All seven patients achieved bony stability.

A number of case series of hangman’s fractures offer similarexperiences with surgical management. McLaurin et al. (51)described their experience with early fusion in two patientswith hangman’s fractures to allow early mobilization. Theauthors acknowledged that both injuries would likely havehealed with external immobilization alone. Salmon (64) de-scribed 20 patients with hangman’s fractures treated withposterior interlaminar wiring and fusion with no morbidity.Verheggen and Jansen (73), in their 1998 report, arguedstrongly for surgical fixation and fusion of Effendi Type II andIII hangman’s fractures. In their opinion, the optimal manage-ment of these injuries remains controversial. They described16 patients with hangman’s fractures they treated with surgi-cal fixation of the posterior arch of the axis with screw fixa-tion. They found that this fixation technique resulted in su-perior functional results as compared with historical controls.They favor this management strategy in the setting of theLevine and Edwards (47) Type IIa fracture. The viewpoint ofVerheggen and Jansen (73) is challenged by Sypert (69a) in hiscomments that accompany their article. Borne et al. (10), in1984, published their approach to the management of pedic-ular fractures of the axis. They used a technique of bilateralposterior screw fixation. They described excellent results anda 100% fusion rate. Despite this, their technique has notgained widespread acceptance.

Fractures of the axis body

A number of authors have addressed the management ofnon-odontoid, non-hangman’s fractures of the axis. They havebeen labeled as miscellaneous fractures of the axis, non-odontoid non-hangman’s fractures, or simply axis body frac-tures (6, 31, 36, 40). Several attempts have been made toclassify the various fracture types within this diverse group.Benzel et al. (6) reported on 15 patients with fracture of theaxis body and divided them into three types: coronal, sagittal,and transverse oriented. The latter group was thought torepresent the same group as the Anderson and D’AlonzoType III odontoid fracture. The authors proposed that theType III odontoid fracture classification be discarded becauseit is misleading (6). The original authors had the same thought(1). Benzel et al. (6) offered a mechanism of injury for each ofthe three fracture types they described. No treatment or out-come data were included in their report. Greene et al. (36)described 67 patients with miscellaneous axis fractures of alltypes. Of the 61 patients available for follow-up (medianfollow-up, 14 mo), all but one was successfully managed witha variety of nonoperative means. The authors note that this isa diverse injury group and describe a treatment algorithmbased on features of fracture stability. Only one patient witha miscellaneous axis fracture required surgical interventionfor delayed nonunion. Fujimura et al. (31) classified 31 axisbody fractures on the basis of their radiographic injury pat-tern: avulsion, transverse, burst, or sagittal. In their series, allnine cases of avulsion fracture and the two cases of transverse

fracture healed with external immobilization. Two of the threeburst fractures were treated with C2–C3 anterior interbodyfusion. Of the 17 sagittal fractures, 15 healed with nonopera-tive treatment. The remaining two patients required surgicalfusion. The authors recommend initial nonoperative treat-ment for all non-odontoid, non-hangman’s axis fractures.Craig and Hodgson (19) added nine cases of axis fracturesinvolving the superior articular facet. In seven patients, therewas an associated odontoid fracture. This fracture occurred ineither the coronal or sagittal plane, resulting in shearing of theanterior or lateral portion of the facet complex. The lateralmass of the atlas was noted to occasionally sublux into thefacet fracture. The authors recommended immobilization fornondisplaced fractures and the consideration of surgical re-duction, fixation, and fusion for fractures that are difficult toreduce. Bohay et al. (8) described three unusual fractures ofthe posterior body of C2, all of which responded to nonop-erative management. Jakim and Sweet (42) contributed a sin-gle case. Korres et al. (45) described 14 patients with avulsionfractures of the anteroinferior portion of the axis that theybelieved to be extension-type injuries. These cases repre-sented only 3% of the cervical spine fractures they managedover a 12-year period. All 14 of these body fracture types weresuccessfully managed with cervical immobilization (meanfollow-up, 8.5 yr).

SUMMARY

Fractures of the odontoid

There is no Class I medical evidence addressing the issue ofmanagement of acute traumatic odontoid fractures. A singleClass II evidence paper reviews the management of Type IIodontoid fractures in halo immobilization devices. This studydemonstrated a 21-fold increase in risk of nonunion with haloimmobilization in patients older than 50 years. All other arti-cles reviewed contain Class III evidence that supports severaltreatments.

Type II odontoid fractures in patients 50 years and oldershould be considered for surgical stabilization and fusion.Type I, Type II, and Type III fractures may be managedinitially with external cervical immobilization. Type II andType III odontoid fractures should be considered for surgicalfixation in cases of dens displacement of 5 mm or more,comminution of the odontoid fracture (Type IIA), and/orinability to achieve or maintain fracture alignment with ex-ternal immobilization. Isolated Type I and Type III odontoidfractures may be treated with cervical immobilization, result-ing in fusion rates of 100 and 84%, respectively. Anteriorsurgical fixation of Type III fractures has been associated witha 100% fusion rate. Type II odontoid fractures may be treatedwith external immobilization or surgical fixation and fusion.Halo immobilization and posterior fixation have both beenused successfully for these injuries. Anterior odontoid-screwfixation has been reported with an up to 90% fusion successrate, except in older patients. Treatment of Type II odontoidfracture with a cervical collar alone or traction and thencervical collar immobilization may also be undertaken, butthese methods have lower success rates.



Traumatic spondylolisthesis of the axis

There is no Class I or Class II medical evidence addressingthe management of traumatic spondylolisthesis of the axis.All articles reviewed contain Class III evidence that supportsa variety of treatments. Most hangman’s fractures heal with12 weeks of cervical immobilization with either a rigid cervi-cal collar or a halo immobilization device. Surgical stabiliza-tion is an option in cases of severe angulation (Francis GradeII and IV, Effendi Type II), disruption of the C2–C3 disc space(Francis Grade V, Effendi Type III), or the inability to establishor maintain fracture alignment with external immobilization.

Fractures of the axis body (miscellaneousaxis fractures)

There is no Class I or Class II medical evidence addressing themanagement of traumatic fractures of the axis body. All articlesreviewed contain Class III evidence that supports the use ofexternal immobilization as the initial treatment strategy.


More data are necessary to determine treatment standardsand/or guidelines for the definitive management of odontoidfractures. For Type I and Type III fractures, the available ClassIII evidence suggests that a well-designed multicenter case-control study could provide sufficient evidence to define theirappropriate management in the early postinjury period. ForType II fractures, the literature suggests that both operativeand nonoperative management remain treatment options. Arandomized or case-control study would be of benefit inestablishing definitive treatment recommendations for thisfracture type. Traumatic spondylolisthesis of the axis andmiscellaneous axis fractures are treated successfully with ex-ternal immobilization in most cases. A multicenter case-control study of patients with these injury types would help todefine optimal treatment of each specific fracture subtype.


REFERENCES

1. Anderson LD, D’Alonzo RT: Fractures of the odontoid process ofthe axis. J Bone Joint Surg Am 56A:1663–1674, 1974.

2. Andersson S, Rodrigues M, Olerud C: Odontoid fractures: Highcomplication rate associated with anterior screw fixation in theelderly. Eur Spine J 9:56–60, 2000.

3. Apfelbaum RI, Lonser RR, Veres R, Casey A: Direct anteriorscrew fixation for recent and remote odontoid fractures.J Neurosurg 93[Suppl 2]:227–236, 2000.

4. Barros TE, Bohlman HH, Capen DA, Cotler J, Dons K, Biering-Sorensen F, Marchesi DG, Zigler JE: Traumatic spondylolisthesisof the axis: Analysis of management. Spinal Cord 37:166–171,1999.

5. Bednar DA, Parikh J, Hummel J: Management of Type II odon-toid process fractures in geriatric patients: A prospective studyof sequential cohorts with attention to survivorship. J SpinalDisord 8:166–169, 1995.

6. Benzel EC, Hart BL, Ball PA, Baldwin NG, Orrison WW,Espinosa MC: Fractures of the C-2 vertebral body. J Neurosurg81:206–212, 1994.

7. Berlemann U, Schwarzenbach O: Dens fractures in the elderly:Results of anterior screw fixation in 19 elderly patients. ActaOrthop Scand 68:319–324, 1997.

8. Bohay D, Gosselin RA, Contreras DM: The vertical axis fracture:A report on three cases. J Orthop Trauma 6:416–419, 1992.

9. Bohler J: An approach to non-union of fractures. Surg Annu14:299–315, 1982.

10. Borne GM, Bedou GL, Pinaudeau M: Treatment of pedicularfractures of the axis: A clinical study and screw fixation tech-nique. J Neurosurg 60:88–93, 1984.

11. Brashear R Jr, Venters G, Preston ET: Fractures of the neural archof the axis: A report of twenty-nine cases. J Bone Joint Surg Am57A:879–887, 1975.

12. Bucholz RW: Unstable hangman’s fractures. Clin Orthop 154:119–124, 1981.

13. Burke JT, Harris JH Jr: Acute injuries of the axis vertebra. Skel-etal Radiol 18:335–346, 1989.

14. Campanelli M, Kattner KA, Stroink A, Gupta K, West S: Poste-rior C1–C2 transarticular screw fixation in the treatment of dis-placed Type II odontoid fractures in the geriatric population:Review of seven cases. Surg Neurol 51:596–601, 1999.

15. Chiba K, Fujimura Y, Toyama Y, Fujii E, Nakanishi T,Hirabayashi K: Treatment protocol for fractures of the odontoidprocess. J Spinal Disord 9:267–276, 1996.

16. Clark CR, White AA III: Fractures of the dens: A multicenterstudy. J Bone Joint Surg Am 67A:1340–1348, 1985.

17. Coric D, Wilson JA, Kelly DL Jr: Treatment of traumatic spon-dylolisthesis of the axis with nonrigid immobilization: A reviewof 64 cases. J Neurosurg 85:550–554, 1996.

18. Coyne TJ, Fehlings MG, Wallace MC, Bernstein M, Tator CH:C1–C2 posterior cervical fusion: Long-term evaluation of resultsand efficacy. Neurosurgery 37:688–693, 1995.

19. Craig JB, Hodgson BF: Superior facet fractures of the axis verte-bra. Spine 16:875–877, 1991.

20. Dai LY, Yuan W, Ni B, Liu HK, Jia LS, Zhao DL, Xu YK: Surgicaltreatment of nonunited fractures of the odontoid process, withspecial reference to occipitocervical fusion for unreducible atlan-toaxial subluxation or instability. Eur Spine J 9:118–122, 2000.

21. Davis D, Bohlman H, Walker AE, Fisher R, Robinson R: Thepathological findings in fatal craniospinal injuries. J Neurosurg34:603–613, 1971.

22. Dickman CA, Sonntag VKH: Surgical management of atlantoax-ial nonunions. J Neurosurg 83:248–253, 1995.

23. Dickman CA, Foley KT, Sonntag VKH, Smith MM: Cannulatedscrews for odontoid fixation and atlantoaxial transarticularscrew fixation. J Neurosurg 83:1095–1100, 1995.

24. Dunn ME, Seljeskog EL: Experience in the management of odon-toid process injuries: An analysis of 128 cases. Neurosurgery18:306–310, 1986.

25. Effendi B, Roy D, Cornish B, Dussault RG, Laurin CA: Fracturesof the ring of the axis: A classification based on the analysis of131 cases. J Bone Joint Surg Br 63B:319–327, 1981.

26. Ekong CE, Schwartz ML, Tator CH, Rowed DW, Edmonds VE:Odontoid fracture: Management with early mobilization usingthe halo device. Neurosurgery 9:631–637, 1981.

27. Ersmark H, Lowenhielm P: Factors influencing the outcome ofcervical spine injuries. J Trauma 28:407–410, 1988.

28. Fielding JW, Francis WR Jr, Hawkins RJ, Pepin J, Hensinger R:Traumatic spondylolisthesis of the axis. Clin Orthop 239:47–52,1989.



29. Francis WR, Fielding JW, Hawkins RJ, Pepin J, Hensinger R:Traumatic spondylolisthesis of the axis. J Bone Joint Surg Br63B:313–318, 1981.

30. Fujii E, Kobayashi K, Hirabayashi K: Treatment in fractures ofthe odontoid process. Spine 13:604–609, 1988.

31. Fujimura Y, Nishi Y, Kobayashi K: Classification and treatmentof axis body fractures. J Orthop Trauma 10:536–540, 1996.

32. Garber J: Abnormalities of the atlas ans axis vertebrae:Congential and traumatic. J Bone Joint Surgery Am 46A:1782–1791, 1964.

33. Govender S, Charles RW: Traumatic spondylolisthesis of theaxis. Injury 18:333–335, 1987.

34. Govender S, Grootboom M: Fractures of the dens: The results ofnon-rigid immobilization. Injury 19:165–167, 1988.

35. Grady MS, Howard MA, Jane JA, Persing JA: Use of the Phila-delphia collar as an alternative to the halo vest in patients withC-2, C-3 fractures. Neurosurgery 18:151–156, 1986.

36. Greene KA, Dickman CA, Marciano FF, Drabier JB, Hadley MN,Sonntag VKH: Acute axis fractures: Analysis of managementand outcome in 340 consecutive cases. Spine 22:1843–1852, 1997.

37. Guiot B, Fessler RG: Complex atlantoaxial fractures. J Neurosurg91:139–143, 1999.

38. Hadley MN, Browner CM, Liu SS, Sonntag VKH: New subtypeof acute odontoid fractures (type IIA). Neurosurgery 22:67–71,1988.

39. Hadley MN, Browner C, Sonntag VKH: Axis fractures: A com-prehensive review of management and treatment in 107 cases.Neurosurgery 17:281–290, 1985.

40. Hadley MN, Dickman CA, Browner CM, Sonntag VKH: Acuteaxis fractures: A review of 229 cases. J Neurosurg 71:642–647,1989.

41. Hanigan WC, Powell FC, Elwood PW, Henderson JP: Odontoidfractures in elderly patients. J Neurosurg 78:32–35, 1993.

42. Jakim I, Sweet MB: Transverse fracture through the body of theaxis. J Bone Joint Surg Br 70B:728–729, 1988.

43. Jenkins JD, Coric D, Branch CL Jr: A clinical comparison of one-and two-screw odontoid fixation. J Neurosurg 89:366–370, 1998.

44. Julien TD, Frankel B, Traynelis VC, Ryken TC: Evidence-basedanalysis of odontoid fracture management. Neurosurg Focus8:Article 1, 2000.

45. Korres DS, Zoubos AB, Kavadias K, Babis GC, Balalis K: The“tear drop” (or avulsed) fracture of the anterior inferior angle ofthe axis. Eur Spine J 3:151–154, 1994.

46. Lennarson PJ, Mostafavi H, Traynelis VC, Walters BC: Manage-ment of type II dens fractures: A case-control study. Spine25:1234–1237, 2000.

47. Levine AM, Edwards CC: The management of traumatic spon-dylolisthesis of the axis. J Bone Joint Surg Am 67A:217–226,1985.

48. Lind B, Nordwall A, Sihlbom H: Odontoid fractures treated withhalo-vest. Spine 12:173–177, 1987.

49. Maiman DJ, Larson SJ: Management of odontoid fractures.Neurosurgery 11:820, 1982 (letter).

50. Marar BC, Tay CK: Fracture of the odontoid process. Aust N ZJ Surg 46:231–236, 1976.

51. McLaurin RL, Vernal R, Salmon JH: Treatment of fractures of theatlas and axis by wiring without fusion. J Neurosurg 36:773–780,1972.

52. Mestdagh H, Letendart J, Sensey JJ, Duquennoy A: Treatment offractures of the posterior axial arch: Results of 41 cases [inFrench]. Rev Chir Orthop Reparatrice Appar Mot 70:21–28,1984.

53. Mirvis SE, Young JW, Lim C, Greenberg J: Hangman’s fracture:Radiologic assessment in 27 cases. Radiology 163:713–717, 1987.

54. Mollan RA, Watt PC: Hangman’s fracture. Injury 14:265–267,1982.

55. Morandi X, Hanna A, Hamlat A, Brassier G: Anterior screwfixation of odontoid fractures. Surg Neurol 51:236–240, 1999.

56. Muller EJ, Wick M, Russe O, Muhr G: Management of odontoidfractures in the elderly. Eur Spine J 8:360–365, 1999.

57. Pepin JW, Hawkins RJ: Traumatic spondylolisthesis of the axis:Hangman’s fracture. Clin Orthop 133–138, 1981.

58. Pepin JW, Bourne RB, Hawkins RJ: Odontoid fractures, withspecial reference to the elderly patient. Clin Orthop 193:178–183,1985.

59. Pitzen T, Caspar W, Steudel WI, Barbier D: Dens fracture inelderly patients and surgical management [in German].Aktuelle Traumatol 24:56–59, 1994.

60. Polin RS, Szabo T, Bogaev CV, Replogle RE, Jane JA: Nonopera-tive management of types II and III odontoid fractures: ThePhiladelphia collar versus the halo vest. Neurosurgery 38:450–456, 1996.

61. Ryan MD, Henderson JJ: The epidemiology of fractures andfracture-dislocations of the cervical spine. Injury 23:38–40, 1992.

62. Ryan MD, Taylor TK: Odontoid fractures: A rational approach totreatment. J Bone Joint Surg Br 64B:416–421, 1982.

63. Ryan MD, Taylor TK: Odontoid fractures in the elderly. J SpinalDisord 6:397–401, 1993.

64. Salmon JH: Fractures of the second cervical vertebra: Internalfixation by interlaminar wiring. Neurosurgery 1:125–127, 1977.

65. Schneider R: “Hangman’s fracture” of the cervical spine.J Neurosurg 22:141–154, 1965.

66. Seljeskog EL, Chou SN: Spectrum of the hangman’s fracture.J Neurosurg 45:3–8, 1976.

67. Seybold EA, Bayley JC: Functional outcome of surgically andconservatively managed dens fractures. Spine 23:1837–1846,1998.

68. Starr JK, Eismont FJ: Atypical hangman’s fractures. Spine 18:1954–1957, 1993.

69. Subach BR, Morone MA, Haid RW Jr, McLaughlin MR, RodtsGR, Comey CH: Management of acute odontoid fractures withsingle-screw anterior fixation. Neurosurgery 45:812–820, 1999.

69a. Sypert GW: Hangman’s fracture: Arguments in favor of surgicaltherapy for type II and III according to Edwards and Levine.Surg Neurol 49:262, 1998 (comment).

70. Tan ES, Balachandran N: Hangman’s fracture in Singapore(1975–1988). Paraplegia 30:160–164, 1992.

71. Torreman M: Long-term prognosis of the hangman’s fracture [inDutch]. Ned Tijdschr Geneeskd 134:1173–1176, 1990.

72. Traynelis VC: Evidence-based management of type II odontoidfractures. Clin Neurosurg 44:41–49, 1997.

73. Verheggen R, Jansen J: Hangman’s fracture: Arguments in favorof surgical therapy for type II and III according to Edwards andLevine. Surg Neurol 49:253–262, 1998.

74. Waddell JP, Reardon GP: Atlantoaxial arthrodesis to treat odon-toid fractures. Can J Surg 26:255–258, 1983.

75. Wang GJ, Mabie KN, Whitehill R, Stamp WG: The nonsurgicalmanagement of odontoid fractures in adults. Spine 9:229–230,1984.

76. White AA, Panjabi MM: Clinical Biomechanics of the Spine. Phila-delphia, J.B. Lippincott, 1990, ed 2.

77. Williams TG: Hangman’s fracture. J Bone Joint Surg Br 57B:82–88, 1975.

78. Wood-Jones F: The ideal lesion produced by judicial hanging.Lancet 1:53, 1913.



CHAPTER 18

Management of Combination Fractures of the Atlas and Axisin Adults

RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS: Treatment of atlas-axis combination fractures based primarily on the specific characteristics of the

axis fracture is recommended. External immobilization of most C1–C2 combination fractures is recom-mended. C1–Type II odontoid combination fractures with an atlantodens interval of 5 mm or more andC1–hangman’s combination fractures with C2–C3 angulation of 11 degrees or more should be consideredfor surgical stabilization and fusion. In some cases, the surgical technique must be modified as a result ofloss of the integrity of the ring of the atlas.

RATIONALE

Combined fractures of the atlas and axis often presentmanagement challenges owing to the unique anatomyand biomechanics of the atlantoaxial complex and the

untoward stresses applied to the atlantoaxial region duringtrauma. Although most isolated atlas and axis fractures havebeen managed with cervical immobilization, the occurrence ofthe two fractures in combination often implies a more signif-icant structural and mechanical injury. Although reports ofcombination C1–C2 fractures are relatively infrequent, suffi-cient evidence exists to allow a review of the management ofa variety of combinations of atlas and axis fractures. Thepurpose of this chapter is to examine the available literature todetermine successful treatment strategies for individualC1–C2 combination fracture types.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject heading “vertebral fracture”in combination with “atlas,” “axis,” and “human” yielded1071 references. The abstracts were reviewed, and articlesfocusing on clinical management and follow-up of combina-tion fractures of the atlas and axis were selected for inclusion.The relative infrequency of these fractures, the small numberof case series, and the numerous case reports with pertinentinformation necessitated rather broad inclusion criteria. Sev-eral papers addressing relevant biomechanics and radiologywere included. The bibliographies of the selected papers werereviewed to provide additional references. These efforts re-sulted in 49 articles describing the clinical features and man-agement of acute traumatic atlas and axis combination frac-tures. Forty-eight of the articles are summarized in Tables 18.1and 18.2. No Class I or II evidence has been generated on the

management of these fractures. Treatment options have beenformulated on the basis of Class III medical evidence.


Overview

In 1920, Sir Geoffrey Jefferson (30) reviewed 46 cases ofatlas fractures. Although his article is best known for thecharacterization of the C1 burst fracture or “Jefferson frac-ture,” Jefferson’s series included 19 fractures that were de-scribed as “combination fractures” of the atlas and the axis.He noted increased morbidity and mortality for patients withcombination injuries. Eleven of the 19 patients he describedwith C1–C2 combination injuries had significant neurologicalinjuries. In 1986, Levine and Edwards (34) reported theirapproach to the management of C1–C2 traumatic fractureinjuries. They suggested that if an atlas or axis injury wasidentified, a careful search for other related injuries was indi-cated. They stressed that each patient and each injury neededto be evaluated independently. They described staged treat-ment for certain injuries to allow healing of one fracture(usually the atlas) before definitively managing the combina-tion injury (typically the axis fracture). Several of their obser-vations are worthy of consideration in the management ofcombination fracture injuries of the atlas and axis today.

Incidence

Combination fractures of the C1–C2 complex are relativelycommon. In reports focusing primarily on odontoid fractures,the occurrence of a concurrent C1 fracture in the presence of aType II or Type III odontoid fracture has been reported in 5 to53% of cases (4, 12, 23, 25, 26, 28, 36, 39, 40, 43–45, 47, 49).Odontoid fractures have been identified in 24 to 53% of patientswith atlas fractures (18, 32, 35, 45). In the presence of a hang-


man’s fracture, the reported incidence of a C1 fracture rangesfrom 6 to 26% (13, 17, 33, 38, 41, 45). Greene et al. (23) reportedon 340 axis fractures and found 48 concurrent atlas fractures(combination injuries), for an incidence of 14%. Ryan and Hen-derson (45) reviewed 717 spine fractures and found combinationatlas-axis fractures in 15% of odontoid fractures and in 9% ofhangman’s fractures. Gleizes et al. (21) reviewed 784 patientswith proximal cervical spine injuries in 2000. One hundred six-teen patients had injuries to C1 and/or C2. Thirty-one patientshad C1 fractures in association with a C2 fracture (combinationinjury), representing 4% of the total cervical spine fracture pop-ulation and 27% of all C1–C2 fracture injuries.

Morbidity and mortality

Various authors have suggested that the morbidity and mor-tality of C1–C2 combination fractures is higher than that associ-ated with isolated fractures of either the atlas or the axis (12, 18,19, 25, 26, 31, 49). Fujimura et al. (19) observed neurologicaldeficits in 82 (34%) of 247 patients with injuries to the C1–C2complex. Those patients with deficit had either burst fractures orfractures of the posterior arch of C1 or a fracture of the C2 bodycoupled with an odontoid or hangman’s fracture. Several au-thors have described a high mortality rate with combinationfractures, in particular C1 fractures combined with Type II odon-toid fractures (18, 25, 26, 49). Fowler et al. (18) found that 6 (86%)of 7 patients with C1–Type II odontoid combination fracturesdied in the early treatment period. Similarly, Hanssen and Ca-banela (26) observed that 5 (83%) of 6 patients with this samecombination fracture pattern died within the first 40 days ofinjury. Both Hanigan et al. (25) and Zavanone et al.(49) reportedearly deaths associated with C1–Type II odontoid fractures. Inother reports on C1–C2 combination fractures, the description ofmorbidity and mortality has been less remarkable (12, 18).Dickman et al. (12) suggested a 12% incidence of neurologicaldeficit for C1–C2 combination fractures compared with a 0% (0of 32) and a 2% (2 of 125) incidence for isolated atlas and axisfractures, respectively. Kesterson et al. (31) described four pa-tients with C1–C2 combination fractures. Only 1 patient (25%)had a neurological deficit. Irrespective of the author, the de-scribed incidence of morbidity and mortality associated withcombination C1–C2 fractures seems to be more than that asso-ciated with isolated atlas and axis fractures.

Treatment

Since Jefferson published the original description of C1–C2combination fracture injuries, nearly every series reviewingeither isolated fractures of the atlas or the axis includes somemention of C1–C2 combination fractures. It is difficult todetermine the specific treatment provided to and outcome formost of those patients. Several authors have focused theirreports specifically on combination C1–C2 fractures and theirmanagement (12, 21, 24).

In 1989, Dickman et al. (12) identified 25 cases of acuteatlas-axis combination fractures in an overall series of 860patients with acute cervical fracture injuries. In their experi-ence, C1–C2 combination fractures represented 3% of theirtotal cervical fracture population. Combination injuries rep-resented 43% of acute atlas fractures (25 of 58 patients) and16% of acute axis fractures (25 of 150 patients). The fracturesof C1 and C2 were identified using plain film x-rays in 76 and92% of the cases, respectively. Computed tomography char-acterized the combination fracture patterns in all cases.Twelve percent of patients (3 of 25 patients) had neurologicaldeficits at admission. Two patients had acute central cordsyndrome, and one patient had a complete neurological in-jury. The etiology of the injury was a motor vehicle accident in60% of cases and a fall in 28%. Four main types of atlas-axisfracture combination were identified: C1–Type II odontoid (10cases, 40%), C1–miscellaneous axis fracture (7 cases, 28%),C1–Type III odontoid (5 cases, 20%), and C1–hangman’s-typefracture (3 cases, 12%). The distribution of the atlas fractureswas reported as multiple ring fractures in 40%, posterior ringfracture in 28%, unilateral ring fracture in 24%, and lateralmass fracture in 8%. Nonoperative therapy was the initialmanagement strategy in 20 (84%) of 25 of patients. Eighteenpatients were placed in a halo orthosis and two in asuboccipital-mandibular immobilizer (SOMI) brace, for a me-dian duration of 12 weeks (range, 10–22 wk). Four patientswere treated with early surgical stabilization and fusion basedon an atlantoaxial interval of 6 mm or more. Three weretreated with posterior C1–C2 wiring and fusion. Follow-upwas accomplished in 23 (92%) of 25 patients. Nineteen (95%)of the 20 patients treated with either a halo or SOMI orthosisachieved stability and fusion. Halo immobilization failed inone patient with an initial atlantoaxial interval of 5 mm, andthe patient was treated with posterior C1–C2 fusion. All pa-tients treated surgically achieved stability using a posteriorfusion technique, four early and one delayed (100%). No

TABLE 18.1. Initial Management of CombinationAxis–Atlas Fracturesa

Combination Fracture Type Treatment Options

C1–Type II odontoid fractureStable Collar, halo, surgical fixation/

fusionUnstable (ADI �5 mm) Halo, surgical fixation/fusion

C1–Type III odontoid fracture Halo

C1–miscellaneous axis Collar, halo

C1–hangman’s fractureStable Collar, haloUnstable (C2–C3angulation �11 degrees)

Halo, surgical fixation/fusion

a ADI, atlantodens interval.

Combination Fractures of the Atlas and Axis


TAB

LE18

.2.

Sum

mar

yof

Rep

orts

onFr

actu

res

ofth

eA

tlas

and

Axi

sa

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ce

Cla

ssC

oncl

usio

ns

And

erss

onet

al.,

2000

(2)

Cas

ese

ries

ofpa

tient

s�

65yr

with

odon

toid

frac

ture

s.III

Incl

udes

3pa

tient

sw

ithC

1–Ty

peII

odon

toid

frac

ture

s.Tr

eatm

ent:

Hal

o,2

patie

nts.

Post

erio

rce

rvic

al

fusi

on,

1pa

tient

.

Gle

izes

etal

.,20

00(2

1)R

etro

spec

tive

epid

emio

logi

cal

revi

ewof

coin

cide

nce

offr

actu

res

inth

eup

per

cerv

ical

spin

e.

III78

4ce

rvic

alsp

ine

inju

ries

.11

6up

per

cerv

ical

spin

ein

juri

es(C

1–C

2)(1

5%).

31/1

16(2

6%)

com

bina

tion

of

C1

and

C2.

70%

ofal

lat

las

frac

ture

soc

curr

edin

com

bina

tion

with

anot

her

frac

ture

.30

%of

all

hang

man

’s

and

odon

toid

frac

ture

soc

curr

edin

com

bina

tion

with

anot

her

frac

ture

.41

.9%

ofpa

tient

sw

ithco

mbi

natio

n

frac

ture

sof

the

uppe

rce

rvic

alsp

ine

unde

rwen

tsu

rgic

alfix

atio

nve

rsus

21.7

%of

thos

ew

ithis

olat

ed

inju

ries

.

Mul

ler

etal

.,20

00(4

1)C

ase

seri

esof

39ca

ses

ofha

ngm

an’s

frac

ture

s.III

Incl

udes

2pa

tient

sw

ithC

1ri

ngfr

actu

res

(5.1

%).

Gui

otan

dFe

ssle

r,19

99(2

4)R

etro

spec

tive

revi

ewof

10pa

tient

sun

derg

oing

surg

ical

fixat

ion

for

com

bina

tion

C1–

C2

frac

ture

s.5/

10re

ferr

edsp

ecifi

cally

for

surg

ical

fixat

ion

afte

rfa

iled

exte

rnal

imm

obili

zatio

n.A

vera

ge

follo

w-u

p,28

.5m

o.

IIITy

pe:

C1–

Type

IIod

onto

id,

9pa

tient

s(9

0%).

C1–

Type

IIIod

onto

idan

dha

ngm

an’s

,1

patie

nt(1

0%).

Tech

niqu

e:O

dont

oid

scre

w,

6(6

0%).

Odo

ntoi

dsc

rew

plus

C2

pedi

cle

scre

ws,

1pa

tient

(10%

).C

1–C

2

tran

sart

icul

arsc

rew

s(p

oste

rior

),2

patie

nts

(20%

).C

1–C

2tr

ansa

rtic

ular

scre

ws

(ant

erio

r),

1pa

tient

(10%

).

Out

com

e:1

unre

late

dde

ath.

All

othe

rsfu

sed

succ

essf

ully

with

out

othe

rco

mpl

icat

ion.

Hen

ryet

al.,

1999

(28)

Cas

ese

ries

of61

case

sof

Type

IIod

onto

idfr

actu

res

trea

ted

with

ante

rior

scre

wfix

atio

nin

whi

chfo

llow

-up

was

avai

labl

e.

IIIIn

clud

es10

com

bina

tion

frac

ture

sof

C1–

C2

(16%

).C

1bu

rst

(Jeffe

rson

)-Ty

peII

odon

toid

,3

patie

nts

(5%

).

C1

ante

rior

arch

-Typ

eII

odon

toid

,3

patie

nts

(5%

).C

1po

ster

ior

arch

-Typ

eII

odon

toid

,4

patie

nts

(6%

).A

ll

patie

nts

inth

ese

ries

wer

etr

eate

dw

ithan

teri

orod

onto

idsc

rew

fixat

ion

with

a92

%su

cces

sra

te.

No

prob

lem

sat

trib

uted

dire

ctly

toth

epr

esen

ceof

the

atla

sfr

actu

re.

Mor

andi

etal

.,19

99(4

0)C

ase

seri

esin

clud

ing

17od

onto

idfr

actu

res

trea

ted

with

ante

rior

scre

wfix

atio

n.

IIIIn

clud

es2

case

sof

C1–

post

erio

rar

chfr

actu

repl

usa

post

erio

rly

disp

lace

dTy

peII

odon

toid

.

Lee

etal

.,19

98(3

2)R

etro

spec

tive

revi

ewof

16ca

ses

ofat

las

frac

ture

.III

Incl

udes

8pa

tient

sw

ithco

mbi

natio

nC

1an

dC

2fr

actu

res.

C1–

Type

IIod

onto

id,

3ca

ses.

1tr

eate

dw

ith

halo

imm

obili

zatio

n.2

trea

ted

with

post

erio

rC

1–C

2fu

sion

.C

1–H

angm

an’s

,2

case

s.B

oth

trea

ted

with

cerv

ical

colla

r.C

1–C

2bo

dyfr

actu

re,

3ca

ses.

All

3tr

eate

dw

ithce

rvic

alco

llar.

Aut

hors

conc

lude

that

the

man

agem

ent

ofth

eco

mbi

natio

nfr

actu

resh

ould

beba

sed

onth

eC

2fr

actu

rean

dth

atha

loim

mob

iliza

tion

isno

tal

way

sre

quir

ed.

Seyb

old

and

Bay

ley,

1998

(47)

Cas

ese

ries

of57

odon

toid

frac

ture

s.III

Incl

udes

3ca

ses

ofC

1ri

ngfr

actu

repl

usTy

peII

odon

toid

.Th

eau

thor

ssu

cces

sful

lym

anag

edtw

opa

tient

s

with

aha

lo.

One

patie

ntw

astr

eate

din

aco

llar

with

a“p

oor”

resu

lt.Th

eov

eral

lfu

sion

rate

for

the

Type

II

odon

toid

frac

ture

sin

this

seri

esw

as65

%.

No

spec

ific

effe

ctfr

omth

eC

1fr

actu

rew

asno

ted.

Apo

stol

ides

etal

.,19

97(3

)C

ase

repo

rt.

IIIPa

tient

with

ante

rior

ring

ofC

1fr

actu

rean

da

Type

IIod

onto

idin

who

mha

loim

mob

iliza

tion

faile

d.

Trea

ted

succ

essf

ully

with

ante

rior

C1–

C2

tran

sart

icul

arfix

atio

nan

dan

odon

toid

scre

w.

Ber

lem

ann

and

Schw

arze

nbac

h,19

97(5

)R

etro

spec

tive

revi

ewof

19pa

tient

sag

e�

65w

ithod

onto

id

frac

ture

s.

IIIIn

clud

es4

patie

nts

with

C1

frac

ture

san

dTy

peII

odon

toid

frac

ture

sal

ltr

eate

dw

ithan

teri

orod

onto

idsc

rew

fixat

ion.

Gre

ene

etal

.,19

97(2

3)La

rge

revi

ewof

340

axis

frac

ture

s.III

48pa

tient

sw

ithan

axis

frac

ture

also

had

anat

las

frac

ture

(14%

).Sp

ecifi

cson

man

agem

ent

are

not

pres

ente

d,bu

tth

eau

thor

sin

dica

teth

atth

em

anag

emen

tin

thes

eca

ses

was

base

don

the

C2

frac

ture

.

Cas

tillo

and

Muk

herj

i,19

96(8

)C

ase

repo

rt.

IIIIn

clud

es1

case

ofJe

ffers

onfr

actu

repl

usTy

peII

odon

toid

trea

ted

with

halo

.

Cor

icet

al.,

1996

(9)

Cas

ese

ries

of57

patie

nts

with

hang

man

’sfr

actu

res.

IIIIn

clud

es7

case

sof

com

bina

tion

frac

ture

(C1–

hang

man

’s).

All

wer

etr

eate

dba

sed

onde

gree

of

disp

lace

men

t.If

disp

lace

men

tw

as�

6m

m,

they

wer

etr

eate

dw

ithno

nrig

idim

mob

iliza

tion.

Fujim

ura

etal

.,19

95(1

9)C

ase

seri

esof

axis

body

frac

ture

s.III

Des

crib

es3

patie

nts

with

C1–

mis

cella

neou

sbo

dyfr

actu

real

ltr

eate

dw

ithce

rvic

alim

mob

iliza

tion.

Aut

hors

reco

mm

end

nono

pera

tive

trea

tmen

tex

cept

inca

ses

ofse

vere

angu

latio

n.Ph

ilade

lphi

aco

llar

used

if

min

imal

angu

latio

n.

Polin

etal

.,19

96(4

4)C

ase

seri

esof

62pa

tient

sw

ithod

onto

idfr

actu

res.

IIIIn

clud

es5

case

sof

com

bina

tion

C1–

C2

frac

ture

(8%

).C

1–Je

ffers

on-T

ype

IIod

onto

id,

4ca

ses.

C1–

mis

cella

neou

sC

2bo

dyfr

actu

re,

1ca

se.

All

patie

nts

inse

ries

man

aged

with

eith

erha

loor

colla

r.

Coy

neet

al.,

1995

(10)

Ret

rosp

ectiv

ere

view

of32

patie

nts

with

odon

toid

frac

ture

s

incl

udes

1co

mbi

natio

nfr

actu

re.

III1

case

ofJe

ffers

on-T

ype

IIod

onto

idtr

eate

dw

ithG

allie

fusi

on.

Fujim

ura

etal

.,19

95(1

9)R

etro

spec

tive

revi

ewof

247

adm

issi

ons

with

uppe

rce

rvic

alsp

ine

frac

ture

s.Fo

cuse

son

82pa

tient

sw

ithne

urol

ogic

alde

ficit.

IIIIn

patie

nts

with

com

bine

din

jury

ofC

1–C

2,tw

one

urol

ogic

alde

ficits

occu

rred

inpa

tient

sw

ithpo

ster

ior

arch

frac

ture

,bu

rst

frac

ture

ofth

eat

las,

orbo

dyfr

actu

reof

the

axis

asso

ciat

edw

ithei

ther

anod

onto

id

frac

ture

orha

ngm

an’s

frac

ture

.



TAB

LE18

.2.

Con

tinu

ed

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ce

Cla

ssC

oncl

usio

ns

Ben

zel

etal

.,19

94(4

)C

ase

seri

es.

IIIIn

clud

es1

case

ofC

1,ve

rtic

ally

orie

nted

C2

mis

cella

neou

sbo

dyfr

actu

re(tr

eatm

ent

not

desc

ribe

d).

The

auth

ordi

scus

ses

the

poss

ible

mec

hani

sms,

incl

udin

ghy

pere

xten

sion

and

axia

llo

adin

g.

Pede

rsen

and

Kos

tuik

,19

94(4

2)C

ase

repo

rtof

70-y

r-ol

dm

anw

ithfr

actu

redi

sloc

atio

nof

C1–

C2

with

20-m

mat

lant

oaxi

aldi

spla

cem

ent.

IIISu

cces

sful

lytr

eate

dw

ithO

–C4

deco

mpr

essi

onan

dpo

ster

ior

fusi

onw

ithco

mpl

ete

reco

very

.

Han

igan

etal

.,19

93(2

5)C

ase

seri

esof

19pa

tient

s�

80yr

ofag

ew

ithod

onto

idfr

actu

res.

IIIIn

clud

es2

patie

nts

with

aC

1–Je

ffers

on-T

ype

IIod

onto

id.

1pa

tient

died

inth

eho

spita

laf

ter

bein

gpl

aced

in

trac

tion.

1pa

tient

had

ast

able

fibro

usno

nuni

onaf

ter

trea

tmen

tin

aha

lo.

Boh

ayet

al.,

1992

(6)

Cas

ese

ries

.III

Incl

udes

aca

seof

aC

1bu

rst

frac

ture

plus

ave

rtic

alC

2bo

dyfr

actu

resu

cces

sful

lytr

eate

din

ace

rvic

al

colla

ral

one.

Hay

san

dB

ernh

ang,

1992

(27)

Cas

ese

ries

ofun

usua

lfr

actu

res

ofth

eat

las.

IIIIn

clud

es2

case

sof

com

bina

tion

frac

ture

s.C

1(a

nter

ior

arch

)–Ty

peII

odon

toid

frac

ture

faile

dha

lo

trea

tmen

t,re

sulti

ngin

anO

–C2

fusi

on.

Jean

nere

tan

dM

ager

l,19

92(2

9)C

ase

seri

esof

59pa

tient

sw

ithod

onto

idfr

actu

res,

30of

whi

ch

wer

etr

eate

dsu

rgic

ally

.

IIIIn

clud

es2

case

sin

whi

chth

epo

ster

ior

arch

ofC

1w

asno

tin

tact

.C

1–Je

ffers

on-T

ype

IIod

onto

id,

1pa

tient

.

C1–

post

erio

rar

ch-T

ype

IIIod

onto

id,

1pa

tient

.A

utho

rsfe

elst

rong

lyth

at,

ifth

epo

ster

ior

arch

ofC

1is

not

inta

ct,

C1–

C2

tran

sart

icul

arfix

atio

nis

indi

cate

d.In

the

com

men

tth

atfo

llow

sth

ear

ticle

,th

epo

int

ism

ade

that

anon

lay

graf

tbe

twee

nC

1an

dC

2po

ster

iorl

yw

ithou

tw

irin

gof

C1

follo

wed

byha

loim

mob

iliza

tion

has

been

used

inth

issi

tuat

ion.

Rya

nan

dH

ende

rson

,19

92(4

5)Ep

idem

iolo

gica

lre

port

of71

7sp

ine

frac

ture

s.III

Atla

sfr

actu

res

occu

rred

with

odon

toid

frac

ture

s(5

3%)

and

with

hang

man

’sfr

actu

res

(24%

).O

dont

oid

frac

ture

soc

curr

edw

ithat

las

frac

ture

s(1

5%).

Han

gman

’sfr

actu

reoc

curr

edw

ithat

las

frac

ture

(9%

).

Cra

igan

dH

odgs

on,

1991

(11)

Cas

ere

port

.III

Jeffe

rson

plus

supe

rior

face

tof

axis

trea

ted

with

colla

r.

Esse

san

dB

edna

r,19

91(1

6)C

ase

repo

rt.

Atla

san

dod

onto

idfr

actu

re.

IIIJe

ffers

onpl

usTy

peII

odon

toid

ina

34-y

r-ol

dm

antr

eate

dsu

cces

sful

lyw

ithco

llar

only

(see

naf

ter

a1-

mo

dela

yin

diag

nosi

s).

Kes

ters

onet

al.,

1991

(31)

Cas

ese

ries

,re

tros

pect

ive

revi

ew.

IIIIn

clud

es4

patie

nts

with

com

bina

tion

frac

ture

ofth

eat

las

and

Type

IIod

onto

idtr

eate

dw

ithO

–C2

fusi

on.

1

ofth

ese

4pa

tient

sha

da

sign

ifica

ntne

urol

ogic

alde

ficit

(25%

).Th

eau

thor

ssu

gges

tsu

rger

yif

unst

able

and

defin

ein

stab

ility

asat

lant

oaxi

alin

terv

alof

�5

mm

orla

tera

lm

ass

disp

lace

men

t�

7m

m.

Levi

nean

dEd

war

ds,

1991

(35)

Cas

ese

ries

of34

patie

nts

with

atla

sfr

actu

res.

IIIIn

clud

es15

patie

nts

with

aco

mbi

natio

nC

1–C

2fr

actu

re(4

4%).

C1–

Type

IIor

Type

IIIod

onto

id,

8pa

tient

s

(24%

).C

1–ha

ngm

an’s

,7

patie

nts

(21%

).D

escr

ibes

2ca

ses

inth

eC

1–od

onto

idfr

actu

regr

oup

inw

hich

the

post

erio

rC

1ar

chal

tere

dth

etr

eatm

ent

plan

.In

1ca

se,

aG

allie

fusi

onfa

iled,

and

inth

ese

cond

,no

wir

ing

was

used

,ju

ston

lay

bone

graf

t.

Mon

tesa

noet

al.,

1991

(39)

Cas

ese

ries

of14

Type

IIod

onto

idfr

actu

res

trea

ted

with

ante

rior

odon

toid

scre

wfix

atio

n.Fo

llow

-up,

24m

o.

III7

patie

nts

had

aC

1fr

actu

re(5

0%).

The

over

all

fusi

onra

tew

as93

%.

No

prob

lem

sat

trib

uted

dire

ctly

toth

e

C1

frac

ture

.

Zav

onon

eet

al.,

1991

(49)

Cas

ese

ries

of23

C1–

C2

frac

ture

s.III

Incl

udes

2co

mbi

natio

nfr

actu

res

(9%

).C

1–Ty

peII

odon

toid

:th

epa

tient

died

.C

1–ha

ngm

an’s

:tr

eate

d

succ

essf

ully

with

trac

tion

redu

ctio

nan

dM

iner

va.

Fow

ler

etal

.,19

90(1

8)C

ase

seri

esof

48at

las

frac

ture

sfr

omse

ries

of86

7C

-spi

ne

frac

ture

s(5

.5%

).

IIIIn

clud

es18

case

sw

itha

com

bina

tion

C1–

C2

frac

ture

(38%

ofto

tal

seri

es).

C1

burs

t(Je

ffers

on)-

Type

II

odon

toid

,6

patie

nts

(33%

).C

1bu

rst

(Jeffe

rson

)-Ty

peIII

odon

toid

,1

patie

nt(6

%).

C1

burs

t(Je

ffers

on)-

mis

cella

neou

sax

is,

2pa

tient

s(1

1%).

C1

burs

t(Je

ffers

on)-

hang

man

’s,

0pa

tient

s(0

%).

C1

arch

-Typ

eII

odon

toid

,8

patie

nts

(44%

).C

1ar

ch-T

ype

IIIod

onto

id,

1pa

tient

(6%

).C

1ar

ch-m

isce

llane

ous

axis

,1

patie

nt(6

%).

C1

arch

-han

gman

’s,

3pa

tient

s(1

6%).

Thes

eau

thor

spr

esen

tda

tasu

ppor

ting

the

incr

ease

d

mor

talit

yas

soci

ated

with

com

bina

tion

C1–

C2

frac

ture

s.6

(86%

)of

the

7ea

rly

deat

hsha

da

C1

frac

ture

asso

ciat

edw

ithei

ther

aTy

peII

orTy

peIII

odon

toid

frac

ture

.

Dic

kman

etal

.,19

89(1

2)R

etro

spec

tive

revi

ewof

25pa

tient

sw

ithfr

actu

res

ofbo

thC

1an

d

C2.

Com

pris

es3%

ofth

eov

eral

lce

rvic

alsp

ine

inju

ryco

hort

(25/

860)

.

IIIFo

urty

pes

note

d:C

1–Ty

peII

odon

toid

,10

patie

nts

(40%

).C

1–m

isce

llane

ous

axis

,7

patie

nts

(28%

).C

1–

Type

IIIod

onto

id,

5pa

tient

s(2

0%).

C1–

hang

man

’s,

3pa

tient

s(1

2%).

Neu

rolo

gica

lde

ficit

in3/

25pa

tient

s

(12%

).Tr

eatm

ent

dete

rmin

edby

type

ofC

2fr

actu

re.

Non

oper

ativ

e,84

%.

Hal

o,18

patie

nts.

SOM

I,2

patie

nts.

In1

ofth

eC

1–Ty

peII

patie

nts,

halo

faile

d,an

dpa

tient

requ

ired

C1–

C2

fusi

on.

Ope

rativ

e(in

itial

man

agem

ent).

C1–

Type

IIod

onto

idw

ith6

mm

disp

lace

men

t,3

patie

nts

trea

ted

with

post

erio

rC

1–C

2

fusi

on;

1pa

tient

trea

ted

with

O–C

2fu

sion

beca

use

ofm

ultip

lefr

actu

res

inC

1.

Combination Fractures of the Atlas and Axis S143


TAB

LE18

.2.

Con

tinu

ed

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ce

Cla

ssC

oncl

usio

ns

Fiel

ding

etal

.,19

89(1

7)C

ase

seri

esof

123

hang

man

’sfr

actu

res.

IIIIn

clud

es:

C1

arch

-han

gman

’s,

10ca

ses

(8%

).C

1bu

rst

(Jeffe

rson

)-ha

ngm

an’s

,2

case

s(2

%).

C1

late

ral

mas

s-

hang

man

’s,

3ca

ses

(3%

).Sp

ecifi

csno

tgi

ven

for

each

subt

ype,

but

over

all

the

auth

ors

reco

mm

end

trea

tmen

tba

sed

onth

eC

2fr

actu

rede

spite

the

pres

ence

ofth

eC

1fr

actu

re.

Reg

ardl

ess

ofth

eC

1fr

actu

re,

the

auth

ors

favo

ran

ante

rior

C2–

C3

fusi

onfo

rth

ose

patie

nts

with

angu

latio

n�

11de

gree

s,as

this

grou

p

had

an85

%no

nuni

onra

tew

ithce

rvic

alim

mob

iliza

tion.

Gov

ende

ran

dC

harl

es,

1987

(22)

Cas

ese

ries

ofup

per

cerv

ical

frac

ture

s.III

Incl

udes

2ca

ses

ofco

mbi

natio

nC

1po

ster

ior

arch

frac

ture

-han

gman

’sfr

actu

re,

trea

ted

succ

essf

ully

with

a

cerv

ical

colla

r(n

onri

gid

cerv

ical

imm

obili

zatio

n).

Han

ssen

and

Cab

enel

a,19

87(2

6)C

ase

seri

esof

42od

onto

idfr

actu

res.

IIIIn

clud

es7

com

bina

tion

frac

ture

s(1

7%).

C1–

Jeffe

rson

-Typ

eII

odon

toid

,6

patie

nts.

5/6

(83%

)di

edw

ithin

first

40d.

1/6

deve

lope

da

stab

leno

nuni

on.

C1–

post

erio

rar

ch-T

ype

IIod

onto

id,

1pa

tient

.H

eale

dw

ith

halo

imm

obili

zatio

n.

Lind

etal

.,19

87(3

6)C

ase

seri

esof

14od

onto

idfr

actu

res

man

aged

inha

loor

thos

es.

IIIIn

clud

es1

case

ofC

1–Je

ffers

on-T

ype

IIod

onto

idm

anag

edin

aha

love

stfo

r12

wk.

Mir

vis

etal

.,19

87(3

8)R

adio

grap

hic

revi

ewof

27C

2fr

actu

res.

IIIN

oted

9as

soci

ated

C1

frac

ture

s(2

6%).

Sega

let

al.,

1987

(46)

Cas

ese

ries

of18

patie

nts

with

atla

sfr

actu

res.

III6

case

sw

ere

com

bina

tion

C1–

C2

frac

ture

s.C

1–Je

ffers

on-o

dont

oid

frac

ture

,5

case

s.3

trea

ted

with

halo

,2

with

trac

tion

follo

wed

byha

lo.

C1–

Jeffe

rson

-han

gman

’s,

1ca

setr

eate

dw

itha

colla

r.

Levi

nean

dEd

war

ds,

1986

(34)

Rev

iew

artic

leon

man

agem

ent

ofC

1–C

2tr

aum

a.III

Com

men

tson

com

bine

din

juri

es:

1.Th

epr

esen

ceof

thre

ein

juri

esto

the

C1–

C2

com

plex

isas

soci

ated

with

ahi

ghlik

elih

ood

ofne

urol

ogic

alin

jury

.2.

If1

inju

ryor

frac

ture

isfo

und,

one

shou

ldlo

okca

refu

lly

for

anot

her.

3.M

echa

nism

ofin

jury

usua

llyis

cons

iste

ntw

ithth

ein

jury

obse

rved

.4.

Each

inju

ryne

eds

to

beev

alua

ted

indi

vidu

ally

;fo

rex

ampl

e,th

epr

esen

ceof

2fr

actu

res

does

not

alw

ays

indi

cate

inst

abili

ty

(pos

teri

orar

chof

C1

plus

ano

ndis

plac

edha

ngm

an’s

frac

ture

).5.

Stag

ing

oftr

eatm

ent

may

bere

quir

ed(a

s

desc

ribe

dby

Lips

onet

al.

belo

w)

with

allo

wan

ceof

one

frac

ture

tohe

albe

fore

trea

ting

defin

itive

ly.

Levi

nean

dEd

war

ds,

1985

(33)

Cas

ese

ries

of53

patie

nts

with

hang

man

’sfr

actu

re.

Des

crib

es

stab

le(T

ype

Iha

ngm

an’s

)an

dun

stab

le(T

ype

IIha

ngm

an’s

)

grou

ps.

IIIIn

clud

es9

case

sof

Type

Iha

ngm

an’s

(sta

ble)

plus

C2

frac

ture

:Ty

peII

odon

toid

,2

case

s.Ty

peIII

odon

toid

,

3ca

ses.

Post

erio

rar

ch,

1ca

se.

Bur

st(Je

ffers

on),

2ca

ses.

Late

ral

mas

s,1

case

.O

nly

1ca

seTy

peII

hang

man

’s(u

nsta

ble)

with

C2

frac

ture

.Po

ster

ior

arch

,1

case

.O

nly

1ca

setr

eate

dsu

rgic

ally

:Ty

peI

hang

man

’spl

usTy

peII

odon

toid

trea

ted

with

post

erio

rC

1–C

2fu

sion

.

Pepi

nan

dH

awki

ns,

1981

(43)

Cas

ese

ries

of41

odon

toid

frac

ture

s.III

Incl

udes

9ca

ses

ofod

onto

idfr

actu

rein

com

bina

tion

with

anot

her

spin

alfr

actu

re,

ofw

hich

the

C1–

Jeffe

rson

-Typ

eII

odon

toid

was

the

mos

tco

mm

on.

All

trea

ted

with

eith

erC

1–C

2fu

sion

orha

lo.

Aut

hor

reco

mm

ends

fusi

onin

the

elde

rly.

Effe

ndi

etal

.,19

81(1

3)C

ase

seri

esof

131

hang

man

’sfr

actu

rew

ithcl

assi

ficat

ion.

IIIIn

clud

esco

mbi

natio

nfr

actu

res:

C1

post

erio

rar

ch-h

angm

an’s

,8

patie

nts

(8/1

31,

6%).

Odo

ntoi

dfr

actu

re-

hang

man

’s,

2pa

tient

s(2

/131

,2%

).Sp

ecifi

cou

tcom

esno

tpr

esen

ted,

but

all

fuse

dw

ithei

ther

ante

rior

or

post

erio

rC

1–C

2fu

sion

orha

lo.

Ove

rall

mor

talit

yw

as9%

.

Ekon

get

al.,

1981

(14)

Cas

ese

ries

of22

patie

nts

with

odon

toid

frac

ture

s.III

Incl

udes

:C

1–Je

ffers

on-T

ype

IIod

onto

id,

1pa

tient

.C

1–Je

ffers

on-T

ype

IIIod

onto

id,

2pa

tient

s.A

lltr

eate

d

with

halo

.H

alo

faile

din

1of

the

C1

Jeffe

rson

-Typ

eIII

odon

toid

patie

nts,

requ

irin

gC

1–C

2po

ster

ior

fusi

on.

Lips

on,

1977

(37)

Cas

ese

ries

of3

case

sof

atla

sfr

actu

repl

usTy

peII

odon

toid

.III

The

auth

ors

reco

mm

end

com

bina

tion

ther

apy

ofha

loim

mob

iliza

tion

for

10–1

2w

kun

tilth

epo

ster

ior

arch

ofth

eat

las

frac

ture

has

heal

ed,

follo

wed

byat

lant

oaxi

alfu

sion

(Gal

liety

pe)

tode

finiti

vely

man

age

the

odon

toid

frac

ture

.

Bra

shea

ret

al.,

1975

(7)

Cas

ese

ries

ofha

ngm

an’s

frac

ture

.III

Incl

udes

2pa

tient

sw

ithC

1po

ster

ior

arch

frac

ture

plus

hang

man

’str

eate

dw

ithre

duct

ion

and

Min

erva

for

3–6

mo.

And

erso

nan

dD

’Alo

nzo,

1974

(1)

Cas

ese

ries

ofod

onto

idfr

actu

res.

IIIIn

clud

es1

patie

ntw

ithco

mbi

ned

C1–

Type

IIod

onto

idfr

actu

retr

eate

dw

ithO

–C2

fusi

on.

Ellio

tet

al.,

1972

(15)

Cas

ese

ries

.III

C1

post

erio

rar

ch-h

angm

an’s

,2

case

str

eate

dw

ithim

mob

iliza

tion.

Sher

kan

dN

icho

lson

,19

70(4

8)C

ase

repo

rt.

III1

case

each

ofa

com

bina

tion

C1–

Type

IIod

onto

idan

da

C1–

hang

man

’s.

Bot

hw

ere

trea

ted

with

imm

obili

zatio

n(r

educ

tion

intr

actio

nfo

llow

edby

aM

iner

vabr

ace)

succ

essf

ully

.

aO

,oc

cipi

tal;

C-s

pine

,ce

rvic

alsp

ine;

SOM

I,su

bocc

ipit

al-m

andi

bula

rim

mob

ilize

r.



patient deteriorated during or as a result of treatment. Sixpatients complained of persistent neck pain or limitation ofneck motion. The authors offered a treatment algorithm basedon the type and displacement of the axis fracture. They be-lieve that every patient with a C1 or C2 fracture should bestudied with computed tomography to rule out a combinationinjury. When present, atlas fractures in combination withType II or Type III odontoid fractures with an atlantoaxialinterval of 5 mm or more should be considered for earlysurgical management. The authors stressed that the integrityof the C1 ring must be assessed to determine whether C1–C2wiring techniques can be used. Their perspectives were of-fered before the popularization of C1–C2 transarticular screwfixation techniques.

Guiot and Fessler (24), in 1999, described a series of 10patients with combination atlas-axis fractures treated withsurgical stabilization and fusion. In five (50%) of these pa-tients, halo immobilization had failed, and the patients werereferred specifically for operative intervention. Ninety per-cent were patients with C1–Type II odontoid fractures, andthe remaining patient had a C1–Type III odontoid combina-tion fracture injury. One patient died of unrelated causes inthe follow-up period. There were no other significant compli-cations in a follow-up period of 28.5 months. All nine otherpatients accomplished successful fusion. An odontoid screwalone was used in five patients, an odontoid screw plus C2pedicle screws in one, posterior transarticular screws in two,and anterior transarticular screws in one patient. The authors’indications for surgery included patients with fractures thatcould not be reduced or maintained with external immobili-zation and unstable fractures with a high likelihood of non-union (including evidence of disruption of the transverseatlantal ligament).

Treatment of C1–Type II odontoidcombination fractures

The treatment of specific fracture combinations has beenthe subject of numerous reports. The C1–Type II odontoidfracture combination seems to be the most frequent and thesubject of the most variability in treatment strategy. As notedwith the management of isolated Type II odontoid fractures,optimal treatment remains controversial (see Chapter 17).Management techniques for C1–Type II odontoid combina-tion fractures include semirigid immobilization (collar), trac-tion and then immobilization in a brace, rigid immobilization(halo, Minerva, SOMI), posterior fusion with and withoutinstrumentation, and anterior odontoid screw fixation. Al-though Esses and Bednar (16) describe a single cases of C1–Type II odontoid combination fracture managed successfullyin a cervical collar, the lower fusion rate described for Type IIodontoid fractures managed in a collar alone should be con-sidered when electing this treatment option (see Chapter 17).Sherk and Nicholson (48) described a single patient success-fully treated with traction reduction and then immobilizationin a Minerva brace. Segal et al. (46) treated two patients withtraction and then rigid immobilization. Some authors havedescribed the treatment of C1–Type II odontoid combination

fractures with rigid immobilization (halo, SOMI, Minerva) (8,12, 14, 26, 32, 36, 46). Dickman et al. (12) described five of sixpatients successfully treated in this way (83% success rate).All six patients had an atlantoaxial interval of less than 6 mm.Halo immobilization failed in one patient with an atlantoaxialinterval of 5 mm, and the patient required posterior C1–C2fusion at 12 weeks postinjury. Segal et al. (46) described threecases of C1–Type II odontoid combination fracture success-fully treated with halo immobilization. Andersson et al. (2)described two patients older than 65 years with this combi-nation fracture injury who were successfully treated with ahalo device. Seybold and Bayley (47) added two more patientstreated with a halo resulting in successful union. Additionalsingle cases managed with halo immobilization have beendescribed (8, 14, 26, 32, 36, 43).

The C1–Type II odontoid combination fracture has beensuccessfully managed with surgical stabilization and fusion.Dickman et al. (12) treated four patients with C1–Type IIodontoid combination fractures with early surgical fusionbased on an atlantoaxial interval of 6 mm or more. Threepatients had posterior C1–C2 fusion, and one patient under-went occipitocervical fusion for multiple fractures of the pos-terior atlantal arch. Andersson et al. (2) treated one patientwith C1–Type II odontoid combination fracture with posteriorC1–C2 fusion in a series of elderly patients. Coyne et al. (10)also treated one patient with this injury pattern with a C1–C2posterior fusion. Several authors have suggested that the C1arch fracture be allowed to heal before undertaking definitiveatlantoaxial arthrodesis for this subtype of combination frac-tures. Other authors have suggested using onlay bone graftfor C1–C2 fusion and then halo immobilization in the settingof posterior C1 arch incompetence (29, 34, 37). Lee et al. (32)described the surgical management of two patients with C1–Type II odontoid combination fractures in whom posteriorC1–C2 fusion was performed. Guiot and Fessler (24) de-scribed two patients with this combination injury patterntreated posteriorly with C1–C2 transarticular screw fixationand fusion. Some investigators have used anterior odontoidscrew fixation in the treatment of C1–Type II odontoid com-bination fractures. Montesano et al. (39), in 1991, describedfour cases successfully managed in this fashion. Berlemannand Schwarzenbach (5) published an additional four cases.The report by Guiot and Fessler (24) included six patients inwhom odontoid screw fixation was accomplished. These au-thors added anterior transarticular fixation in one patient. In1999, Henry et al. (28) described a fusion success rate of 90%in 10 patients with C1–Type II odontoid combination fracturestreated with anterior odontoid screw fixation. Apostolides etal. (3) described a single case in which three screws wereplaced, all from an anterior trajectory, to stabilize the C1–C2articulation bilaterally and the odontoid fracture. Occipitocer-vical fusion has been reported in the management of C1–TypeII odontoid combination fractures (1, 2, 12, 27, 31, 42). It seemsthat this approach is reserved for patients with disruption ofthe C1 arch and gross C1–C2 instability.

In summary, a variety of treatment options have been ef-fective in C1–Type II odontoid combination fractures. Exter-nal orthoses have been successfully used in the management



of most of these injuries. Combination fractures of this sub-type with C1–C2 instability as defined by an atlantodensinterval of 5 mm or more have a high failure rate with externalimmobilization alone and have been successfully managedwith operative reduction, internal fixation, and fusion.

Treatment of C1–Type III odontoidcombination fractures

Dickman et al. (12) described five patients with C1–Type IIIodontoid combination fractures. All were successfully treatedwith halo immobilization for an average of 12 weeks. Ekong etal. (14) identified two similar cases. One was managed suc-cessfully in a halo. In the second patient, halo immobilizationfailed, and a delayed posterior C1–C2 fusion was required.Guiot and Fessler (24) reported a patient with a C1–Type IIIodontoid-hangman’s combination fracture, which they suc-cessfully treated with ventral odontoid screw fixation andthen posterior pedicle screw fixation and fusion. It seems thatexternal immobilization is effective in the management ofthese injuries in most patients.

Treatment of C1–hangman’s combination fractures

Most reported combination injuries of the atlas and theposterior elements of the axis have been successfully man-aged with semirigid or rigid external immobilization (with orwithout initial traction) (7, 9, 12, 15, 32, 46, 49). Coric et al. (9)and Lee et al. (32) described the successful treatment of ninepatients with this combination fracture type with a cervicalcollar only. Dickman et al. (12) reported three patients withC1–hangman’s combination fractures successfully treatedwith either a halo or SOMI device. The reports of Elliott et al.(15), Brashear et al. (7), Segal et al. (46), Govender and Charles(22), and Zavanone et al. (49) each describe patients withsimilar injuries successfully treated with nonoperative tech-niques. As with an isolated unstable hangman’s fracture, sur-gical fixation may be an option. The report by Fielding et al.(17) included 15 patients with C1–hangman’s combinationfractures. These authors recommended that fractures withangulation between C2 and C3 of 11 degrees or more betreated surgically. These combination fractures with angula-tion of more than 11 degrees were associated with an 85%nonunion rate with nonoperative management, in their expe-rience. This combination injury subtype seems to be managedeffectively with external immobilization alone. Unstable inju-ries, as defined by C2–C3 angulation of 11 degrees or more,may require surgical management.

Treatment of C1-miscellaneous C2 bodycombination fractures

Combination fractures of the atlas associated with miscel-laneous axis body fractures have been treated with both rigidand nonrigid immobilization (6, 11, 12, 20, 32, 44). Dickman etal. (12) reported seven cases of this combination fracturesubtype treated successfully with either a halo or SOMI brace.The cases described by Fujimura et al. (20), Lee et al. (32),Craig and Hodgson (11), and Bohay et al. (6) were all man-aged successfully with a cervical collar alone. A single case

described by Polin et al. (44) was treated with traction andsubsequent halo immobilization. Nonoperative managementof this combination injury subtype is effective.

SUMMARY

Combination fractures involving fractures of both the atlasand axis occur relatively frequently. A higher incidence ofneurological deficit is associated with C1–C2 combinationfractures compared with either C1 or C2 fractures in isolation.The C1–Type II odontoid combination fracture seems to be themost common combination injury subtype, and then C1–miscellaneous axis, C1–Type III odontoid, and C1–hangman’scombination fractures. No Class I or Class II evidence ad-dressing the management of patients with combination atlasand axis fractures is available. All of the articles revieweddescribe case series or case reports containing Class III evi-dence supporting a variety of treatment strategies for theseunique fracture injuries.

In most circumstances, the specifics of the axis fracture willdictate the most appropriate management of the combinationfracture injury. As reported for isolated atlas and axis frac-tures, most atlas-axis combination fractures can be effectivelytreated with rigid external immobilization. Combinationatlas-axis fractures with an atlantoaxial interval of 5 mm ormore or angulation of C2–C3 of 11 degrees or more may beconsidered for surgical fixation and fusion. The integrity ofthe ring of the atlas must often be taken into account whenplanning a specific surgical strategy using instrumentationand fusion techniques. If the posterior arch of C1 is inade-quate, both incorporation of the occiput into the fusion con-struct (occipitocervical fusion) and posterior C1–C2 transar-ticular screw fixation and fusion have been successful.


The identification of which of the atlas-axis combinationfracture subtypes are at greatest risk for nonunion and sub-sequent instability would be useful in determining appropri-ate management for combination fracture injuries. A uniformand clinically useful definition of cranial, C1, and C2 instabil-ity in association with these fractures would be of benefit.Prospective data collection and case-control studies at manyinstitutions would provide meaningful data addressing theseissues. The relative infrequency of combined atlas-axis frac-tures would make a randomized study difficult. Patients witha C1–Type II odontoid combination fracture should be stud-ied to compare operative and nonoperative management andshould be evaluated in terms of management morbidity, long-term success, economic benefit, patient satisfaction, and re-turn to preinjury activities.


REFERENCES

1. Anderson LD, D’Alonzo RT: Fractures of the odontoid process ofthe axis. J Bone Joint Surg Am 56A:1663–1674, 1974.




3. Apostolides PJ, Theodore N, Karahalios DG, Sonntag VKH: Tripleanterior screw fixation of an acute combination atlas-axis fracture:Case report. J Neurosurg 87:96–99, 1997.

4. Benzel EC, Hart BL, Ball PA, Baldwin NG, Orrison WW, EspinosaMC: Fractures of the C-2 vertebral body. J Neurosurg 81:206–212,1994.


6. Bohay D, Gosselin RA, Contreras DM: The vertical axis fracture:A report on three cases. J Orthop Trauma 6:416–419, 1992.

7. Brashear R Jr, Venters G, Preston ET: Fractures of the neural archof the axis: A report of twenty-nine cases. J Bone Joint Surg Am57A:879–887, 1975.

8. Castillo M, Mukherji SK: Vertical fractures of the dens. AJNRAm J Neuroradiol 17:1627–1630, 1996.

9. Coric D, Wilson JA, Kelly DL Jr: Treatment of traumatic spon-dylolisthesis of the axis with nonrigid immobilization: A reviewof 64 cases. J Neurosurg 85:550–554, 1996.


11. Craig JB, Hodgson BF: Superior facet fractures of the axis verte-bra. Spine 16:875–877, 1991.

12. Dickman CA, Hadley MN, Browner C, Sonntag VKH: Neurosur-gical management of acute atlas-axis combination fractures: Areview of 25 cases. J Neurosurg 70:45–49, 1989.

13. Effendi B, Roy D, Cornish B, Dussault RG, Laurin CA: Fracturesof the ring of the axis: A classification based on the analysis of 131cases. J Bone Joint Surg Br 63B:319–327, 1981.


15. Elliott JM Jr, Rogers LF, Wissinger JP, Lee JF: The hangman’sfracture: Fractures of the neural arch of the axis. Radiology 104:303–307, 1972.

16. Esses SI, Bednar DA: Screw fixation of odontoid fractures andnonunions. Spine 16[Suppl 10]:S483–S485, 1991.


18. Fowler JL, Sandhu A, Fraser RD: A review of fractures of the atlasvertebra. J Spinal Disord 3:19–24, 1990.

19. Fujimura Y, Nishi Y, Chiba K, Kobayashi K: Prognosis of neuro-logical deficits associated with upper cervical spine injuries. Para-plegia 33:195–202, 1995.

20. Fujimura Y, Nishi Y, Kobayashi K: Classification and treatment ofaxis body fractures. J Orthop Trauma 10:536–540, 1996.

21. Gleizes V, Jacquot FP, Signoret F, Feron JM: Combined injuries inthe upper cervical spine: Clinical and epidemiological data over a14-year period. Eur Spine J 9:386–392, 2000.

22. Govender S, Charles RW: Traumatic spondylolisthesis of the axis.Injury 18:333–335, 1987.

23. Greene KA, Dickman CA, Marciano FF, Drabier JB, Hadley MN,Sonntag VKH: Acute axis fractures: Analysis of management andoutcome in 340 consecutive cases. Spine 22:1843–1852, 1997.

24. Guiot B, Fessler RG: Complex atlantoaxial fractures. J Neurosurg91[Suppl 2]:139–143, 1999.


26. Hanssen AD, Cabanela ME: Fractures of the dens in adult pa-tients. J Trauma 27:928–934, 1987.

27. Hays MB, Bernhang AM: Fractures of the atlas vertebra: A three-part fracture not previously classified. Spine 17:240–242, 1992.

28. Henry AD, Bohly J, Grosse A: Fixation of odontoid fractures by ananterior screw. J Bone Joint Surg Br 81B:472–477, 1999.

29. Jeanneret B, Magerl F: Primary posterior fusion C1/2 in odontoidfractures: Indications, technique, and results of transarticularscrew fixation. J Spinal Disord 5:464–475, 1992.

30. Jefferson G: Fractures of the atlas vertebra: Report of four casesand a review of those previously reported. Br J Surg 7:407–422,1920.


32. Lee TT, Green BA, Petrin DR: Treatment of stable burst fracture ofthe atlas (Jefferson fracture) with rigid cervical collar. Spine 23:1963–1967, 1998.

33. Levine AM, Edwards CC: The management of traumatic spon-dylolisthesis of the axis. J Bone Joint Surg Am 67A:217–226, 1985.

34. Levine AM, Edwards CC: Treatment of injuries in the C1–C2complex. Orthop Clin North Am 17:31–44, 1986.

35. Levine AM, Edwards CC: Fractures of the atlas. J Bone Joint SurgAm 73A:680–691, 1991.

36. Lind B, Nordwall A, Sihlbom H: Odontoid fractures treated withhalo-vest. Spine 12:173–177, 1987.

37. Lipson SJ: Fractures of the atlas associated with fractures of theodontoid process and transverse ligament ruptures. J Bone JointSurg Am 59A:940–943, 1977.

38. Mirvis SE, Young JW, Lim C, Greenberg J: Hangman’s fracture:Radiologic assessment in 27 cases. Radiology 163:713–717, 1987.

39. Montesano PX, Anderson PA, Schlehr F, Thalgott JS, Lowrey G:Odontoid fractures treated by anterior odontoid screw fixation.Spine 16[Suppl 3]:S33–S37, 1991.


41. Muller EJ, Wick M, Muhr G: Traumatic spondylolisthesis of theaxis: Treatment rationale based on the stability of the differentfracture types. Eur Spine J 9:123–128, 2000.

42. Pedersen AK, Kostuik JP: Complete fracture-dislocation of theatlantoaxial complex: Case report and recommendations for anew classification of dens fractures. J Spinal Disord 7:350–355,1994.

43. Pepin JW, Hawkins RJ: Traumatic spondylolisthesis of the axis:Hangman’s fracture. Clin Orthop 157:133–138, 1981.

44. Polin RS, Szabo T, Bogaev CA, Replogle RE, Jane JA: Nonopera-tive management of Types II and III odontoid fractures: ThePhiladelphia collar versus the halo vest. Neurosurgery 38:450–457, 1996.

45. Ryan MD, Henderson JJ: The epidemiology of fractures andfracture-dislocations of the cervical spine. Injury 23:38–40, 1992.



48. Sherk HH, Nicholson JT: Fractures of the atlas. J Bone Joint SurgAm 52A:1017–1024, 1970.

49. Zavanone M, Guerra P, Rampini P, Crotti F, Vaccari U: Traumaticfractures of the craniovertebral junction: Management of 23 cases.J Neurosurg Sci 35:17–22, 1991.



CHAPTER 19

Os Odontoideum

RECOMMENDATIONSDIAGNOSIS:Standards: There is insufficient evidence to support diagnostic standards.Guidelines: There is insufficient evidence to support diagnostic guidelines.Options: Plain x-rays of the cervical spine (anteroposterior, open-mouth odontoid, and lateral) and plain

dynamic lateral x-rays performed in flexion and extension are recommended. Tomography (computed orplain) and/or magnetic resonance imaging of the craniocervical junction may be considered.

MANAGEMENT:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options:• Patients with os odontoideum, either with or without C1–C2 instability, who have neither symptoms nor

neurological signs may be managed with clinical and radiographic surveillance.• Patients with os odontoideum, particularly with neurological symptoms and/or signs, and C1–C2 instability

may be managed with posterior C1–C2 internal fixation and fusion.• Postoperative halo immobilization as an adjunct to posterior internal fixation and fusion is recommended

unless successful C1–C2 transarticular screw fixation and fusion can be accomplished.• Occipitocervical fusion with or without C1 laminectomy may be considered in patients with os odontoi-

deum who have irreducible cervicomedullary compression and/or evidence of associated occipitoatlantalinstability.

• Transoral decompression may be considered in patients with os odontoideum who have irreducible ventralcervicomedullary compression.

RATIONALE

The definition of os odontoideum is uniform throughoutthe literature: an ossicle with smooth circumferentialcortical margins representing the odontoid process that

has no osseous continuity with the body of C2 (16, 22). Theetiology of os odontoideum remains debated in the literaturewith evidence for both acquired and congenital causes (18, 23,25). The etiology of os odontoideum, however, does not playan important role in its diagnosis or subsequent management.

Diagnosis

Os odontoideum can present with a wide range of clinicalsymptoms and signs; it can also be an incidental finding onimaging. The literature has focused on three groups of pa-tients with os odontoideum: 1) those with occipitocervicalpain alone, 2) those with myelopathy, and 3) those withintracranial symptoms or signs from vertebrobasilar ischemia(4). Patients with os odontoideum and myelopathy have beensubcategorized further into those with 1) transient myelopa-thy (commonly after trauma), 2) static myelopathy, and 3)progressive myelopathy (10). Because patients with occipito-

cervical pain, myelopathy, or vertebrobasilar ischemia likelywill have etiologies other than os odontoideum, the diagnosisof os odontoideum is not usually considered until imaging isobtained. The presence of an os odontoideum is usually firstsuggested after obtaining plain cervical spine x-rays. Mostoften, plain cervical spine x-rays are sufficient to obtain adiagnosis (15).

Os odontoideum has been classified into two anatomictypes, orthotopic and dystopic. Orthotopic defines an ossiclethat moves with the anterior arch of C1, whereas dystopicdefines an ossicle that is functionally fused to the basion. Thedystopic os odontoideum may sublux anterior to the arch ofC1 (10). Tomography and computed tomography (CT) havebeen used to better define the bony anatomy of the os odon-toideum and the odontoid process. Plain dynamic x-rays inflexion/extension have been used to depict the degree ofabnormal motion between C1 and C2. Most often, there isanterior instability, with the os odontoideum subluxing for-ward in relation to the body of C2. However, at times one willsee either no discernible instability or “posterior instability,”with the os odontoideum moving posteriorly into the spinalcanal during neck extension (10, 20).


With regard to diagnosis, there are two issues of concern inthe imaging of os odontoideum. First, although plain x-raysare often diagnostic for os odontoideum, the sensitivity andspecificity of plain cervical x-rays for os odontoideum havenot been reported. The usefulness of confirmatory studies,such as CT and plain tomography and magnetic resonanceimaging (MRI), has not been well defined. Second, after thediagnosis of os odontoideum on plain cervical x-rays, insta-bility and osseous anomalies associated with os odontoideumcan influence clinical management. The best methods of fur-ther evaluating or excluding these complicating factors de-serve definition.

Management

The natural history of untreated os odontoideum covers awide spectrum. The literature provides many examples ofboth asymptomatic and symptomatic patients with known osodontoideum who have never been treated and who have hadno reported new problems in follow-up over many years (22).Conversely, examples of sudden spinal cord injury in associ-ation with os odontoideum after minor trauma have also beenreported (17). The natural history of os odontoideum is vari-able, and predictive factors for deterioration, particularly inthe asymptomatic patient, have not been identified. Indica-tions for surgical stabilization include the simple existence ofan os odontoideum, os odontoideum in association with oc-cipitocervical pain alone, and/or os odontoideum in associa-tion with neurological deficit (10, 22). Other factors that mayassist in determining the need for stabilization and/or decom-pression include C1–C2 instability, associated deformities,and spinal cord compression. A variety of techniques havebeen used to stabilize C1 and C2 in patients with os odontoi-deum (2, 3, 5, 6, 10, 20, 22, 26, 27). Fusion success rates andcomplication rates for these various procedures may provideevidence as to whether a preferred method of C1–C2 arthro-desis is supported by the literature.

Finally, neural compression is an important considerationin patients with os odontoideum. Neural compression may beanterior from a combination of bone and soft tissue, or pos-terior from the dorsal arch of C1. Surgical techniques to sta-bilize and fuse across the craniocervical junction with or with-out C1 laminectomy and techniques that provide ventraldecompression have been reported in the treatment of os odon-toideum with irreducible neural compression (6, 24). The litera-ture will be examined in light of the risks and benefits thesetechniques may provide to patients with os odontoideum.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The key phrase “os odontoideum” identified 121articles. Articles written in English were reviewed. Twenty-seven articles that described the clinical aspects and manage-

ment of patients with os odontoideum were identified andused to generate these guidelines. None of the articles meet-ing selection criteria represented Class I or Class II studies. All27 provided Class III evidence regarding the diagnosisand/or management of os odontoideum. These 27 articlesrepresent the basis for this review and are summarized inTable 19.1. In addition, one article germane to the topic but notmeeting criteria for inclusion in the Evidentiary Tables isreferenced in the Scientific Foundation section and includedin the references.


Diagnostic evaluation

There is no literature that describes the sensitivity andspecificity of imaging studies for os odontoideum. Dai et al.(6), in their review of 44 patients with os odontoideum, usedtomography, CT, and MRI in addition to “routine” plaincervical x-rays (anteroposterior, lateral, open-mouth, flexion/extension x-rays) in 39, 27, and 22 patients, respectively. Mat-sui et al. (16) described only the plain x-rays of 12 patientswith os odontoideum. They excluded patients with Down’ssyndrome and Klippel-Feil anomalies. The authors made nomention of any other studies to obtain or confirm the diag-nosis in these 12 patients. Likewise, Watanabe et al. (27) andSpierings and Braakman (22) described the plain x-rays of 34and 37 patients, respectively, with os odontoideum, withoutreference to other imaging studies. Fielding et al. (10) de-scribed 35 patients with os odontoideum in which each pa-tient had extensive roentgenographic investigation, includingmultiple roentgenograms of the cervical spine and oftenflexion-extension lateral roentgenograms and flexion-extension laminagrams (10). No mention was made as towhether additional studies beyond static plain cervical spinex-rays were necessary to confirm the diagnosis of os odontoi-deum in their series of patients.

The literature supports the ability of plain cervical spinex-rays to establish the diagnosis of os odontoideum. There isno compelling evidence in the literature that supports theneed for additional studies to confirm the diagnosis of osodontoideum. Specific characteristics or associated abnormal-ities of os odontoideum, including C1–C2 instability, softtissue masses, spinal canal diameter, associated osseousanomalies, spinal cord appearance, and vertebral artery com-promise have been investigated with a variety of imagingstudies. The imaging of abnormal motion and spinal cordcompression in association with os odontoideum has receivedthe most attention in the reported clinical series.

Instability of C1–C2 in association with os odontoideum hasbeen investigated with multiple imaging modalities. Usingflexion/extension lateral cervical spine x-ray studies in 33patients, Fielding et al. (10) reported 22 patients (67%) withanterior instability who had a mean atlantodens interval of

Os Odontoideum


TAB

LE19

.1.

Sum

mar

yof

Rep

orts

onO

sO

dont

oide

uma

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ceC

lass

Con

clus

ions

Apf

elba

umet

al.,

2000

(1)

18pa

tient

sw

ithod

onto

idfr

actu

res

incu

rred

�18

mo

befo

retr

eatm

ent

who

wer

e

trea

ted

with

ante

rior

odon

toid

scre

wfix

atio

n.

III16

patie

nts

with

follo

w-u

p.25

%fu

sion

rate

.3

with

scre

wfr

actu

rean

d1

with

scre

w

pullo

ut.

Bro

ckm

eyer

etal

.,20

00(2

)R

evie

wof

tran

sart

icul

arsc

rew

plac

emen

tin

31ch

ildre

n.12

child

ren

with

os

odon

toid

eum

(age

d4–

16yr

).

III55

/62

poss

ible

site

sde

emed

suita

ble

for

tran

sart

icul

arsc

rew

s.A

llch

ildre

nw

ithos

odon

toid

eum

wer

eab

leto

have

2sc

rew

spl

aced

.

Dai

etal

.,20

00(6

)A

revi

ewof

44pa

tient

sag

ed7–

56yr

with

osod

onto

ideu

m.

Mea

nfo

llow

-up,

6.5

yr.

III7

patie

nts

wer

eas

ympt

omat

ic.

5of

thes

e7

wer

etr

eate

dw

itha

cerv

ical

colla

ron

ly

and

have

rem

aine

dst

able

.39

unde

rwen

tfu

sion

succ

essf

ully

(9at

lant

oaxi

alan

d33

occi

pito

cerv

ical

).Sy

mpt

oms

and

sign

sdi

sapp

eare

din

26an

dim

prov

edin

13.

Tagg

ard

etal

.,20

00(2

4)A

revi

ewof

cran

iove

rteb

ral

junc

tion

abno

rmal

ities

in36

Dow

n’s

synd

rom

e

patie

nts.

Os

odon

toid

eum

pres

ent

in12

.

III27

unde

rwen

tsu

rgic

alpr

oced

ures

.O

f11

with

basi

lar

inva

gina

tion,

itw

as

irre

duci

ble

in5,

and

tran

sora

lde

com

pres

sion

was

perf

orm

ed.

Dic

kman

and

Sonn

tag,

1998

(7)

Rev

iew

of12

1pa

tient

str

eate

dw

ithpo

ster

ior

C1–

C2

tran

sart

icul

arsc

rew

san

d

wir

edpo

ster

ior

bone

stru

ts.

Os

odon

toid

eum

was

pres

ent

in9.

This

grou

pw

as

com

pare

dw

ith74

patie

nts

trea

ted

with

post

erio

rw

irin

gte

chni

ques

alon

e.

III2

failu

res

inth

etr

ansa

rtic

ular

grou

p.Th

eet

iolo

gyof

the

C1–

C2

inst

abili

tyw

asno

t

stat

edfo

rth

ese

2fa

ilure

s.1/

8pa

tient

sw

ithos

odon

toid

eum

inth

epo

ster

ior

wir

ing

grou

pha

da

nonu

nion

.O

vera

llfu

sion

rate

for

tran

sart

icul

arw

as98

%ve

rsus

86%

for

post

erio

rw

irin

gte

chni

ques

.

Wan

get

al.,

1999

(26)

16ch

ildre

ntr

eate

dfo

rat

lant

oaxi

alin

stab

ility

.4

ofth

emha

dos

odon

toid

eum

and

wer

etr

eate

dw

ithC

1–C

2tr

ansa

rtic

ular

scre

ws

and

post

erio

rw

irin

gan

dfu

sion

tech

niqu

es.

IIIA

llfu

sed.

No

halo

imm

obili

zatio

n.

Tran

sart

icul

arsc

rew

sw

ere

succ

essf

ully

used

inch

ildre

nas

youn

gas

4yr

.

Kuh

nset

al.,

1998

(14)

4ch

ildre

nw

ithos

odon

toid

eum

unde

rwen

tM

RI

exam

inat

ions

.III

All

4ch

ildre

nha

dch

ange

sin

the

nuch

alco

rdco

nsis

tent

with

inju

ry.

Low

ryet

al.,

1997

(15)

Are

view

of25

child

ren

requ

irin

gup

per

cerv

ical

fusi

ons.

11ch

ildre

nha

dos

odon

toid

eum

.

III10

unde

rwen

ta

Bro

oks-

type

C1–

C2

fusi

on.

2of

thes

ech

ildre

ndi

dno

tfu

se.

1

unde

rwen

ta

Gal

lie-t

ype

fusi

on.

This

child

rem

aine

dun

stab

lean

dw

asre

vise

dto

a

Bro

oks-

type

fusi

on,

whi

chw

assu

cces

sful

.

Mat

sui

etal

.,19

97(1

6)R

evie

wof

the

plai

nra

diog

raph

icm

orph

olog

yof

C2

eval

uate

din

12pa

tient

s(1

5–

71yr

old)

with

osod

onto

ideu

mun

rela

ted

toan

ysy

ndro

me.

III3

conf

igur

atio

nsde

scri

bed

from

anan

tero

post

erio

rvi

ew:

roun

d,co

ne,

and

blun

t-

toot

h.M

yelo

path

yw

asm

ore

seve

rein

the

grou

pw

itha

roun

dco

nfig

urat

ion.

Ver

ska

and

And

erso

n,19

97(2

5)R

epor

tof

apa

irof

iden

tical

twin

s,1

with

osod

onto

ideu

m,

and

1w

ithou

tos

odon

toid

eum

.

IIIH

isto

ryof

trau

ma

inth

etw

inw

ithan

osod

onto

ideu

m.

Fell

atag

e3

yr,

had

tort

icol

lisan

dne

ckpa

info

rse

vera

lm

onth

s.

Wat

anab

eet

al.,

1996

(27)

Rev

iew

of34

patie

nts

with

osod

onto

ideu

m(5

–76

yrol

d).

Div

ided

into

grou

psby

Row

land

clas

sific

atio

n(1

�lo

cal

sym

ptom

s,2

�po

sttr

aum

atic

tran

sien

t

mye

lopa

thy,

3,4

�pr

ogre

ssiv

em

yelo

path

yor

intr

acra

nial

sym

ptom

s).

Late

ral

neut

ral

and

dyna

mic

x-ra

ysob

tain

ed.

Sagi

ttal

plan

ero

tatio

nan

gle,

min

imum

dist

ance

,an

din

stab

ility

inde

xw

ere

mea

sure

d.

IIILo

wco

rrel

atio

nbe

twee

nsa

gitta

lpl

ane

rota

tion

angl

ean

din

stab

ility

inde

x.Sa

gitta

l

plan

ero

tatio

nan

gle

of�

20de

gree

sor

inst

abili

tyin

dex

of�

40%

corr

elat

esw

ith

mye

lopa

thy.

Cle

men

tset

al.,

1995

(4)

Rep

ort

ofno

nope

rativ

etr

eatm

ent

ofan

inci

dent

ally

disc

over

edos

odon

toid

eum

with

out

C1–

C2

inst

abili

tyat

diag

nosi

s.

IIIA

fter

5yr

,pr

ofou

ndC

1–C

2in

stab

ility

and

sym

ptom

sha

dde

velo

ped,

nece

ssita

ting

post

erio

rin

stru

men

tatio

nan

dfu

sion

.

Coy

neet

al.,

1995

(5)

Rev

iew

ofpo

ster

ior

C1–

C2

fusi

onan

din

stru

men

tatio

nte

chni

ques

.5/

32pa

tient

s

had

osod

onto

ideu

m.

III3/

5w

ithos

odon

toid

eum

faile

dw

ithpo

ster

ior

wir

ing

tech

niqu

es.

All

wer

e

imm

obili

zed

inha

los.

2/5

deve

lope

dne

wne

urol

ogic

alde

ficits

asop

erat

ive

com

plic

atio

ns.

Stev

ens

etal

.,19

94(2

3)R

evie

wof

abno

rmal

odon

toid

san

dC

1–C

2in

stab

ility

.24

/62

patie

nts

with

os

odon

toid

eum

.9

child

ren

and

15ad

ults

.

IIIPe

riod

onto

idso

fttis

sue

thic

keni

ngw

aspr

esen

ton

lyin

thos

ew

ithM

orqu

io’s

dise

ase.

Afte

rfu

sion

,th

eod

onto

idw

asno

ted

topa

rtia

llyor

com

plet

ely

rege

nera

te

inca

ses

ofM

orqu

io’s

dise

ase.

Men

ezes

and

Ryk

en,

1992

(17)

Rev

iew

of18

Dow

n’s

synd

rom

epa

tient

sw

ithsy

mpt

omat

icce

rvic

omed

ulla

ry

com

prom

ise.

4ha

dos

odon

toid

eum

.

IIIA

ll4

had

gros

sin

stab

ility

ondy

nam

icx-

rays

.Su

cces

sful

fusi

onw

ithpo

ster

ior

wir

ing

tech

niqu

esan

dfu

ll-th

ickn

ess

rib

graf

ts.

Imm

obili

zed

for

a“m

inim

umof

3m

onth

s.”

Dic

kman

etal

.,19

91(8

)R

evie

wof

36pa

tient

str

eate

dw

ithC

1–C

2po

ster

ior

wir

ing

and

fusi

onfo

rva

riou

s

reas

ons.

4pa

tient

s(a

ged

16,

25,

38,

43yr

)ha

dos

odon

toid

eum

.A

llpa

tient

s

wer

em

aint

aine

din

aha

lofo

r12

wk

afte

rsu

rger

y.

IIIO

fth

e4

with

osod

onto

ideu

m,

3de

velo

ped

osse

ous

unio

nsan

d1

had

ast

able

fibro

usun

ion

(follo

w-u

p,15

–44

mo)

.N

oco

mpl

icat

ions

for

thes

e4

patie

nts.

Hos

ono

etal

.,19

91(1

3)C

iner

adio

grap

hic

eval

uatio

nof

6pa

tient

sw

ithos

odon

toid

eum

.III

2ty

pes

ofC

1po

ster

ior

arch

tran

slat

ion:

stra

ight

(ver

tical

)(n

�4)

and

sigm

oid

(n�

2).

Cor

rela

ted

abno

rmal

mot

ion

with

biom

echa

nics

ofpo

ster

ior

wir

ing

tech

niqu

es.

Smith

etal

.,19

91(2

1)R

evie

wof

17ch

ildre

nop

erat

edon

for

C1–

C2

inst

abili

ty.

11ha

dos

odon

toid

eum

.Po

ster

ior

wir

ing

tech

niqu

es,

auto

logo

usbo

ne,

and

halo

used

inal

l.

III2/

11w

ithos

odon

toid

eum

had

nonu

nion

s.1

cord

inju

ryth

ough

tse

cond

ary

to

subl

amin

arw

ire

pass

age.



TAB

LE19

.1.

Con

tinu

ed

Seri

es(R

ef.

No.

)D

escr

iptio

nof

Stud

yEv

iden

ceC

lass

Con

clus

ions

Shir

asak

iet

al.,

1991

(20)

9pa

tient

sw

ithos

odon

toid

eum

and

post

erio

rin

stab

ility

had

3ra

diog

raph

ic

para

met

ers

mea

sure

d.D

ista

nce

betw

een

the

osan

dC

2sp

inou

spr

oces

sin

exte

nsio

n(D

ext),

dist

ance

betw

een

the

osan

dpo

ster

ior

C1

arch

(Dat

l),an

d

“deg

ree

ofin

stab

ility

”(In

st).

Thes

efin

ding

sw

ere

com

pare

dto

thei

rne

urol

ogic

al

stat

us.

IIITh

ose

with

out

hist

ory

orev

iden

ceof

mye

lopa

thy

had

aD

ext

of�

16m

m.

Dex

tw

as

�16

mm

inth

ose

with

mye

lopa

thy.

The

pres

ence

orab

senc

eof

mye

lopa

thy

was

not

rela

ted

toth

eIn

st.

Inth

ose

with

mye

lopa

thy

and

Dat

l�

13m

m,

ther

ew

as

reve

rsib

leco

rdco

mpr

essi

onin

exte

nsio

n;in

thos

ew

itha

Dat

lof

�13

mm

,th

eco

rd

rem

aine

dco

mpr

esse

din

flexi

onan

dex

tens

ion.

Mor

gan

etal

.,19

89(1

8)R

epor

tof

3fa

mily

mem

bers

with

C2–

C3

Klip

pel-

Feil

abno

rmal

ities

and

os

odon

toid

eum

.

IIIA

ges

16(in

dex

case

),39

(fath

er),

and

64yr

(pat

erna

lgr

andm

othe

r).

Non

ew

ith

neur

olog

ical

sign

sor

sym

ptom

s.

Yam

ashi

taet

al.,

1989

(28)

Cor

rela

tion

ofcl

inic

alst

atus

,M

RI

scan

s,an

dx-

rays

in29

patie

nts

with

C1–

C2

inst

abili

ty.

4ha

dos

odon

toid

eum

.

IIITh

eA

DI

did

not

corr

elat

ew

ithth

ede

gree

ofm

yelo

path

y,bu

tM

RI

degr

eeof

cord

com

pres

sion

did

corr

elat

ew

ithde

gree

ofm

yelo

path

y.

Fren

chet

al.,

1987

(11)

Rev

iew

ofdy

nam

icce

rvic

alsp

ine

x-ra

ysin

185

patie

nts

with

Dow

n’s

synd

rom

e.III

6ha

dab

norm

alod

onto

ids

cons

iste

ntw

ithos

odon

toid

eum

for

anin

cide

nce

of3%

.

3ha

dpr

evio

usx-

rays

show

ing

noab

norm

ality

.1

had

anex

agge

rate

dA

DI

of6

mm

.

Spie

ring

san

dB

raak

man

,19

82(2

2)37

patie

nts

with

osod

onto

ideu

m.

20tr

eate

dco

nser

vativ

ely.

IIIO

f20

man

aged

cons

erva

tivel

y,1

was

lost

tofo

llow

-up.

15ha

dno

mye

lopa

thy

(med

ian

follo

w-u

p,5

yr),

and

none

deve

lope

dm

yelo

path

y.O

f4

with

mye

lopa

thy

(follo

w-u

p,0.

5,1,

7,an

d14

yr),

1is

dead

from

canc

er,

1ha

sne

ckpa

in,

1ha

sne

ck

pain

and

pare

sthe

sias

,an

d1

has

head

ache

s.

Fiel

ding

etal

.,19

80(1

0)35

patie

nts

(3–6

5yr

old)

with

osod

onto

ideu

m.

25pa

tient

sw

ere

sym

ptom

atic

.III

22pa

tient

sha

dan

teri

orin

stab

ility

with

am

ean

AD

Iof

10.3

mm

.5

had

post

erio

r

inst

abili

ty.

3ha

dno

dete

ctab

lem

otio

n.3

had

�3

mm

ofC

1–C

2m

otio

n.26

unde

rwen

tpo

ster

ior

fusi

onsu

cces

sful

ly.

5w

ere

not

oper

ated

on,

3w

ere

asym

ptom

atic

with

noin

stab

ility

.Th

eyre

mai

ned

wel

lw

ithno

inst

abili

tyat

1,2,

and

3yr

,re

spec

tivel

y.1

patie

ntw

ithin

stab

ility

refu

sed

surg

ery

but

was

wel

lat

2-yr

follo

w-u

p.1

patie

ntdi

edof

rena

lfa

ilure

.

Bro

oks

and

Jenk

ins,

1978

(3)

3ch

ildre

n(8

,11

,an

d12

yrol

d)w

ithos

odon

toid

eum

trea

ted

with

subl

amin

ar

C1–

C2

wir

esan

dau

tolo

gous

iliac

cres

tgr

aft.

Min

erva

cast

imm

obili

zatio

n.

IIIA

llfu

sed.

Spon

tane

ous

exte

nsio

nof

fusi

onto

C3

in1

child

.

Dyc

k,19

78(9

)R

evie

wof

8ch

ildre

n(a

ges

7–17

yr)

with

osod

onto

ideu

m.

6w

ere

trea

ted

with

post

erio

rw

irin

gan

dfu

sion

ofC

1–C

3.Ex

tern

alim

mob

iliza

tion

for

“usu

ally

”3–

4

mo.

III6

child

ren

unde

rwen

tpo

ster

ior

fusi

onby

the

auth

or.

2re

quir

edre

oper

atio

n.

Gri

swol

det

al.,

1978

(12)

4pa

tient

sw

ithos

odon

toid

eum

wer

etr

eate

dw

ithsu

blam

inar

C1–

C2

wir

esan

d

auto

logo

usili

accr

est.

III3

fuse

d.1

did

not

fuse

afte

r3

atte

mpt

s.

aM

RI,

mag

neti

cre

sona

nce

imag

ing;

AD

I,at

lant

oden

sin

terv

al.

Os Odontoideum S151


10.3 mm, five patients (15%) with posterior instability (meanposterior translation of the os odontoideum during extension,8.4 mm), three patients (9%) with less than 3 mm of C1–C2motion, and three patients (9%) with no detectable C1–C2 mo-tion. Eight patients (23%) had both anterior and posteriorinstability. The authors noted that cineradiography was help-ful in examining range of motion at C1–C2 in these patients,but it was not of benefit in the measurement of the degree ofinstability. Of note is that almost one-fifth of the patients intheir series manifested no radiographic evidence of C1–C2instability.

Spierings and Braakman (22) studied 21 of their 37 patientswith os odontoideum with flexion/extension cervical spinex-rays or tomograms. They measured the maximal distancethe os odontoideum moved in the sagittal plane, the innerdiameter of the atlas, and the minimal spinal canal diameter(the distance between the posterior aspect of the C2 body andthe dorsal arch of C1 during flexion). They compared thesemeasurements in two groups, those with and without my-elopathy. The degree of C1–C2 instability did not correlatewith neurological status, but the measured minimal spinalcanal diameter was significantly smaller (P � 0.05) in thegroup with myelopathy. They identified 13 mm as the criticalanteroposterior spinal diameter. Watanabe et al. (27) madesimilar measurements in 34 patients using plain lateral cervi-cal x-rays in flexion/extension. As in the Spierings and Braak-man study, the degree of instability in their patients did notcorrelate with the presence of myelopathy. Shirasaki et al. (20)described radiographic findings on lateral flexion/extensionx-rays in nine patients with os odontoideum. They reportedthat a distance of 13 mm or less between the os odontoideumand the dorsal arch of C1 specifically defined severe cervicalmyelopathy (20) in their patients. They, too, found that thedegree of C1–C2 instability did not correlate with the presenceof myelopathy. Yamashita et al. (28) studied atlantoaxial sub-luxation with plain radiography and MRI and correlated theimaging studies with the degree of myelopathy in 29 patients(4 with os odontoideum). They found that the degree ofmyelopathy did not correlate with the distance of subluxationof C1 on C2 on plain x-rays. The degree of cord compressionon MRI did correlate well with the degree of myelopathymeasured clinically. Matsui et al. (16) classified os odontoi-deum into three types according to the shape of the os odon-toideum on plain x-rays. Three types were described: round,cone, and blunt-tooth. They compared these three os odon-toideum types to the degree of clinical myelopathy and foundthat the degree of myelopathy correlated most closely withthe “round” os odontoideum type. Kuhns et al. (14) describedthe MRI appearance of os odontoideum in four children andidentified signal changes within the posterior ligaments con-sistent with trauma. They could not discern whether thesechanges represented a primary or secondary phenomenonwith respect to atlantoaxial instability.

These studies provide two consistent conclusions: 1) thedegree of C1–C2 instability does not seem to correlate withneurological status in patients with os odontoideum; and 2)sagittal spinal canal diameter on plain x-rays of 13 mm or lessis strongly associated with myelopathy.

Beyond plain spine x-rays and flexion-extension x-rays,imaging to assist with operative planning of unstable os odon-toideum receives brief mention in several reports (11, 17, 24,26). Important factors to consider before proceeding withsurgical intervention for this disorder are the ability to reduceC1–C2, spinal cord compression, an assimilated atlas, an in-complete C1 ring, the course of the vertebral arteries at C1 andC2, and the presence of an associated congenital fusion of thecervical spine (e.g., Klippel-Feil). Plain x-rays, tomography,and CT provide information regarding the ability to achieveanatomic alignment of C1 on C2 and the presence or absenceof a congenital fusion. MRI is the best modality for viewingcord compression even after apparent C1–C2 realignment(28). CT can provide important information about the bonyanatomy at the craniocervical junction, including the com-pleteness of the atlas ring and the position of the transverseforamina at C1 and C2 (19). Hosono et al. (13) made interest-ing observations on the different motions of the posterior archof C1 in relation to C2 in patients with os odontoideum. Theyobserved two patterns of motion, linear and sigmoid. Theythought that, in those patients with a sigmoid-shaped motionpattern, posterior wiring techniques may not provide ade-quate stability. The selection of and necessity for additionalimaging studies in the evaluation of os odontoideum seems tobe made on a patient-by-patient basis. The literature providesno convincing evidence as to which patients should undergosupplemental imaging (CT or MRI) after the diagnosis of osodontoideum has been made.

Management

The universal theme of the various management strategiesoffered in the treatment of patients with os odontoideum hasbeen either to confirm or to secure cervical spinal stability atthe C1–C2 level. The earliest reports of os odontoideum de-scribe small pediatric case series treated surgically. In 1978,Griswold et al. (12) described four children with os odontoi-deum who underwent posterior C1–C2 wiring and autolo-gous iliac fusion. Three children had successful arthrodesis.The fourth child did not achieve fusion/stability despite threeattempts. In the same year, Brooks and Jenkins (3) describedtheir technique of C1–C2 wiring and fusion and reportedthree children with os odontoideum who were immobilizedpostoperatively in Minerva jackets. All three patientsachieved successful fusion. In summary, six of the sevenchildren with os odontoideum described in these two reportswere successfully treated.

Two larger series, reported in the early 1980s, includedadults and children with os odontoideum and described bothoperative and nonoperative management strategies for thesepatients. Fielding et al. (10) described 35 patients with osodontoideum, 27 with radiographic evidence of instability.Twenty-six of these 27 patients underwent successful poste-rior C1–C2 fusion (Gallie type). Fusions were noted to be“solid” after 2 months of immobilization in children and 3months in adults. One patient with instability refused surgeryand remained well at the 2-year follow-up examination. Theeight remaining patients with no evidence of C1–C2 instabil-



ity managed nonoperatively remained well at the lastfollow-up (1–3 yr). Spierings and Braakman (22) describedtheir management of 37 patients with os odontoideum. Sev-enteen were treated surgically. They provide 20 patients foranalysis of the natural history of os odontoideum. Informa-tion about radiographic stability was provided for only 21 ofthe 37 patients they reported. Sixteen patients in their seriespresented with neck pain only or had an incidentally discov-ered os odontoideum. Nine of these 16 patients had flexion/extension x-rays. Of these nine patients, seven had abnormalmotion of 8 mm or more. With a median follow-up of 7 years,none of these 16 patients developed a neurological deficit.Four additional patients who presented with myelopathywere treated nonoperatively with follow-up from 6 months to14 years. Three of these four patients presented with transientmyelopathy and had no recurrence at last follow-up, despiteabnormal motion of C1 on C2 of 8 to 16 mm. The fourthpatient had a stable monoparesis at last follow-up. Of the17 patients who underwent surgery, one patient had neuro-logical worsening and two died. Eight of these 17 patientstreated surgically had a posterior C1–C2 fusion. Nine patientsunderwent occipitocervical fusion with C1 laminectomy. Theauthors did not report a single failed fusion. They had acombined surgical morbidity and mortality of 18% (3 of 17patients). The authors conclude that patients with os odontoi-deum without C1–C2 instability can be managed withoutsurgical stabilization and fusion with good results. Althoughthey did not provide operative treatment to every os odon-toideum patient with C1–C2 instability, those with myelopa-thy and greater amounts of instability were more likely toundergo surgery. If these two series are considered represen-tative of patients with os odontoideum, the implication is thatminimally symptomatic or asymptomatic patients with osodontoideum without C1–C2 instability can be managed non-operatively with little or no morbidity over time. Althoughpatients with os odontoideum and myelopathy or C1–C2instability have been managed conservatively, most patientswith myelopathy or instability are treated surgically.

Clements et al. (4), in 1995, reported a patient who had adocumented os odontoideum without instability who, at 5years follow-up, developed symptomatic frank C1–C2 insta-bility that required surgical stabilization and fusion. It seemsthat a lack of C1–C2 instability at initial diagnosis does notguarantee that instability will not develop in these patients. Itis recommended, therefore, that clinical and radiographicfollow-up be provided to patients with os odontoideum whoare found to have radiographic C1–C2 stability on initialassessment.

More recent series reported in the literature provide betterdescriptions of the operative procedures and postoperative im-mobilization techniques used for patients with os odontoideum(2, 5–9, 15, 17, 21–26). Smith et al. (21) described 11 children withos odontoideum who underwent posterior wiring and at-tempted fusion. Autologous bone graft and halo immobilizationwere used in all children. Two children had fusion failure withnonunion. One child incurred an intraoperative cord injury sec-ondary to sublaminar wire passage. Lowry et al. (15) also de-scribed 11 children with os odontoideum who were treated with

C1–C2 fusion and posterior wiring. One child treated with aGallie-type procedure had continued instability and fusion fail-ure. The C1–C2 construct was revised successfully with aBrooks-type fusion procedure. The remaining 10 children weresuccessfully treated with Brooks C1–C2 wiring and fusion pro-cedures. Coyne et al. (5), in a review of posterior C1–C2 fixationtechniques, described five patients with os odontoideum. Threeof these five had unsuccessful attempted posterior fusions de-spite halo immobilization. Two developed new neurologicaldeficits after surgery. Dai et al. (6) described 44 patients with osodontoideum with a mean follow-up of 6.5 years. Seven patientswere asymptomatic at presentation. Five of these seven refusedsurgery and were treated with a cervical collar only; they re-mained stable at last follow-up. The remaining 39 patients un-derwent successful fusion procedures after skeletal traction. Theauthors reported that nine patients underwent atlantoaxial fu-sion and 33 required occipitocervical fusion (42 operations in 39patients). Symptoms and signs disappeared in 26 of their oper-ative patients and improved in the remaining 13 at last follow-up. They used occipitocervical constructs with fusion with orwithout C1 laminectomy in those patients with irreducible de-formities because of the concern that sublaminar passage ofwires or cables might result in neurological morbidity.

Wang et al. (26) reported 16 children with C1–C2 instability,four of whom had os odontoideum. These four children weretreated with C1–C2 transarticular screw fixation with poste-rior C1–C2 wiring and fusion. The youngest child was 4 yearsold. All achieved stable fusion arthrodesis without complica-tions. Halo immobilization devices were not used. Brockm-eyer et al. (2) also reported 31 children they treated withC1–C2 transarticular screw fixation and fusion. Twelve ofthese children had os odontoideum. Bilateral screws wereplaced successfully without complication in all children withos odontoideum. The authors did not comment on the type ofpostoperative immobilization devices they used. In 1991,Dickman et al. (8) reviewed their experience with fusion plus12 weeks of halo immobilization in the treatment of C1–C2instability. The authors described 36 patients with C1–C2 in-stability, four of whom had os odontoideum. Three of four osodontoideum patients they treated in this way developedosseous union. One had a stable fibrous union at last follow-up. In a subsequent report in 1998, Dickman and Sonntag (7)compared their series of patients undergoing C1–C2 transar-ticular screw fixation with posterior wiring and fusion withpatients treated with posterior wiring and fusion alone. Thefusion rates in the two groups were 98 and 86%, respectively.No patient with os odontoideum treated with C1–C2 transar-ticular fusion techniques experienced failure to fuse. Only oneof eight patients with os odontoideum in the posterior wiringand fusion group developed a nonunion (previously de-scribed). In contrast to the patients treated with posteriorwiring and fusion only, no patient treated with transarticularscrew fixation required postoperative halo immobilization.Menezes and Ryken (17) described four children with osodontoideum and Down’s syndrome whom they successfullytreated with posterior wiring and fusion, using full-thicknessautograft rib and at least 3 months of postoperative haloimmobilization. Dyck (9) reported eight children with os

Os Odontoideum S153


odontoideum, six of whom were treated with posterior C1–C3wiring and fusion techniques. All were externally immobi-lized in a four-poster brace “usually” for 3 to 4 months post-operatively. Two of six children required reoperation fornonunions.

Apfelbaum et al. (1) described their experience in treatingrecent and remote (�18 mo after injury) odontoid injurieswith anterior screw fixation. They reported a fusion rate of25% in 16 “remote” odontoid injuries. If an os odontoideumwere considered anatomically similar to a “remote” odontoidfracture, then the rate of fusion for os odontoideum treatedwith an odontoid screw fixation would also be expected to bepoor. Anterior C1–C2 transfacetal fixation techniques mayhave merit in the surgical treatment of os odontoideum, butthere are no descriptions of their application for os odontoi-deum in the literature.

The surgical treatment of patients with C1–C2 instability inassociation with os odontoideum has been demonstrated to besuccessful when combined fusion and internal fixation tech-niques are used, usually in conjunction with postoperativehalo immobilization. Fusion success rates and reports of op-erative morbidity varied considerably among the clinical caseseries reported in the literature. Although the numbers aresmall, transarticular C1–C2 screw fixation and fusion has beenassociated with higher rates of fusion than posterior wiringand fusion techniques alone. Of note is that patients treatedwith transarticular screw fixation have been managed in hardcollars postoperatively, obviating the need for halo immobi-lization devices. If transarticular screw fixation is not used inthe treatment of unstable os odontoideum, postoperative haloimmobilization as an adjunct to dorsal internal fixation andfusion is recommended.

Ventral or transoral decompression for irreducible ventralcervicomedullary compression in association with os odon-toideum has been suggested (24). Reports of the managementof ventral compression and os odontoideum are scant. In areview of 36 patients with Down’s syndrome and craniover-tebral junction abnormalities, Taggard et al. (24) described 12patients with os odontoideum. Eleven of the 36 patients re-ported were noted to have basilar invagination. Five of these11 patients with basilar invagination had irreducible ventralspinal cord compression and were treated with transoral de-compression. The authors reported stable to excellent out-comes without complications after transoral decompression inall five patients; however, the total number of patients whohad basilar invagination due to os odontoideum was notdescribed. The report implies, however, that selected patientswith atlantoaxial instability and irreducible symptomatic ven-tral cervicomedullary compression may benefit from ventraldecompression. On the other hand, Dai et al. (6) reported thesuccessful use of occipitocervical fusion with or without C1laminectomy in cases of irreducible deformity with cervi-comedullary neural compression in 33 patients with os odon-toideum. These authors described improvement in all patientsand no complications related to their dorsal-only approach.Although it may seem intuitive to remove ventral neuralcompression in association with os odontoideum, the litera-

ture suggests that dorsal stabilization and fusion withoutventral decompression is an effective management option.

SUMMARY

Plain cervical spine x-rays seem to be adequate for makinga diagnosis of os odontoideum in most patients with thisdisorder. Lateral flexion/extension x-rays can provide usefulinformation regarding C1–C2 instability. Tomography (com-puted or plain) may be helpful to define the osseous relation-ships at the cranial base, C1, and C2 in patients in whom thecraniovertebral junction is not well visualized on plain x-rays.The degree of C1–C2 instability identified on cervical x-raysdoes not correlate with the presence of myelopathy. A sagittaldiameter of less than 13 mm in the spinal canal at the C1–C2level does correlate with myelopathy detected on clinicalexamination. MRI can depict spinal cord compression andsignal changes within the cord that correlate with the pres-ence of myelopathy.

Surgical treatment is not required for every patient inwhom os odontoideum is identified. Patients who have noneurological deficit and have no instability at C1–C2 onflexion/extension studies can be managed without operativeintervention. Even patients with documented C1–C2 instabil-ity and neurological deficit have been managed nonopera-tively without clinical consequence during finite follow-upperiods. Most investigators of this disorder favor operativestabilization and fusion of C1–C2 instability associated withos odontoideum. The concern exists that patients with osodontoideum who have C1–C2 instability have an increasedlikelihood of future spinal cord injury. Although not sup-ported by Class I or Class II evidence from the literature,multiple case series (Class III evidence) suggest that stabili-zation and fusion of C1–C2 is meritorious in this circumstance(6, 15, 24, 26). Because a patient with an initially stable osodontoideum has been reported to develop delayed C1–C2instability, and because there are rare examples of patientswith stable os odontoideum who have developed neurologi-cal deficits after minor trauma, longitudinal clinical and ra-diographic surveillance of patients with os odontoideumwithout instability is recommended (4, 10).

Posterior C1–C2 arthrodesis in the treatment of os odontoi-deum provides effective stabilization of the atlantoaxial jointin most patients. Posterior wiring and fusion techniques sup-plemented with postoperative halo immobilization providedsuccessful fusion in 40 to 100% of cases reported (3, 5, 6, 22,26). Atlantoaxial transarticular screw fixation and fusionseems to have merit in the treatment of C1–C2 instability inassociation with os odontoideum and seems to obviate theneed for postoperative halo immobilization. Neural compres-sion in association with os odontoideum has been treated withreduction of deformity, dorsal decompression of irreducibledeformity, and ventral decompression of irreducible defor-mity, each in conjunction with C1–C2 or occipitocervical fu-sion and internal fixation. Each of these combined approacheshas provided satisfactory results. Odontoid screw fixation hasno role in the treatment of this disorder.




A cooperative multi-institutional natural history study ofpatients with os odontoideum without C1–C2 instabilitywould provide demographic and clinical information thatmight indicate predictive factors for the development of sub-sequent instability. In a related study, the prevalence of osodontoideum as an incidental finding should be established.The literature supports essentially no treatment for os odon-toideum, even with C1–C2 subluxation. Whether activity re-striction is called for in these patients would be helpful andshould be studied. A cooperative multi-institutional prospec-tive randomized trial of posterior wiring and fusion tech-niques with and without C1–C2 transarticular screw fixationfor patients with os odontoideum and C1–C2 instabilitywould help to definitively identify the risks and merits of eachof the two procedures in this patient population.


REFERENCES

1. Apfelbaum RI, Lonser RR, Veres R, Casey A: Direct anterior screwfixation for recent and remote odontoid fractures. J Neurosurg93[Suppl 2]:227–236, 2000.

2. Brockmeyer DL, York JE, Apfelbaum RI: Anatomic suitability ofC1–2 transarticular screw placement in pediatric patients.J Neurosurg 92[Suppl 1]:7–11, 2000.

3. Brooks AL, Jenkins EB: Atlanto-axial arthrodesis by the wedgecompression method. J Bone Joint Surg Am 60A:279–284, 1978.

4. Clements WD, Mezue W, Mathew B: Os odontoideum: Congen-ital or acquired? That’s not the question. Injury 26:640–642, 1995.


6. Dai L, Yuan W, Ni B, Jai L: Os odontoideum: Etiology, diagnosis,and management. Surg Neurol 53:106–109, 2000.

7. Dickman CA, Sonntag VKH: Posterior C1–C2 transarticular screw fix-ation for atlantoaxial arthrodesis. Neurosurgery 43:275–281, 1998.

8. Dickman CA, Sonntag VK, Papadopoulos SM, Hadley MN: Theinterspinous method of posterior atlantoaxial arthrodesis.J Neurosurg 74:190–198, 1991.

9. Dyck P: Os odontoideum in children: Neurological manifesta-tions and surgical management. Neurosurgery 2:93–99, 1978.

10. Fielding JW, Hensinger RN, Hawkins RJ: Os odontoideum.J Bone Joint Surg Am 62A:376–383, 1980.

11. French HG, Burke SW, Roberts JM, Johnston CE II, Whitecloud T,Edmunds JO: Upper cervical ossicles in Down syndrome.J Pediatr Orthop 7:69–71, 1987.

12. Griswold DM, Albright JA, Schiffman E, Johnson R, SouthwickW: Atlanto-axial fusion for instability. J Bone Joint Surg Am60A:285–292, 1978.

13. Hosono N, Yonenobu K, Ebara S, Ono K: Cineradiographic mo-tion analysis of atlantoaxial instability in os odontoideum. Spine16[Suppl 10]:S480–S482, 1991.

14. Kuhns LR, Loder RT, Farley FA, Hensinger RN: Nuchal cordchanges in children with os odontoideum: Evidence for associ-ated trauma. J Pediatr Ortho 18:815–819, 1998.

15. Lowry DW, Pollack IF, Clyde B, Albright AL, Adelson PD: Uppercervical spine fusion in the pediatric population. J Neurosurg87:671–676, 1997.

16. Matsui H, Imada K, Tsuji H: Radiographic classification of osodontoideum and its clinical significance. Spine 22:1706–1709,1997.

17. Menezes AH, Ryken TC: Craniovertebral abnormalities inDown’s syndrome. Pediatr Neurosurg 18:24–33, 1992.

18. Morgan MK, Onofrio BM, Bender CE: Familial os odontoideum:Case report. J Neurosurg 70:636–639, 1989.

19. Paramore CG, Dickman CA, Sonntag VKH: The anatomical suit-ability of the C1–2 complex for transarticular screw fixation.J Neurosurg 85:221–224, 1996.

20. Shirasaki N, Okada K, Oka S, Hosono N, Yonenobu K, Ono K: Osodontoideum with posterior atlantoaxial instability. Spine 16:706–715, 1991.

21. Smith MD, Phillips WA, Hensinger RN: Fusion of the uppercervical spine in children and adolescents: An analysis of 17patients. Spine 16:695–701, 1991.

22. Spierings EL, Braakman R: The management of os odontoideum:Analysis of 37 cases. J Bone Joint Surg Br 64B:422–428, 1982.

23. Stevens JM, Chong WK, Barber C, Kendall BE, Crockard HA: Anew appraisal of abnormalities of the odontoid process associatedwith atlantoaxial subluxation and neurological disability. Brain117:133–148, 1994.

24. Taggard DA, Menezes AH, Ryken TC: Treatment of Downsyndrome-associated craniovertebral junction abnormalities.J Neurosurg 93[Suppl 2]:205–213, 2000.

25. Verska JM, Anderson PA: Os odontoideum: A case report of oneidentical twin. Spine 22:706–709, 1997.


27. Watanabe M, Toyama Y, Fujimura Y: Atlantoaxial instability in osodontoideum with myelopathy. Spine 21:1435–1439, 1996.

28. Yamashita Y, Takahashi M, Sakamoto Y, Kojima R: Atlantoaxialsubluxation: Radiography and magnetic resonance imaging cor-related to myelopathy. Acta Radiol 30:135–140, 1989.

Plate from Gautier D, Duverney M: Essai D’anatomie, en TableauxImprimés. . . . Paris, 1745. Courtesy, Dr. Irwin J. Pincus, Los Angeles,California.

Os Odontoideum S155

CHAPTER 20

Treatment of Subaxial Cervical Spinal Injuries

RECOMMENDATIONSSUBAXIAL CERVICAL FACET DISLOCATION INJURIES:Standards: There is insufficient evidence to recommend treatment standards.Guidelines: There is insufficient evidence to recommend treatment guidelines.Options:• Closed or open reduction of subaxial cervical facet dislocation injuries is recommended.• Treatment of subaxial cervical facet dislocation injuries with rigid external immobilization, anterior arth-

rodesis with plate fixation, or posterior arthrodesis with plate or rod or interlaminar clamp fixation isrecommended.

• Treatment of subaxial cervical facet dislocation injuries with prolonged bedrest in traction is recommendedif more contemporary treatment options are not available.

SUBAXIAL CERVICAL INJURIES EXCLUDING FACET DISLOCATION INJURIES:Standards: There is insufficient evidence to recommend treatment standards.Guidelines: There is insufficient evidence to recommend treatment guidelines.Options:• Closed or open reduction of subluxations or displaced subaxial cervical spinal fractures is recommended.• Treatment of subaxial cervical spinal injuries with external immobilization, anterior arthrodesis with plate

fixation, or posterior arthrodesis with plate or rod fixation is recommended.

RATIONALE

Subaxial cervical vertebral fracture-dislocation injuriesare common after nonpenetrating cervical trauma andare often associated with neurological injury. Before the

advent of spinal instrumentation, many of these injuries weremanaged with traction, postural reduction, or external ortho-ses with frequent success. However, the morbidity and mor-tality associated with prolonged immobilization for 3 monthsor more prompted surgeons to investigate the usefulness ofinternal fixation in the management of these injuries. To de-velop treatment recommendations for closed subaxial cervicalspinal injuries, an analysis of the articles examining theirmanagement is undertaken in this chapter. In particular, thisfocused review assessed the usefulness of closed reductionwith or without external immobilization compared with arth-rodesis with or without internal fixation.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The search terms “cervical vertebrae,” “spinalfractures,” and “dislocations,” limited to English languagearticles, yielded 15,124, 3,010, and 17,811 citations, respec-tively. “Cervical vertebrae” combined with “spinal fractures”yielded a subset of 688 citations, and “cervical vertebrae”combined with “dislocations” yielded a subset of 1,159 cita-

tions. An exploded search of “therapeutics” or “treatment,”limited to English language articles, yielded 1,566,596 cita-tions. This search term was combined with each of the twoearlier subsets, resulting in 198 and 287 citations, respectively.The abstracts were reviewed, and only those containing 10 ormore cases of subaxial cervical injury after nonpenetratingcervical trauma were included. An exception was made forankylosing spondylitis because so few reports included morethan 10 patients with this disorder. Sixty-three articles met theselection criteria and provide the basis for this review. Sixtyarticles providing Class III medical evidence are summarizedin Tables 20.1 to 20.6.


The variety and heterogeneity of subaxial cervical spinalinjuries require accurate characterization of the mechanicsand type of injury to enable a comparison of the efficacy ofoperative and nonoperative treatment strategies. The absenceof a uniformly accepted classification scheme for cervicalvertebral injuries limits the ability to compare the effects oftreatment reported in many clinical studies. In 26 articlesdescribing series of patients with cervical injuries, sub-axial cervical injuries are not differentiated. The Allen andFerguson classification system (3) has been the most com-monly used scheme to differentiate and characterize subaxialvertebral injuries. Although few authors reported injuries by


subtype, many of the reports described cervical injuries thatcould be grouped into the following broad categories as de-scribed by Allen et al.: distractive flexion, compressiveflexion/vertical compression, extension, and subluxation. Theeffectiveness of treatment of subaxial cervical spinal facetdislocation injuries, a subset of distractive flexion injuries,may be different from that of other subaxial cervical spinalfracture-dislocation injuries and is considered separately (53).Finally, four articles described unique characteristics of themanagement of subaxial cervical injuries in patients withankylosing spondylitis and are included in this review.

Several general principles can be formulated from the re-view of 26 articles that describe the treatment of subaxialcervical injuries without accurate differentiation into specificinjury types. Although closed reduction was successful in 64to 91% of patients with traumatic subaxial cervical malalign-ment (5, 28, 34a, 35, 43), patients with delayed treatment ofinjuries had a higher failure rate of closed reduction (22.5%)compared with those treated early (5). Halo vest applicationwas used successfully to immobilize patients with subaxialcervical injuries on arrival at the hospital to facilitate transportand workup; none had neurological worsening (32). Orthosesfailed to maintain reduction of subaxial cervical fracture-dislocation injuries in 7 to 56% of patients (10, 16, 18, 21–23,24a, 27, 35, 46, 53). Overall, 30% of these injuries (222 of 752injuries) had recurrent displacement or inadequate alignmentduring external immobilization (503 halo vest, 249 traction).Six of these patients were reported to heal with good ultimatealignment after readjustment of the halo device (three pa-tients) or continued postural reduction (three patients) (27,46). Nineteen percent (140 of 752 patients) were maintained inexternal immobilization despite displaced injuries and healedin an unreduced, nonanatomic position (10, 16, 18, 21, 23, 24a,27, 35, 46, 53). Eleven percent (82 of 752 patients) underwentsubsequent surgical treatment, typically for correcting cervi-cal malalignment (10, 16, 18, 22, 23, 24a, 27, 35, 46, 53). Severalrisk factors were identified in association with failure of non-operative management of subaxial cervical injuries. Patientswith more than 40% compression of a cervical vertebra, morethan 15 degrees of kyphotic angulation, or more than 20%subluxation of one vertebra on another were more likely toexperience failure of treatment with external immobilization(craniocervical traction alone or traction and then externalorthosis) (35).

In contrast, failure to maintain anatomic reduction of sub-axial cervical fracture-dislocation injuries after operativetreatment ranged from 1 to 18% (7, 10, 23, 28, 36, 46, 48, 51, 56).Anterior cervical fusion procedures (28, 48, 56, 59) were asso-ciated with less frequent failure to maintain reduction (10[5%] of 213 patients) when compared with posterior cervicalfusion procedures (38 [14%] of 280 patients) among all pa-tients with subaxial cervical injuries treated operatively (7, 36,51). Overall, 9% (61 of 704 patients) had recurrent angulationor subluxation despite surgical management (7, 10, 23, 28, 33,

35, 36, 46, 48, 51, 56, 59). A second operation in treatment ofprogressive deformity was rare in these patients. Successfularthrodesis occurred in nearly every patient reported (7, 15,33, 36, 43, 46, 48, 56). Surgical complications were relativelycommon in these series, ranging from 9 to 25% (7, 20, 27, 33,46, 48, 56, 59). In particular, graft extrusion after anteriorcervical surgery without plate fixation was observed in asmany as 10% of patients managed in this way (20, 27). Overall4 (4%) of 104 patients experienced graft displacement (20, 23,27, 35, 52, 59) after anterior fusion without plate fixation,compared with none of 291 patients treated with anteriorfusion with plate fixation (5, 28, 33, 48, 56). Complicationshave been reported using posterior plate or rod fixation aswell (7, 36, 51); radiculopathy occurred in 25% of patients inone report (36) describing these techniques.

Subaxial cervical facet dislocation injuries

Twenty-eight articles provided sufficient information toevaluate patients with subaxial cervical distractive flexioninjuries. Most reports were retrospective series of patientswith subaxial cervical spinal facet dislocation injuries (unilat-eral, bilateral, or both) (8, 11–13, 16, 19, 27, 30, 37, 39, 41, 45, 47,49, 50, 53, 54, 55, 57, 61). Overall, 26% (181 of 701 patients)with cervical spinal facet dislocation injuries had failure toachieve closed reduction with craniocervical traction (8, 11,12, 17, 19, 30, 39, 41, 42, 45, 47, 50, 53–55, 57, 61). The issue ofclosed reduction of cervical fracture dislocation injuries, in-cluding the potential of an associated ventral disc herniation,is described in the chapter “Initial Closed Reduction of Cer-vical Spine Fracture-Dislocation Injuries” (Chapter 6). Reduc-tion, when accomplished, could not be maintained in 28% ofpatients (112 of 393 patients) treated with subsequent externalimmobilization (8, 11–13, 16, 19, 27, 30, 37, 39, 41, 45, 47, 49, 50,53–55, 57, 61). Mortality associated with closed treatment offacet dislocation injuries was 7% (28 of 392 patients) in seriesreporting this complication (8, 11–13, 19, 30, 37, 39, 41, 45, 50,53, 57, 61). Prolonged bedrest and cervical traction alone for 12to 16 weeks’ duration was associated with the highest mor-tality of all treatment strategies reported for these injuries,27% in one series of 41 patients managed in this way (13).

Vertebral subluxation, facet injury (ligamentous or frac-ture), and a locked/perched facet on the initial x-rays orsubsequent computed tomography or magnetic resonance im-aging studies have been cited as factors associated with fail-ure of nonoperative treatment (8, 11, 30, 31, 57). Facet frac-tures associated with cervical spinal facet dislocation injuriesmay preclude successful closed reduction (30, 57). They havealso been associated with a high rate of arthrodesis withexternal immobilization alone (halo device) if closed reduc-tion can be accomplished, 97% in one report on this issue (30).Ligamentous disruption without facet fracture is associatedwith an increased likelihood of failure of external immobili-zation (halo device, Minerva cast) in the treatment of these

Subaxial Cervical Spinal Injuries


TABLE 20.1. Summary of Reports on Subaxial Cervical Spinal Injuries, Mixed Typesa


ClassConclusions

Kalff et al., 1993 (33) Retrospective study. 97 cervical injuries (79 anterior and 18 anterior-posterior

fusion/instrumentation procedures). 16 DF, 14 VC, 64 fracture-dislocation

injuries.

III 9% operative complications related to fixation devices, but less than half

required reoperation. All patients achieved fusion.

Lemons and Wagner,

1993 (35)

Retrospective study. 64 cervical fractures: 14 VC, 12 CF, 12 UFD, 16 BFD,

10 extension. Treatment: 38 halo, 38 fusion (12 failed halo, 26 primary), 26

posterior fusion/instrumentation, 4 anterior fusions, 4 anterior-posterior.

III 12/38 halo-treated injuries were unstable and were fused. 4 healed

malaligned. None with extension injuries were unstable. 5/38 treated with

fusion were unstable. Risk for orthosis failure: �40% compression, �15

degree angulation, �20% subluxation.

Cybulski et al., 1992

(20)

Retrospective study. 21 cervical injuries failing posterior wiring treated with

anterior fusion.

III 2/21 had graft extrusions.

Della Torre and

Rinonapoli, 1992 (21)

Retrospective study. 28 cervical injuries.

3 CE, 7 CF, 4 DF treated with halo.

III 4/7 CF injuries were not reducible.

All were stable with immobilization.

Heary et al., 1992 (32) Retrospective study. 78 cervical injuries.

Halo for transport. 49 subaxial fractures, 45 subaxial subluxations.

III No patient worsened neurologically in halo before receiving definitive

treatment.

Levine et al., 1992 (36) Retrospective study. 24 facet fractures.

Posterior fusion with instrumentation.

III 11 complications including 4 who lost correction and 6 with radiculopathy.

All achieved fusion.

Roy-Camille et al.,

1992 (51)


89% posterior fusion, 11% anterior fusion.

III 15% developed kyphosis after surgery.

Nazarian and Louis,

1991 (43)


Posterior fusion with plates. 11 UFD, 4 BFD, 3 subluxations, 5 facet

fractures.

III 3/12 failed closed reduction, and 3 were unstable in an orthosis. All achieved

fusion.

Ripa et al., 1991 (48) Retrospective study. 92 cervical injuries. Anterior fusion instrumentation with

plate. 48 multiple fractures, 20 VC, 13 DF, 6 extension.

III No patient worsened neurologically. 12/15 complications were hardware-

related. 1 patient had pseudarthrosis.

Sears and Fazl, 1990

(53)

Retrospective study. 173 cervical injuries. 103 non-facet-dislocation injuries.

Halo treatment.

Operative procedure unreported.

III Nonoperative treatment failed in 31/103 patients (3 were irreducible, 10 were

neurologically worse, 16 subluxed in halo, 2 had late instability). Subluxation

and angulation predicted failed treatment, whereas fracture did not.

Benzel and Kesterson,

1989 (7)

Retrospective study. 50 cervical injuries. 25 fracture-subluxation. Posterior

fusion/instrumentation with wire.

III 1 patient with complete injury of 25 patients died.

Remainder healed.

Goffin et al., 1989 (28) Retrospective study. 41 cervical injuries. Anterior fusion with plate. III 2/41 subluxed, requiring surgery. 3/12 dislocations were irreducible. All 4

deaths were in quadriplegics.

Shoung and Lee, 1989

(56)

Retrospective study. 37 cervical injuries. Anterior fusion with plate. III All 37 healed. No graft extrusion. 1 death, 1 infection, 2 screw loosening.

Argenson et al., 1988

(5)

Retrospective study. 47 cervical injuries. 7 posterior fusion, 40 anterior

fusion.

III 17/22 were reducible, but 5 old dislocations were irreducible. 1 died of

vertebrobasilar thrombosis.

Bucci et al., 1988 (10) Retrospective study. 49 cervical injuries. Treatment: 20 halo alone (1

refused), 28 fusion with immobilization, procedure unreported.

III 12/20 with halo stable.

26/28 fused were stable (P � 0.01).

2 in each group lost reduction. 1 in each group were neurologically worse.

Donovan et al., 1987

(23)

Retrospective study. 61 cervical injuries. Treatment: 17 fusion with

immobilization (4 anterior, 13 posterior), 43 6-wk treatment in halo, 1

laminectomy without fusion.

III 18/43 had alignment in halo.

3/9 DF unstable: 2 surgery/1 asymptomatic.

All patients treated with fusion were stable, but 3 developed kyphosis.

Savini et al., 1987 (52) Retrospective study. 12 cervical injuries with late instability after closed

treatment.

III No grafts dislodged when anterior fusion was performed before posterior

reduction.

Ersmark and Kalen,

1986 (24a)

Retrospective study. 64 cervical injuries. Treatment with halo vest (36

subaxial).

III 29 dislocations and 5 VC injuries were stable after halo vest treatment.

Glaser et al., 1986 (27) Retrospective study. 245 cervical injuries, 125 complex fractures. Halo

treatment. Fusion posteriorly with wire or anteriorly without plate.

III 17/86 lost alignment in the halo. 2 interbody grafts displaced after anterior

surgery without plate.

De Smet et al., 1984

(22)

Retrospective study. 28 cervical injuries. Traction. III Early reduction failed in 4/28 patients. 2/24 had late instability.

Cahill et al., 1983 (15) Retrospective study of 25 DF or CF injuries. Treated with posterior fusion

with wiring.

III 18/18 with 3-mo follow-up were stable, and none had complications.

Chan et al., 1983 (16) Retrospective study. 188 cervical injuries. 150 subaxial with follow-up. Halo

treatment.

III 4/55 fracture-dislocation or complex fractures, 13/53 had UFD/BFD, and 2/41

VC were unstable with halo treatment.

Cooper et al., 1979 (18) Retrospective study. 33 cervical injuries treated with halo. III 2/11 “complex” fractures. 1/3 subluxations unstable with halo.

Verbiest, 1969 (59) Retrospective study. 47 cervical injuries. Anterior fusion without plate. III 5 patients died, 4 with complete spinal cord injuries. 6 had residual

malalignment, and 1 other had reoperation for lost alignment.

Paeslack et al., 1973

(46)

Retrospective study. 221 cervical injuries. 68 CF, 114 DF treated with

postural reduction, traction. 31 cervical injuries treated with anterior or

posterior fusion.

III 75 aligned, 67 wedged, 43 partially reduced, 36 failed reduction.

4/221 had late instability, 3 stable with further treatment and 1 with surgery.

2/31 were unstable after surgery.

Koskinen and

Nieminen, 1967 (34a)

Retrospective study.

159 cervical injuries.

Various treatments.

III No difference in pain, neck mobility, radiculopathy, or mortality when

operative and nonoperative treatments were compared.

a DF, distractive flexion; VC, vertical compression; CF, compressive flexion; UFD, unilateral facet dislocation; BFD, bilateral facet dislocation;CE, compressive extension.



TABLE 20.2. Summary of Reports on Subaxial Cervical Spinal Injuries: Distractive Flexiona


ClassConclusions

Ordonez et al., 2000 (44) Retrospective study. 6 UFD, 4 BFD. 9 anterior reduction and fusion,

1 anterior-posterior-anterior.

III 10 with stable fusions, although 1 was incompletely reviewed.

Shapiro et al., 1999 (55) Retrospective study. 51 UFD. 24 SP posterior wiring, 22 SP wiring

and plates. 5 anterior-posterior-anterior.

III 1/24 with wire failed and 13/24 had late kyphosis. All patients with plate

fixation had stable fusions.

Fehlings et al., 1994 (25) Retrospective study. 44 cervical injuries. 19 facet dislocations.

Posterior fusion with plate.

III 5/19 patients had complications, including 2 late failed reductions.

Lieberman and Webb,

1994 (37)

Retrospective study. 41 cervical injuries. 9 facet dislocations.

Patients �65 yr old.

III 5 patients died, 1 treated with traction and 4 with halo. 3/4 survivors treated

with traction healed. All 4 survivors with halo treatment healed.

Lukhele, 1994 (40) Retrospective study. 43 facet dislocations. 12 with laminar fractures.

Posterior fusion with wire.

III 5/12 patients developed kyphosis.

Pasciak and Doniec, 1993

(47)

Retrospective study. 32 UFD. 23 nonoperative (treatment with halo

or plaster vest). 9 operative.

III All 9 treated with surgery healed. 8/23 in whom closed reduction failed were

fused. 7/15 with failure to maintain reduction achieved fusion.

Shapiro, 1993 (54) Retrospective study. 24 UFD. Posterior fusion with wire. III 23/24 patients with surgery healed. 1 with resubluxation healed with ACF.

9/17 in whom closed reduction failed had fractures of laminae.

Hadley et al., 1992 (30) Retrospective study. 31 UFD, 37 BFD. III 18/29 UFD and 20/37 BFD had successful closed reduction. 16 UFD and 15

BFD healed in halo. 7/31 had failed halo treatment. (5/7 without associated

facet fractures). When facet fractures were present, once reduced, 97%

success rate in halo. UFD/BFD results were similar.

Mahale and Silver, 1992

(41)

Retrospective study. 13 missed BFD with neurologically worse and

late treatment.

III All 13 reduced (10 completely). 12/13 healed with traction, 1 needed surgery.

Beyer et al., 1991 (8) Retrospective study. 36 UFD with and without fractures. 24 treated

with halo or orthosis. 10 posterior ORIF.

III 15/24 reduced in halo. 8/10 reduced with surgery. 11/24 had failed treatment

in halo. All 10 healed with surgery. Pain was more frequent despite healing if

unreduced.

Wolf et al., 1991 (61) Retrospective study. 52 BFD. 44 posterior fusion with wire, 3

anterior fusion and plate, 2 both.

III 12/52 had failed closed reduction. All 3 deaths had complete quadriplegia.

Cotler et al., 1990 (19) Retrospective study. 23 UFD (10 nonoperative), 12 BFD (4

nonoperative). 30 fused (21 primary).

III 1/2 had failed halo. 8/12 had failed traction. Complications were not reported.

Rockswold et al., 1990

(49)

Retrospective study. 140 cervical injuries. 8 facet dislocations (6

UFD, 2 BFD). Treated with halo or surgery.

III 1/6 UFD had failed halo. 0/4 had failed surgery. 1/2 BFD had failed halo. 2/9

had failed surgery.

Sears and Fazl, 1990 (53) Retrospective study. 173 cervical injuries. 70 dislocation injuries (38

UFD, 32 BFD).

III 19 healed with halo, 16 in good alignment. 16 had failed reduction and 23

subluxed in halo required surgery. Subluxation and angulation were not

associated with failure of halo. UFD/BFD results similar.

Benzel and Kesterson,

1989 (7)

Retrospective study. 50 cervical injuries. 19 UFD, 6 BFD. Posterior

fusion with wiring.

III All healed with fusion. 1 BFD neurologically worse required ACF, 1 UFD

incomplete patient died.

Bucholz and Cheung,

1989 (11)

Retrospective study. 124 cervical injuries. 20 distractive flexion

injuries. Treated with halo or surgery.

III 9/20 had failed halo. 1 neurologically worse postoperatively, unreported if

distractive flexion or subluxation patient.

Osti et al., 1989 (45) Retrospective study. 167 dislocations. 82 nonoperative (traction), 85

operative (anterior fusion without plate).

III 6/82 who had failed reduction were fused. 14/76 with late instability were

fused. 7 operatively treated within 24 h died (all ASIA A).

Lind et al., 1988 (39) Retrospective study. 83 injuries. Treated with halo. III 4/31 had failed halo. Loose pins common.

Rorabeck et al., 1987 (50) Retrospective study. 26 UFD. III 20/26 had failed closed reduction. 10 healed in halo. 8/10 remained reduced

with surgery. Pain common with failed reduction.

Glaser et al., 1986 (27) Retrospective study. 245 cervical injuries. 17 dislocations. III 3/12 UFD had failed halo. 1/5 BFD had failed halo.

Maiman et al., 1986 (42) Retrospective study. 26 BFD. III 10/18 reduced with closed reduction. 3 died, all complete injuries.

Chan et al., 1983 (16) Retrospective study. 188 cervical injuries. 150 subaxial with follow-

up. 40 halo alone, 20 halo and posterior fusion.

III 27/40 healed with halo. All 20 with primary surgery healed.

Dorr et al., 1982 (24) Retrospective study. 117 cervical injuries. 25 flexion-rotation

injuries.

III 2/3 with ACF had complications (1 graft displaced, 1 kyphosis).

Sonntag, 1981 (57) Retrospective study. 15 BFD. Halo or surgery. III 10/15 reduced with closed reduction. 4 halo healed (2 no follow-up), 8

posterior fusion with wire healed, 1 died (complete).

Stauffer and Kelly, 1977

(58)

Retrospective review. 10 dislocations. 5 fractures. 1 fracture

subluxation. Anterior fusion.

III 16/16 had recurrent angular deformity after ACF without plate. 3/16 fused

angulated.

Burke and Tiong, 1975

(13)

Retrospective review. 175 cervical injuries. Treated with traction,

traction-manipulation, or collar.

III 2/14 UFD and 0/13 BFD had failed nonoperative treatment.

Burke and Berryman, 1971

(12)

Retrospective review. 76 facet dislocations. 41 UFD, 35 BFD. 41

manipulation. 35 traction (3/35 failed manipulation). 3 fusion

primarily.

III 4/41 failed manipulation and 4 of remaining 37 had late instability. 0/32 had

failed traction.

Cheshire, 1969 (17) Retrospective review. 257 cervical injuries. Treated with traction or

collar (33 excluded).

III 3/40 UFD and 2/35 BFD had failed nonoperative treatment.

a UFD, unilateral facet dislocation; BFD, bilateral facet dislocation; SP, spinal process; ACF, anterior cervical fusion; ORIF, open reduction withinterior fixation; ASIA, American Spinal Injury Association scale.

Subaxial Cervical Spinal Injuries S159


injuries (30, 31). Laminar fractures have been associated withan increased risk of late kyphosis after surgical treatment ofcervical spinal facet dislocation injuries (40). Although pa-tients with unreduced facet dislocations treated with externalimmobilization often achieve spinal stability once treatmenthas been completed, arthrodesis in a position of malalignmenthas been associated with persistent cervical pain (8, 50, 55).No differences were observed in the success of achievingclosed reduction and/or maintaining cervical spinal align-

ment in patients with unilateral facet dislocations comparedwith patients with bilateral facet dislocation injuries.

In contrast, open reduction was achieved in all but 1 of 24patients treated with anterior fusion procedures (42, 44, 55)and in all but 7 of 167 patients treated with posterior fusionprocedures in series that reported this finding (7, 8, 11, 19, 30,42, 47, 50, 54, 57, 61). Delayed instability occurred in 6 (6%) of101 patients treated with anterior fusion procedures (12, 13,24, 40, 42, 55, 61), and 6 (3%) of 237 patients treated with

TABLE 20.3. Summary of Reports on Subaxial Cervical Spinal Injuries: Compressive Flexion or Vertical Compressiona


ClassConclusions

Fehlings et al., 1994 (25) Retrospective study. 44 cervicalinjuries. Posterior fusion with plates.

III Complications in 6/17 including 1residual kyphosis and 1 new kyphosis(reoperation).

Lieberman and Webb, 1994 (37) Retrospective study. 41 cervicalinjuries. Patients �65 yr old.

III 1/4 died. 2 in collar and 1 fusedpatient were stable.

Kiwerski, 1993 (34) Retrospective “crossover” study. 273VC. 1st 70 traction. 2nd 203 anteriorfusion.

III Fewer died and more recoveredneurological function when treatedwith surgery.

Aebi et al., 1991 (1) Retrospective study. 22 cervicalinjuries. Anterior corpectomy withplate.

III All 22 achieved stable fusion. 2 screwcomplications occurred.

Anderson et al., (4) Prospective study. 30 cervicalinjuries. Posterior fusion with plate.

III All 9 achieved stable fusion, although1 had late kyphosis.

Bucholz and Cheung, 1989 (11) Retrospective study. 32 cervicalinjuries. 19 VC and CF injuries.

III 1/19 had failed halo treatment. Patienthad failed posterior fusion with wire.

Cabanela and Ebersold, 1988 (14) Retrospective study. 8 teardropfracture. Anterior fusion with plate.

III All 8 achieved stable fusion with nonedeveloping kyphosis.

Lind et al., 1988 (39) Retrospective study. 83 cervicalinjuries. Halo treatment.

III 2/19 were unstable. Drainage andloose pins were common.

Chan et al., 1983 (16) Retrospective study. 188 cervicalinjuries. 150 subaxial with follow-up.Halo treatment.

III All 22 burst fractures and 17 teardropfractures achieved stable fusion.

Dorr et al., 1982 (24) Retrospective study. 117 cervicalinjuries. 32 VC injuries.

III 1/11 had graft displacement afteranterior fusion without plate.

Burke and Tiong, 1975 (13) Retrospective study. 175 cervicalinjuries. Treated with traction,traction-manipulation, collar.

III 1/46 had failed nonoperativetreatment.

Frankel et al., 1973 (26) Retrospective study. 218 cervicalinjuries. 45 burst, 97 teardrop. Closedtreatment.

III 7/142 had failed postural reduction.103 had residual deformities.

Cheshire, 1969 (17) Retrospective review. 257 cervicalinjuries. Treated with traction orcollar (33 excluded).

III 3/63 had failed nonoperativetreatment.

Beatson, 1963 (6) Retrospective study. 59 cervicalinjuries. All immobilized.

III All 16 were stable withimmobilization.

a VC, vertical compression; CF, compressive flexion.



posterior fusion procedures (7, 8, 11, 25, 30, 37, 40, 42, 49, 50,55, 57, 61). None of these six patients who failed to achievestability after anterior fusion was treated with plate fixation inaddition to fusion. Seven (8%) of 85 patients treated withanterior fusion procedures developed kyphosis; none hadbeen treated with anterior plate fixation (7, 8, 11, 25, 30, 37, 40,42, 49, 50, 55, 57, 61). Sixteen patients described by Staufferand Kelly (58) also developed kyphotic angulation after ante-rior cervical fusion without internal fixation. In contrast, 22(13%) of 165 patients developed kyphosis after posterior cer-vical fusion with wiring (25, 40, 54, 55), whereas only 1 (3%)of 40 patients developed kyphosis after posterior fusion withlateral mass plate fixation (25, 55). Alternatively, Halifax in-terlaminar clamps were successfully used in five patients withfacet dislocations treated with posterior arthrodesis (2).

Graft displacement was the most common complicationafter attempted anterior arthrodesis without internal fixation(7 [8%] of 85 patients) (6, 13, 24, 40, 42, 44, 55). Seven percent(8 of 113 patients) died after anterior fusion procedures (13,24, 40, 42, 44, 45, 55, 61); 3% (7 of 268 patients) died afterposterior fusion procedures (7, 11, 19, 25, 30, 42, 47, 49, 50, 55,57, 61). In these reports, all but 1 of the 15 patients who diedafter surgery that was performed in an attempt to correctdeformity and stabilize the spine had complete cervical spinalcord injuries (7, 42, 57, 61).

Subaxial cervical spinal injuries excluding facetdislocation injuries

Fourteen articles provided sufficient information to evalu-ate patients with subaxial cervical spinal compression fracture

injuries (1, 4, 6, 11, 13, 14, 16, 17, 24–26, 34, 37, 39). Althoughsome authors differentiated compressive flexion injuries fromvertical compression injuries, others considered these injuriestogether. Many nonoperative treatment strategies were de-scribed, including traction and external immobilization incollar, plaster jacket, or halo vest. Overall, 5% (17 of 349patients) treated with immobilization for compressive injuriesof the subaxial cervical spine had persistent instability afternonoperative treatment used for 8 to 12 weeks (6, 11, 13, 17,26, 37, 39). In contrast, nearly every patient with these injuriestreated with anterior (22 of 22 patients) or posterior (26 of 27patients) fusion procedures developed a stable union (1, 4, 25,27). Subluxation or kyphosis developed in 2 of 18 patientswho were treated with posterior fusion (11, 25). Operativecomplications were more common in patients treated withposterior fusion procedures (10 [37%] of 27 patients) thananterior fusion procedures (3 [9%] of 33 patients) (1, 4, 24, 25).Graft displacement was the most common complication de-scribed in patients treated with anterior cervical fusion with-out internal fixation (3 [9%] of 33 patients) (1, 24).

Only seven articles reported sufficient data to analyze pa-tients treated for extension injuries of the subaxial cervicalspine (4, 11, 13, 24, 37, 38, 43). Twenty-four percent (19 of 79patients) failed treatment with external immobilization (11,13, 37, 38, 43). In contrast, none of 19 patients failed treatmentwith anterior cervical fusion (37, 38). Two patients had irre-ducible vertebral displacements, and three patients developedkyphotic deformities among 11 patients with cervical spinalextension injuries treated with attempted posterior cervicalfusion (38).

TABLE 20.4. Summary of Reports on Subaxial Cervical Spinal Injuries: Extensiona


ClassConclusions

Lifeso and Colucci, 2000 (38) Retrospective and prospective study.32 CE1 injuries (3 lost to follow-up).

III All 18 treated with brace failed (17were unreduced). 9/11 healed withPCF, but 3 had stable kyphosis.2/11 healed without reduction.

Lieberman and Webb, 1994 (37) Retrospective study. 41 cervicalinjuries. Patients �65 yr old.

III 1/3 healed with collar, and 1/3healed with surgery.

Anderson et al., 1991 (4) Prospective study. 30 cervical injuries.Posterior fusion with plates.

III All 30 healed, but 1 had screwloosening.

Rockswold et al., 1990 (49) Retrospective study. 140 cervicalinjuries. Treated with halo or surgery.

III All 3 treated with halo healed. All3 treated with surgery healed.

Bucholz and Cheung, 1989 (11) Retrospective study. 32 cervicalinjuries, 12 extension injuries.

III 1/12 failed halo treatment. 1patient stable after posterior fusionwith wire.

Dorr et al., 1982 (24) Retrospective study. 117 cervicalinjuries, 45 extension injuries.

III 40/45 were treated with brace. Ofall cervical injuries treated withbrace, 5/86 failed.

Burke and Tiong, 1975 (13) Retrospective review. 175 cervicalinjuries. Treated with traction, traction-manipulation, collar (30 excluded).

III All 45 healed without surgery.

a CE, compressive extension; PCF, posterior cervical fusion.



Eight articles reported sufficient data to analyze patientstreated for vertebral subluxation injuries of the cervical spine(4, 6, 11, 13, 17, 18, 25, 49). Sixty-four percent of patients withthese injuries had successful treatment with external immo-bilization; patients with more than 50% subluxation were

twice as likely to maintain anatomic cervical realignment afterclosed reduction (72% versus 36%) (6). Thirty-six percent ofpatients (39 of 108 patients) had failure of external immobili-zation after closed reduction (11, 13, 17, 18, 49), comparedwith 7% of patients in whom these injuries were managed

TABLE 20.5. Summary of Reports on Subluxation


ClassConclusions

Fehlings et al., 1994 (25) Retrospective study. 44 cervicalinjuries. Posterior fusion with plates.

III 2/5 lost reduction, including 1 whodied.

Anderson et al., 1991 (4) Prospective study. 30 cervical injuries.Posterior fusion with plates.

III 19/19 healed with fusion. 8/19 hadcomplications, including 2 withincreased kyphosis and 3 requiringadditional levels to be fused.

Rockswold et al., 1990 (49) Retrospective study. 140 cervicalinjuries. Treated with halo or surgery.

III 12/26 had failed halo treatment. 2/10had failed surgical treatment.

Bucholz and Cheung, 1989 (11) Retrospective study. 32 cervicalinjuries, 6 subluxation injuries.

III 2/6 had failed halo treatment. 1 worsepostoperatively, unreported if distractiveflexion or subluxation patient.

Cooper et al., 1979 (18) Retrospective study. 33 cervicalinjuries. Treated with halo.

III 1/3 had failed halo treatment.

Burke and Tiong, 1975 (13) Retrospective study. 175 cervicalinjuries. Treated with traction,traction-manipulation, collar (30excluded).

III 1/14 had failed nonoperative treatment.

Cheshire, 1969 (17) Retrospective study. 257 cervicalinjuries. Treated with traction or collar(33 excluded).

III 4/19 had failed nonoperative treatment.

Beatson, 1963 (6) Retrospective study. 59 cervicalinjuries (3 excluded).

III 8/22 with �50% subluxation reduced.2/14 remaining had surgery. 13/18 with�50% subluxation reduced. 5/5remaining had surgery.

TABLE 20.6. Summary of Reports on Ankylosing Spondylitisa


ClassConclusions

Weinstein et al., 1982 (60) Retrospective study. 13 AS. 7traumatic cervical, 6 quadriplegic.2 central cords without fracture.

III 2 treated with traction died ofpneumonia. 2 treated with traction/bracehealed. 1 worse with halo treatedsurgically. 1 laminectomy/fusion worse, 1laminectomy/fusion had pseudarthrosis.

Bohlman, 1979 (9) Retrospective study. 300 cervicalinjuries. 8 AS.

III 5/8 patients died. 2 healed after bracetreatment and 1 after laminectomy.

Cheshire, 1969 (17) Retrospective study. 257 cervicalinjuries. 1 AS.

III 1 C5–C6 extension injury healed withsurgical fusion.

Grisolia et al., 1967 (29) Retrospective study. 6 AS. III 3/4 healed with brace with or withouttraction. 2 with laminectomy and PCFdied of PE.

a AS, ankylosing spondylititis; PCF, posterior cervical fusion; PE, pulmonary embolism.



surgically (4, 49). A kyphotic deformity developed in 4% ofreported patients (3 of 74 patients) treated with posteriorcervical fusion and lateral mass plate or rod fixation proce-dures (4, 25).

Several characteristics of subluxation injuries of the sub-axial cervical spine were associated with failure of nonopera-tive treatment (6, 53). Patients with subluxation or kyphoticangulation frequently failed to achieve a good anatomic resultafter treatment with halo vest immobilization (46 [45%] of 103patients). Combined fractures to all parts of the cervical spinalcolumn and the presence of facet fractures were not associatedwith a higher likelihood of failure of treatment with externalimmobilization (53). Closed reduction was more successfulwith a subluxation of more than 50% of the vertebral bodydiameter (6).

Comparatively few studies examined the specific difficul-ties associated with the management of patients with anky-losing spondylitis who sustained cervical spinal injuries (9, 17,29, 60). In four articles reporting patients with this entity andinjury, 9 of 22 patients died. Four patients managed nonop-eratively died. Two of nine survivors treated with externalimmobilization experienced failure of treatment. One wors-ened neurologically when placed in a halo and subsequentlywas successfully treated with laminectomy and fusion. Theother patient had persistent cervical subaxial spinal instabilitybut refused further therapy. In contrast, five of nine ankylos-ing spondylitis patients with cervical fracture injuries treatedprimarily with surgery died. One patient was neurologicallyworse after surgery. Three patients healed successfully with-out instability.

SUMMARY

In conclusion, closed reduction is successful for most sub-axial cervical spinal fracture-dislocation injuries. Failure ofclosed reduction is more common with facet dislocation inju-ries. Similarly, treatment with external immobilization is fre-quently successful in the management of most subaxial cer-vical spinal injuries, although failure to maintain reduction ismore frequent with facet dislocation injuries as well. Virtuallyall forms of external immobilization have been used in thetreatment of subaxial cervical spinal injuries. More rigid or-thoses (halo, Minerva) seem to have better success rates thanless rigid orthoses (collars, traction only) for fracture-dislocation injuries after reduction has been accomplished.Treatment with traction and prolonged bedrest has been as-sociated with increased morbidity and mortality.

Both anterior and posterior cervical fusion procedures aresuccessful in achieving spinal stability for most patients withsubaxial cervical spinal injuries. Indications for surgical treat-ment offered in the literature include failure to achieve ana-tomic injury reduction (irreducible injury), persistent instabil-ity with failure to maintain reduction, ligamentous injurywith facet instability, spinal kyphotic deformity more than 15degrees, vertebral body fracture compression of 40% or more,vertebral subluxation of 20% or more, and irreducible spinalcord compression. Anterior fusion without plate fixation isassociated with an increased likelihood of graft displacement

and the development of late kyphosis, particularly in patientswith distractive flexion injuries. Similarly, late displacementwith kyphotic angulation is more common in patients treatedfor facet dislocation injuries with posterior fusion and wiringcompared with those treated with posterior fusion and lateralmass plate or rod or interlaminar clamp fixation. Althoughpatients with persistent or recurrent cervical spinal malalign-ment often achieve spinal stability with either external immo-bilization or surgical fusion with or without internal fixation,a higher proportion of these patients have residual cervicalpain than similarly treated patients for whom anatomic spinalalignment is achieved and maintained.


To better compare the advantages and disadvantages ofnonoperative versus operative treatment strategies for sub-axial cervical injuries, future studies must differentiate be-tween the mechanisms of injury that have resulted in subaxialinjury. Although the Allen and Ferguson classification offers acommonly used framework for stratifying these patients,many investigators find that the number of subtypes in theirscheme precludes obtaining sufficient numbers of patientswith specific injuries. A broader classification of patients intocompressive flexion, distractive flexion, vertical compression,and extension injuries would allow comparison of most pa-tients who sustain subaxial cervical spinal injuries. A multi-center study would allow more rapid accumulation of pa-tients with these various categories of subaxial cervicalinjuries. A prospective examination of the efficacy of rigidexternal immobilization compared with surgical arthrodesiswith internal fixation (anterior and posterior approaches) mayfurther refine the most effective treatment for patients withsubaxial cervical spinal injuries.

Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue, Birmingham, AL 35294-3295.

REFERENCES

1. Aebi M, Zuber K, Marchesi D: Treatment of cervical spine inju-ries with anterior plating: Indications, techniques, and results.Spine 16[Suppl 3]:S38–S45, 1991.

2. Aldrich EF, Crow WN, Weber PB, Spagnolia TN: Use of MRimaging-compatible Halifax interlaminar clamps for posteriorcervical fusion. J Neurosurg 74:185–189, 1991.

3. Allen BL Jr, Ferguson RL, Lehmann TR, O’Brien RP: A mecha-nistic classification of closed, indirect fractures and dislocationsof the lower cervical spine. Spine 7:1–27, 1982.

4. Anderson PA, Henley MB, Grady MS, Montesano PX, Winn HR:Posterior cervical arthrodesis with AO reconstruction plates andbone graft. Spine 16[Suppl 3]:S72–S79, 1991.

5. Argenson C, Lovet J, Sanouiller JL, de Peretti F: Traumaticrotatory displacement of the lower cervical spine. Spine 13:767–773, 1988.

6. Beatson TR: Fractures and dislocation of the cervical spine.J Bone Joint Surg Br 45B:21–35, 1963.



7. Benzel EC, Kesterson L: Posterior cervical interspinous compres-sion wiring and fusion for mid to low cervical spinal injuries.J Neurosurg 70:893–899, 1989.

8. Beyer CA, Cabanela ME, Berquist TH: Unilateral facet disloca-tions and fracture-dislocations of the cervical spine. J Bone JointSurg Br 73B:977–981, 1991.


10. Bucci MN, Dauser RC, Maynard FA, Hoff JT: Management ofpost-traumatic cervical spine instability: Operative fusion versushalo vest immobilization—Analysis of 49 cases. J Trauma 28:1001–1006, 1988.

11. Bucholz RD, Cheung KC: Halo vest versus spinal fusion forcervical injury: Evidence from an outcome study. J Neurosurg70:884–892, 1989.

12. Burke DC, Berryman D: The place of closed manipulation in themanagement of flexion-rotation dislocations of the cervicalspine. J Bone Joint Surg Br 53B:165–182, 1971.

13. Burke DC, Tiong TS: Stability of the cervical spine after conser-vative treatment. Paraplegia 13:191–202, 1975.

14. Cabanela ME, Ebersold MJ: Anterior plate stabilization for burst-ing teardrop fractures of the cervical spine. Spine 13:888–891,1988.

15. Cahill DW, Bellegarrigue R, Ducker TB: Bilateral facet to spinousprocess fusion: A new technique for posterior spinal fusion aftertrauma. Neurosurgery 13:1–4, 1983.

16. Chan RC, Schweigel JF, Thompson GB: Halo-thoracic brace im-mobilization in 188 patients with acute cervical spine injuries.J Neurosurg 58:508–515, 1983.

17. Cheshire DJ: The stability of the cervical spine following theconservative treatment of fractures and fracture-dislocations.Paraplegia 7:193–203, 1969.

18. Cooper PR, Maravilla KR, Sklar FH, Moody SF, Clark WK: Haloimmobilization of cervical spine fractures: Indications and re-sults. J Neurosurg 50:603–610, 1979.

19. Cotler HB, Cotler JM, Alden ME, Sparks G, Biggs CA: Themedical and economic impact of closed cervical spine disloca-tions. Spine 15:448–452, 1990.

20. Cybulski GR, Douglas RA, Meyer PR Jr, Rovin RA: Complica-tions in three-column cervical spine injuries requiring anterior-posterior stabilization. Spine 17:253–256, 1992.

21. Della Torre P, Rinonapoli E: Halo-cast treatment of fractures anddislocations of the cervical spine. Int Orthop 16:227–231, 1992.

22. De Smet L, Vercauteren M, Verdonk R, Claessens H: Severeacute cervical spine injuries: Conservative treatment. ActaOrthop Belg 50:512–520, 1984.

23. Donovan WH, Kopaniky D, Stolzmann E, Carter RE: The neu-rological and skeletal outcome in patients with closed cervicalspinal cord injury. J Neurosurg 66:690–694, 1987.

24. Dorr LD, Harvey JP Jr, Nickel VL: Clinical review of the earlystability of spine injuries. Spine 7:545–550, 1982.

24a. Ersmark H, Kalen R: A consecutive series of 64 halo-vest-treatedcervical spine injuries. Arch Orthop Trauma Surg 105:243–246,1986.

25. Fehlings MG, Cooper PR, Errico TJ: Posterior plates in the man-agement of cervical instability: Long-term results in 44 patients.J Neurosurg 81:341–349, 1994.

26. Frankel H, Michaelis L, Paeslack V, Ungar G, Walsh JJ: Closedinjuries of the cervical spine and spinal cord: Results of conser-vative treatment of vertical compression injuries of the cervicalspine. Proc Veterans Adm Spinal Cord Inj Conf 19:28–32, 1973.

27. Glaser JA, Whitehill R, Stamp WG, Jane JA: Complications as-sociated with the halo-vest: A review of 245 cases. J Neurosurg65:762–769, 1986.

28. Goffin J, Plets C, Van den Bergh R: Anterior cervical fusion andosteosynthetic stabilization according to Caspar: A prospectivestudy of 41 patients with fractures and/or dislocations of thecervical spine. Neurosurgery 25:865–871, 1989.

29. Grisolia A, Bell RL, Peltier LF: Fractures and dislocations of thespine complicating ankylosing spondylitis: A report of six cases.J Bone Joint Surg Am 49A:339–344, 1967.


31. Halliday AL, Henderson BR, Hart BL, Benzel EC: The manage-ment of unilateral lateral mass/facet fractures of the subaxialcervical spine: The use of magnetic resonance imaging to predictinstability. Spine 22:2614–2621, 1997.

32. Heary RF, Hunt CD, Krieger AJ, Antonio C, Livingston DH:Acute stabilization of the cervical spine by halo/vest applicationfacilitates evaluation and treatment of multiple trauma patients.J Trauma 33:445–451, 1992.

33. Kalff R, Kocks W, Grote W, Schmit-Neuerburg KP: Operativespondylodesis in injuries of the lower cervical spine. NeurosurgRev 16:211–220, 1993.

34. Kiwerski JE: Early anterior decompression and fusion for crushfractures of cervical vertebrae. Int Orthop 17:166–168, 1993.

34a. Koskinen EV, Nieminen R: Fractures and dislocations of thecervical spine: Treatment and results of 159 cases. Int Surg47:472–485, 1967.

35. Lemons VR, Wagner FC Jr: Stabilization of subaxial cervicalspinal injuries. Surg Neurol 39:511–518, 1993.

36. Levine AM, Mazel C, Roy-Camille R: Management of fractureseparations of the articular mass using posterior cervical plating.Spine 17[Suppl 10]:S447–S454, 1992.

37. Lieberman IH, Webb JK: Cervical spine injuries in the elderly.J Bone Joint Surg Br 76B:877–881, 1994.

38. Lifeso RM, Colucci MA: Anterior fusion for rotationally unstablecervical spine fractures. Spine 25:2028–2034, 2000.

39. Lind B, Sihlbom H, Nordwall A: Halo-vest treatment of unstabletraumatic cervical spine injuries. Spine 13:425–432, 1988.

40. Lukhele M: Fractures of the vertebral lamina associated withunifacet and bifacet cervical spine dislocations. S Afr J Surg32:112–114, 1994.


42. Maiman DJ, Barolat G, Larson SJ: Management of bilaterallocked facets of the cervical spine. Neurosurgery 18:542–547,1986.

43. Nazarian SM, Louis RP: Posterior internal fixation with screwplates in traumatic lesions of the cervical spine. Spine 16[Suppl3]:S64–S71, 1991.

44. Ordonez BJ, Benzel EC, Naderi S, Weller SJ: Cervical facet dis-location: Techniques for ventral reduction and stabilization.J Neurosurg 92[Suppl 1]:18–23, 2000.

45. Osti OL, Fraser RD, Griffiths ER: Reduction and stabilisation ofcervical dislocations: An analysis of 167 cases. J Bone Joint SurgBr 71B:275–282, 1989.

46. Paeslack V, Frankel H, Michaelis L: Closed injuries of the cervi-cal spine and spinal cord: Results of conservative treatment offlexion fractures and flexion rotation fracture dislocation of thecervical spine with tetraplegia. Proc Veterans Adm Spinal CordInj Conf 19:39–42, 1973.



47. Pasciak M, Doniec J: Results of conservative treatment of unilat-eral cervical spine dislocations. Arch Orthop Trauma Surg 112:226–227, 1993.

48. Ripa DR, Kowall MG, Meyer PR Jr, Rusin JJ: Series of ninety-twotraumatic cervical spine injuries stabilized with anterior ASIFplate fusion technique. Spine 16[Suppl 3]:S46–S55, 1991.

49. Rockswold GL, Bergman TA, Ford SE: Halo immobilization and sur-gical fusion: Relative indications and effectiveness in the treatment of140 cervical spine injuries. J Trauma 30:893–898, 1990.

50. Rorabeck CH, Rock MG, Hawkins RJ, Bourne RB: Unilateralfacet dislocation of the cervical spine: An analysis of the resultsof treatment in 26 patients. Spine 12:23–27, 1987.

51. Roy-Camille R, Saillant G, Laville C, Benazet JP: Treatment oflower cervical spinal injuries: C3 to C7. Spine 17[Suppl 10]:S442–S446, 1992.

52. Savini R, Parisini P, Cervellati S: The surgical treatment of lateinstability of flexion-rotation injuries in the lower cervical spine.Spine 12:178–182, 1987.

53. Sears W, Fazl M: Prediction of stability of cervical spine fracturemanaged in the halo vest and indications for surgical interven-tion. J Neurosurg 72:426–432, 1990.

54. Shapiro SA: Management of unilateral locked facet of the cervi-cal spine. Neurosurgery 33:832–837, 1993.

55. Shapiro SA, Snyder W, Kaufman K, Abel T: Outcome of 51 cases ofunilateral locked cervical facets: Interspinous braided cable for lateralmass plate fusion compared with interspinous wire and facet wiringwith iliac crest. J Neurosurg 91[Suppl 1]:19–24, 1999.

56. Shoung HM, Lee LS: Anterior metal plate fixation in the treat-ment of unstable lower cervical spine injuries. Acta Neurochir(Wien) 98:55–59, 1989.

57. Sonntag VKH: Management of bilateral locked facets of thecervical spine. Neurosurgery 8:150–152, 1981.

58. Stauffer ES, Kelly EG: Fracture-dislocations of the cervicalspine: Instability and recurrent deformity following treatmentby anterior interbody fusion. J Bone Joint Surg Am 59A:45–48, 1977.

59. Verbiest H: Anterolateral operations for fractures and dislocations inthe middle and lower parts of the cervical spine: Report of a series offorty-seven cases. J Bone Joint Surg Am 51A:1489–1530, 1969.

60. Weinstein PR, Karpman RR, Gall EP, Pitt M: Spinal cord injury,spinal fracture, and spinal stenosis in ankylosing spondylitis.J Neurosurg 57:609–616, 1982.


Drawing by Leonardo da Vinci. Courtesy, Dr. Edwin Todd, Pasadena, California.


CHAPTER 21

Management of Acute Central Cervical Spinal Cord Injuries

RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS:• Intensive care unit (or other monitored setting) management of patients with acute central cervical spinal

cord injuries, particularly patients with severe neurological deficits, is recommended.• Medical management, including cardiac, hemodynamic, and respiratory monitoring, and maintenance of

mean arterial blood pressure at 85 to 90 mm Hg for the first week after injury to improve spinal cordperfusion is recommended.

• Early reduction of fracture-dislocation injuries is recommended.• Surgical decompression of the compressed spinal cord, particularly if the compression is focal and anterior,

is recommended.

RATIONALE

Central spinal cord injuries are among the most com-mon, well-recognized spinal cord injury patterns iden-tified in neurologically injured patients after acute

trauma. Originally described by Schneider et al. (19) in 1954,this pattern of neurologically incomplete spinal cord injury ischaracterized by disproportionately more motor impairmentof the upper than of the lower extremities, bladder dysfunc-tion and varying degrees of sensory loss below the level of thelesion (19). It has been associated with hyperextension injuriesof the cervical spine, even without apparent damage to thebony spine, but has also been described in association withvertebral body fractures and fracture-dislocation injuries. Thenatural history of acute central cervical spinal cord injuriesindicates gradual recovery of neurological function for mostpatients, albeit usually incomplete and related to the severityof the original injury and the age of the patient (4, 13, 15, 17,19–21). The role of surgery and its timing for patients withacute central spinal cord injuries without fracture compres-sion or dislocation injuries are the subjects of considerabledebate (3, 5–8, 19, 20). The optimal management of patientswho have sustained acute central cervical spinal cord injuriesis the subject of this review.

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject heading “spinal cord injury”combined with “central cord syndrome” and “incompletecervical spinal cord injury” yielded approximately 1450 cita-tions. Non-English language citations were excluded. Titlesand abstracts of the remaining publications were reviewed,

and relevant articles were selected to develop the guidelines.We focused on the specific issues of the natural history, med-ical management, and surgical treatment of human acutecentral cervical spinal cord injuries. These efforts resulted in13 articles (all Class III studies) specifically describing man-agement and outcomes of patients with central cervical spinalcord injuries. The Bibliography includes several articles onmagnetic resonance imaging (MRI) of central cervical spinalcord injuries, many articles (all Class III studies) describingseries of patients with acute spinal cord injuries, most ofwhom had incomplete cervical spinal cord injuries, and sev-eral general review articles that address issues of acute spinalcord injuries, including pathophysiology and treatment. The13 case series describing the management of patients withacute central cervical spinal cord injuries are summarized inTable 21.1.


In 1951, Schneider (18) described two patients with acuteneurologically incomplete cervical spinal cord injuries forwhom he suggested that early operation was indicated. Bothpatients presented after trauma with sudden loss of motorfunction in the distal upper extremities, the torso, and thelower extremities, but with preservation of touch and vibra-tion sense. Both patients had anterior spinal cord compressionfrom acute traumatic cervical disc herniations (one had anassociated vertebral endplate fracture). The diagnosis andanatomic localization were based on the clinical examination.Both patients made incomplete but significant neurologicalrecoveries after delayed surgical decompression via laminec-tomy, dentate ligament sectioning, and transdural discec-tomy. Three years later, Schneider et al. (19) described eight


patients they managed along with six other patients culledfrom the available literature. All but two of these patientspresented with disproportionately more motor impairment in

the upper extremities than in the lower extremities, bladderdysfunction with retention, and varying degrees of sensoryloss below the level of the lesion. Two of the six patients

TABLE 21.1. Summary of Reports on Acute Cervical Spinal Cord Injuriesa


Dai and Jia, 2000 (8) Retrospective review of 24 patients with acute discherniation as cause of ACCSCI treated with ACDF.

III Disc herniation common cause. Surgery successful in allpatients, more rapid improvement. Poor outcome with fracturedislocation injuries.

Newey et al., 2000 (15) Retrospective review of 32 patients with ACCSCImanaged conservatively.

III Improvement seen in most patients over time. Older patientshad worse outcome.

Chen et al., 1998 (6) Retrospective review of 37 patients with ACSI withpreexisting spondylosis. Many with central cord injurypattern. MRI assessment of compression, cord injury.16 managed with surgical decompression, 21medically.

III MRI modality of choice to image cord compression/injury.Surgical decompression associated with more rapidimprovement, shorter hospital and rehabilitation stay. Nodifference in outcome at 2-yr follow-up.

Chen et al., 1997 (7) Retrospective review of 114 patients with acute orchronic CCSCI. 28 patients managed with surgery (3chronic patients), 86 medically. No randomization.

III Surgery associated with more rapid and complete recovery,particularly in upper extremities, compared with similar patientsmanaged medically. Patients with long-segment stenosis hadpoor prognosis.

Bridle et al., 1990 (4) Random late assessment of 18 patients with ACCSCI. III Most patients improved over time, although most with long-term deficits, pain, and dysfunction.

Roth et al., 1990 (17) Retrospective review of 81 rehabilitation patients afterACCSCI.

III 2 age groups of patients, marked heterogeneity. In general,most patients improved over time. Outcome related to age andseverity of initial injury.

Merriam et al., 1986 (13) Retrospective review of 77 patients with ACCSCI. Nopatient with surgical decompression, 30 underwentlate stabilization and fusion.

III Marked variation among patients and injury patterns. Mostimproved. Outcome related to age and severity of initial injury.

Bose et al., 1984 (3) Retrospective review of 28 patients with ACCSCI, 14managed with aggressive medical therapy, 14 withmedical therapy and surgical treatment. Norandomization. Follow-up at time of discharge.

III No patient worse with treatment, medical or surgical. Surgeryprovided more rapid, more complete recovery at time ofdischarge.

Brodkey et al., 1980 (5) Retrospective review of 7 patients with ACCSCIoperated on late after injury. All had stable, profounddeficits and myelographic evidence of cordcompression.

III All had accelerated neurological improvement after surgery. 3patients normal at late follow-up. Surgery of benefit in selectedpatients with persistent deficits and evidence of cordcompression.

Shrosbree, 1977 (21) Retrospective review with late follow-up of 99patients with ACCSCI managed conservatively.

III 2 groups identified.Younger patients with flexion rotation injuries.Older patients with hyperextension injuries.Outcome related to severity of initial injury.

Bosch et al., 1971 (2) Retrospective review and long-term follow-up of 42patients with ACCSCI managed conservatively.

III Most patients improved over time. 75% regained ambulatoryskills, 56% regained functional hands. 10/42 patients had latedeterioration after initial gains (“chronic central cordsyndrome”).

Schneider et al., 1958 (20) Retrospective review of 12 additional patients withACCSCI. 11 managed expectantly, 1 managed withsurgical decompression 13 h after injury.

III 2 age groups of patients.Young patients with fracture dislocation injuries.Older patients with hyperextension injuries often without bonyvertebral damage.Most patients improved.Expectant management is ideal treatment.

Schneider et al., 1954 (19) Retrospective review (and first description) of 8patients with ACCSCI they managed (6 expectantly, 2surgically) and discussion of 6 cases from literature.

III Most patients with ACCSCI improved with time and expectantmanagement.Injury and its recovery follows specific pattern. Surgerycontraindicated for this injury.

a ACDF, anterior cervical discectomy with fusion; ACCSCI, acute central cervical spinal cord injury; ACSI, acute cervical spinal cord injury;CCSCI, central cervical spinal cord injury; MRI, magnetic resonance imaging.

Acute Central Cervical Spinal Cord Injuries


identified in the literature review had complete motor injuriesin both the upper and lower extremities with some preserva-tion of sensory function below the level of injury. Theseincomplete neurological deficits were related to acute trau-matic central cervical spinal cord injuries, usually—but notexclusively—as a result of hyperextension of the head andneck relative to the torso. In several patients, there was nodamage to the bony spine. In these instances, it was presumedthat hypertrophic changes (spurs, ridges, thickened liga-ments) within the spinal canal caused anterior and posteriorcord compression in the position of hyperextension, resultingin injury to the central substance of the cervical spinal cord.Other patients had cervical compression fractures or fracture-dislocation injuries of the cervical spine that contributed to thecentral spinal cord injury. The authors operated on the firsttwo of the eight patients they treated with this disorder. Bothhad central cervical spinal cord injuries without bony damageor displacement. Both were treated in delayed fashion vialaminectomy with sectioning of the dentate ligaments andthen transdural exploration anterior to the cervical spinalcord. In both cases, anterior bony osteophytes were identifiedbut were not removed. Patient 1 was quadriplegic postoper-atively. Patient 2 was neither better nor worse after surgery.Six additional consecutive patients were managed withoutsurgical decompression (Patient 7 underwent delayed dorsalcervical fusion in treatment of fracture instability). Five of sixpatients were managed expectantly (Patient 4 had progressionof neurological deficits despite immobilization and ultimatelydied 3 wk after injury), and three of six patients from theliterature improved without surgery (75%). Most patients ex-perienced permanent loss of hand function and strength. Oneof six patients from their series and three of six from theliterature died without neurological recovery (25%). Con-versely, one (50%) of two patients treated surgically wasimmediately neurologically worse; the second (50%) made aprogressive albeit incomplete recovery over time, much likethat observed in comparable patients managed withoutsurgery. It was on the basis of this early experience thatSchneider et al. determined that the prognosis after acutecentral cervical spinal cord injury was reasonably good. Sur-gery for these patients, they concluded, was contraindicatedand in fact known to harm rather than improve them (19).

In 1958, Schneider et al. (20) added observations on 12additional patients they managed with acute central cervicalspinal cord injuries. One patient died of pneumonia withoutneurological recovery, one patient had a full neurologicalrecovery, and the remaining 10 improved compared withtheir initial postinjury neurological status but were pro-foundly impaired at last follow-up. The authors noted twodistinct age groups with acute central cervical spinal cordinjuries. They described an older group of patients (mean age,59 yr) without bony vertebral damage but with hypertrophicchanges compromising the cervical spinal canal, and ayounger group (mean age, 31 yr) with fracture or fracture-dislocation injuries of the cervical spine. They reported thatcentral cord edema, venous congestion, and ischemia werecomponents of the pathophysiology of this unique injurytype. They advocated expectant management, including

closed reduction with skeletal traction for all patients withthis syndrome (as for four patients in their study), despiteimportant, near-complete neurological recovery in a 17-year-old patient after operative reduction and decompression of aunilateral facet dislocation injury within 13 hours of injury.

Schneider et al.’s (20) collective reported experience in themanagement of patients with acute central cervical cord inju-ries numbered 20 patients at the time of the report. Of the 20patients, 17 were managed medically: 2 patients died withoutimprovement, 14 patients improved but had profound resid-ual deficits, and 1 patient regained normal function. Threepatients were treated with surgical decompression: one early(hours) and two late (weeks). The patient with early decom-pression improved dramatically. One late decompression pa-tient neither improved nor worsened immediately after sur-gery but showed progressive long-term improvement; theother was quadriplegic after surgery. From this experience,they concluded that accurate diagnosis is stressed, with em-phasis placed on the fact that operation is contraindicated,that the prognosis may be good and that should recoveryoccur it will follow a definite pattern (20). These suggestionshave guided the care of acute central cervical spinal cordinjury patients ever since publication of their report.

In 1971, Bosch et al. (2) described observations made duringtheir management of 42 patients with subacute central cervi-cal spinal cord injuries treated at a rehabilitation hospital,with a follow-up period of 4 months to 26 years. At admis-sion, 19% could walk independently, 14% were partial walk-ers, and 67% could not walk. Twenty-six percent had func-tional hands at admission. At discharge, 57% had functionalwalking skills, 20% were partial walkers, and 42% had func-tional hands. Bladder control improved from 17% at admis-sion to 53% at discharge. A similar improvement in bowelcontrol was documented. Importantly, these authors notedlate deterioration in 24% of patients who showed initial im-provements in neurological function after central cervical spi-nal cord injury. Ten (24%) of 42 patients experienced the lateneurological sequelae of “chronic central cord syndrome” andlost walking, hand, and bladder control skills, as observed inlong-term follow-up. The authors concluded that at leastsome return of neurological function in the immediate postin-jury period could be expected in about 75% of cases, with 56%of patients regaining functional hands. In the long term, only59% of the patients with central cervical spinal cord injuriesthey followed retained functional skills with conventionalmedical management.

In the same year, Turnbull (26) reported his studies on themicrovasculature of the human spinal cord and postulatedmechanisms of vascular insufficiency associated with varioustypes of spinal cord displacement. His work speaks to theanatomic basis of the pathophysiology of acute cervical cen-tral spinal cord injuries, particularly those that occur in olderpatients with underlying cervical spondylosis who sustainacute central spinal cord injury without bony vertebral injury.He found that as the cord becomes compressed, whetherowing to a mass lesion or progressive cervical spondylosis, itbecomes flattened and widened. The vasculature of the cord isaffected by cord distortion. Pial vessels become more tortu-



ous. Arteries of the lateral columns are elongated, narrowed,and flattened. Branches from the central arteries that reach thegray matter run laterally and are similarly stretched length-wise and are compressed from side to side. Turnbull reportedthat vessels chronically deformed by cervical spondylosis can-not respond to additional anteroposterior flattening of thecord as would normal arteries in a younger patient. A littleadditional compression would pinch off side branches at theirorigins (26). He concluded that mechanical distortion of thecord and its blood supply plays a major role in the patho-physiology of spinal disease and spinal cord injury (26–28).

In 1977, Shrosbree (21) reported 99 patients with acutecentral cervical spinal cord injuries managed at a South Afri-can spinal cord injury center. Most of the patients were ad-mitted within 72 hours of injury. All patients were treatedconservatively. Fracture/injury reduction was accomplishedvia closed means in 92% of patients with dislocation, either bytraction or by reduction under anesthesia, within 72 hours ofadmission. Two age groups of patients were identified.Younger patients (21–50 yr) had flexion-rotation injuries anda higher incidence of dislocation injuries. Older patients (�50yr) were more likely to have hyperextension injuries super-imposed on preexisting cervical spondylosis. Outcome wasrelated to the severity of the initial neurological deficit. Only5 (22%) of 23 patients with severe motor deficits becameindependent walkers. All of these patients had residual defi-cits in the hands. The author summarized by noting that earlyreduction may well be a factor in promoting more favorableneurological recovery (21) among patients with facet disloca-tion injuries, but he provided no data to support his claim.

In 1977, Maroon (12) reported that “burning hands” (severedysesthesias in the hands and fingers after trauma despitenormal motor function) may indicate acute central spinal cordinjury. He described two football players with dysestheticsymptoms in the hands referable to modest injury to thecentral cervical spinal cord and warned physicians, trainers,and coaches of the importance of this syndrome.

In 1980, Brodkey et al. (5) revisited the management of theacute central cervical spinal cord injury syndrome. They pro-vided operative treatment to seven patients with traumaticcentral cervical spinal cord injuries within 18 to 45 days afteracute injury who had profound residual neurological deficitsafter attentive medical treatment. Myelography revealed sig-nificant defects in all of these patients. Four patients under-went anterior cervical discectomy with fusion (ACDF), onewas treated with multilevel laminectomy, one had multilevelACDF, and one received multilevel laminectomy and thendelayed (4 yr) multilevel ACDF. All patients had an acceler-ated neurological recovery after the surgical procedure, eventhe patient who deteriorated years after laminectomy andrequired late multilevel ACDF. Three patients returned tonormal after severe injuries that persisted until surgical de-compression. The authors concluded that cord compressiondoes play an important role in the pathophysiology of centralcord syndrome and that, when present in the setting of astable poor neurological condition after injury, decompres-sion of the spinal cord may be of benefit.

Bose et al. (3) retrospectively analyzed their management of28 patients with acute central cervical spinal cord injuries. Intheir 1984 report, they noted significantly improved motorscores at the time of discharge in patients managed withcombined medical therapy and surgery, compared with thosemanaged only medically. All were treated aggressively in theintensive care unit setting. Surgical patients had myelo-graphic evidence of cord compression or evidence of cervicalspinal instability. Although selection bias (surgical patientshad cord compression and/or instability or subluxation) andseveral other study flaws precluded direct comparison be-tween the groups, the authors noted that no surgical patientworsened as a result of surgery and all improved neurologi-cally, several substantially. They argued that decompressionof the compressed spinal cord in patients with acute centralcervical spinal cord injury syndrome may be of benefit inselected patients.

Merriam et al. (13), Roth et al. (17), Bridle et al. (4), andNewey et al. (15) described the late outcomes of individualseries of selected groups of patients after central cervicalspinal cord injury. All four groups of investigators notedmarked heterogeneity among patients in the central cervicalspinal cord injury population. All patients were managedmedically. Most patients improved somewhat over time, withmore recovery in the lower extremities than in the upperextremities. All of these authors concluded that outcome wasin general good for patients younger than 70 years. The finalneurological result was influenced by patient age, particularlyage older than 70 years, and the degree of initial neurologicalimpairment. Hand function impairment was the most com-mon long-term disability, even among patients with a “good”outcome. Only Merriam et al. (13) made reference to surgicaltreatment, involving 30 of 77 patients in their series, presum-ably to provide spinal stabilization and fusion. No associationbetween surgical management and outcome was discerned.

Chen et al. (7) reported their experience with 114 patientswith acute and chronic traumatic central cervical cord syn-drome. Twenty-eight patients were managed with surgicaltreatment, and 86 were managed medically. The authors didnot randomize patients to one treatment group or another.Selection criteria for surgical decompression included failureto improve with medical therapy or deterioration in neuro-logical function despite medical treatment with radiographic(MRI or computed tomographic/myelographic) evidence offocal cord compression, or gross instability of the spine. Theyoperated on three patients late (8, 12, and 24 mo after injury)for “chronic” central cord syndrome. Their 1997 retrospectivereview found that younger patients had better long-term re-sults than did older patients (in both management groups)and that surgical decompression was associated with morerapid and complete motor improvement compared with pa-tients managed medically, even if the operation was per-formed a long time after injury. Both management groups hadsimilar outcomes over time with respect to lower extremityand bladder function. Patients selected for surgery had morerapid and more complete recovery of function, particularly inthe upper extremities. The authors noted that patients withstenosis at multiple levels who were managed conservatively

Acute Central Cervical Spinal Cord Injuries S169


had a poorer prognosis and a relatively higher chance todevelop late myelopathy. The authors did not describe theoutcome of similar patients with multilevel stenosis managedwith operative decompression.

In 1998, Chen et al. (6) described the management of 37patients with preexisting cervical spondylosis who sustainedacute incomplete neurological cervical spinal cord injuriesafter trauma. Many of these patients had acute central spinalcord injuries. No patient sustained a bony vertebral columninjury. In their retrospective review, patients were treatedwith surgical decompression if they did not improve morethan one motor grade within 9 days of injury (range, 3–14 d).Patients were studied with MRI to document cord compres-sion and/or signal change within the spinal cord. In total, 16patients underwent surgical decompression, 9 anteriorly and7 posteriorly. Twenty-one patients were managed medically.Thirteen (81%) of 16 surgical patients improved “remarkably”immediately after surgery. Thirteen (62%) of 21 patients man-aged medically improved to the same degree over time. Aswith surgical patients, patients with cervical stenosis in morethan three vertebral levels fared less well than did patientswith focal compression or with stenosis in three vertebrallevels or fewer. There were no reported differences in out-come between patients in the two groups at 2-year follow-up.The authors concluded that surgical decompression might beassociated with more rapid neurological improvement, earlymobilization, and shorter periods of hospitalization and reha-bilitation. They consider MRI to be the imaging modality ofchoice to assess the spinal cord in patients with acute centralcervical spinal cord injuries, a conclusion consistent withthose of other investigators of the role of MRI in the assess-ment of patients with spinal cord injuries (9, 14, 16, 25).

In 2000, Dai and Jia (8) described their experience with 24patients with acute traumatic disc herniation as the cause ofacute central cervical spinal cord injuries. Acute disc herniationwas confirmed with preoperative MRI. The authors provided aretrospective assessment of patients operated on anteriorly(ACDF without internal fixation) for cord decompression andspinal stabilization. The timing of surgery relative to injury wasnot described. They noted an inverse correlation between rate ofrecovery and age and found that patients with fracture-dislocation injuries with acute disc herniation were more im-paired preoperatively and fared less well than patients withoutfracture-dislocation injuries at late follow-up. They reported thatsurgical decompression, stabilization, and fusion were success-ful in all patients and described marked improvement in neuro-logical function in most patients treated.

Contemporary reviews confirm early reports that most pa-tients with incomplete cervical spinal cord injuries meetingthe clinical neurological criteria of acute central spinal cordinjury will show neurological improvement over time (2, 13,15, 17, 19–21). Some patients with these injuries will die, andmany will remain profoundly impaired at late follow-up.These patients in general are older, have spinal cord injurieswithout bony vertebral injury, and have medical comorbidity,or they are younger but have fracture-dislocation injuries as acause of their neurological deficits. A large portion of patientswill regain walking skills over time but will not have useful

hands. A smaller portion of patients will demonstrate signif-icant neurological recovery and regain hand function. Thesepatients are typically younger, do not have fracture-dislocation injuries, and have less severe neurological deficitsat the outset. Up to 24% of patients managed nonoperativelywill improve early but decline again years later (“chroniccentral cord injury syndrome”) (2).

Surgery for decompression of the spinal cord in patientswith acute central cervical spinal cord injuries has been de-nounced on the basis of Schneider’s early poor experiencewith a single patient who underwent surgery (19, 20). Thatpatient, quadriplegic after dorsal cervical exploration anddecompression, experienced significant manipulation of theinjured cord during the process of dentate ligament sectioningand transdural anterior cord exploration (19), a procedureunlikely to be performed in similar fashion today. The samegroup had a rewarding experience with early (13 h afterinjury) surgical decompression and facet fracture reduction ina 17-year-old boy with profound early central cord neurolog-ical deficits (20). Many other authors, including those report-ing three contemporary series of patients with this disorder,have described good to excellent outcomes without neurolog-ical complications for surgical decompression of patients withspinal cord compression, particularly focal anterior cord com-pression (3, 5–8). However, no study to date has provided arandomized direct comparison of surgical patients with sim-ilar patients managed without surgery. Nor has any studyadequately assessed the potential role of dorsal spinal decom-pression for multilevel cervical cord compression in patientswith this disorder, particularly those with acute central cer-vical cord syndrome without bony vertebral damage. Surgerymay have a role in the management of patients with acutecentral cervical spinal cord injury, but, as yet, that role has notbeen accurately defined by scientific study.

Schneider et al.’s (20) conclusion that central cord edema,venous congestion, and ischemia were important componentsof the pathophysiology of these injuries, combined with thehypothesis of Turnbull (26, 27) and Turnbull et al. (28) thatvascular compression and distortion attributable to antero-posterior flattening of the cord plays a major role in thepathophysiology of cord injury, suggest several potential op-portunities for treatment. The compression of the cord anddistortion and compression of its blood supply might berelieved by surgical decompression. Ischemia of the cord,caused by either the primary injury or secondary events,might be improved with augmentation of spinal cord perfu-sion. Although Turnbull (26, 27) and Turnbull et al. (28) didnot offer specific strategies, they did offer an anatomic andpathophysiological rationale for the potential of maintenanceof spinal cord perfusion pressures and cervical cord decom-pression for patients who sustain an acute central cervicalcord injury, particularly those with preexisting cervical spon-dylosis. Maintenance or increases in systemic blood pressuremay improve perfusion to the injured, distorted spinal cord(1, 3, 10, 11, 22, 23, 29). Several contemporary series of patientswith spinal cord injuries treated with aggressive medicalmanagement including maintenance of mean arterial bloodpressure at high normal ranges (85–90 mm Hg) have sug-



gested improved neurological outcomes with this manage-ment plan (10, 11, 24, 29–31). Decompression of the cord hasthe potential to eliminate both cord compression and vascularcompression and distortion (3, 5–8, 30). Either or both of thesetreatment strategies may improve spinal cord blood flow inthe acute central cervical spinal cord injury setting, whichcould translate into preservation of neurological tissue andrecovery of neurological function. The benefit may or may notbe realized at the site of primary injury, but rather at vulner-able adjacent spinal cord levels fed by sulcal and collateralarteries that pass through the injury site but supply the cordrostral and caudal to the site of injury (1, 22, 23, 27, 28).

SUMMARY

The ideal management strategy for patients with acutecentral cervical spinal cord injuries seems to be multifaceted.As Schneider et al. (20) insisted years ago, a rapid, accuratediagnosis is essential. A detailed clinical examination, cervicalspinal x-rays to assess vertebral column injury (see recom-mendations in Chapter 5), and MRI assessment of the cervicalspinal cord for intrinsic injury and/or compression will ac-complish this goal. Many of these patients may require man-agement in the intensive care unit setting (see recommenda-tions in Chapter 7) for monitoring and treatment of cardiac,pulmonary, and blood pressure disturbances. Blood pressureaugmentation to mean arterial blood pressure levels of 85 to90 mm Hg may be of benefit (see recommendations in Chap-ter 8). Early reduction of fracture or fracture-dislocation inju-ries should be accomplished (see recommendations in Chap-ter 20). Administration of pharmacological agents may be ofbenefit according to specific parameters (see recommenda-tions in Chapter 9). Surgical decompression of the compressedspinal cord, particularly if the compression is focal and ante-rior and is approached anteriorly, seems to be of benefit inselected patients.

KEY ISSUES FOR FUTURE INVESTIGATION:

A prospective, controlled, randomized investigation of pa-tients with acute central cervical spinal cord injuries treatedwith aggressive medical therapy alone (intensive care unitmanagement, blood pressure augmentation, closed fracture-dislocation reduction), compared with patients managed withaggressive medical therapy and early surgical decompressionof the spinal cord, is needed.


REFERENCES


2. Bosch A, Stauffer ES, Nickel VL: Incomplete traumatic quadriple-gia: A ten-year review. JAMA 216:473–478, 1971.


4. Bridle MJ, Lynch KB, Quesenberry CM: Long term functionfollowing the central cord syndrome. Paraplegia 28:178–185,1990.

5. Brodkey JS, Miller CF Jr, Harmody RM: The syndrome of acutecentral cervical spinal cord injury revisited. Surg Neurol 14:251–257, 1980.

6. Chen TY, Dickman CA, Eleraky M, Sonntag VKH: The role ofdecompression for acute incomplete cervical spinal cord injuryin cervical spondylosis. Spine 23:2398–2403, 1998.

7. Chen TY, Lee, ST, Lui TN, Wong CW, Yeh YS, Tzaan WC, HungSY: Efficacy of surgical treatment in traumatic central cord syn-drome. Surg Neurol 48:435–441, 1997.

8. Dai L, Jia L: Central cord injury complicating acute cervical discherniation in trauma. Spine 25:331–336, 2000.

9. Kalfas I, Wilberger JE, Goldberg A, Prostko ER: Magnetic reso-nance imaging in acute spinal cord trauma. Neurosurgery 23:295–299, 1988.

10. Levi L, Wolf A, Belzberg H: Hemodynamic parameters inpatients with acute cervical cord trauma: Description, interven-tion, and prediction of outcome. Neurosurgery 33:1007–1017,1993.

11. Levi L, Wolf A, Rigamonti D, Ragheb J, Mirvis S, Robinson WL:Anterior decompression in cervical spine trauma: Does thetiming of surgery affect the outcome? Neurosurgery 29:216–222, 1991.

12. Maroon JC: “Burning hands” in football spinal cord injuries.JAMA 238:2049–2051, 1977.

13. Merriam WF, Taylor TK, Ruff SJ, McPhail MJ: A reappraisal ofacute traumatic central cord syndrome. J Bone Joint Surg Br65B:708–713, 1986.

14. Mirvis SE, Geisler FH, Jelinek JJ, Joslyn JN, Gellad F: Acutecervical spine trauma: Evaluation with 1.5-T MR imaging. Radi-ology 166:807–816, 1988.

15. Newey ML, Sen PK, Fraser RD: The long-term outcome af-ter central cord syndrome. J Bone Joint Surg Br 82B:851–855,2000.

16. Quencer RM, Bunge RP, Egnor M, Green BA, Puckett W, NaidichTP, Post MJ, Norenberg M: Acute traumatic central cord syn-drome: MRI-pathological correlations. Neuroradiology 34:85–94,1992.

17. Roth EJ, Lawler MH, Yarkony GM: Traumatic central cord syn-drome: Clinical features and functional outcomes. Arch PhysMed Rehabil 71:18–23, 1990.

18. Schneider RC: A syndrome in acute cervical spine injuries forwhich early operation is indicated. J Neurosurg 8:360–367,1951.

19. Schneider RC, Cherry G, Pantek H: The syndrome of acute centralcervical spinal cord injury. J Neurosurg 546–577, 1954.

20. Schneider RC, Thompson JC, Bebin J: The syndrome of acutecentral cervical spinal cord injury. J Neurol Neurosurg Psychia-try 21:216–227, 1958.

21. Shrosbree RD: Acute central cervical spinal cord syndrome: Ae-tiology, age incidence and relationship to the orthopaedic injury.Paraplegia 14:251–258, 1977.

22. Tator CH: Ischemia as a secondary neuronal injury, in SalzmanSK, Faden AI (eds): Neurobiology of Central Nervous System Trauma.New York, Oxford University Press, 1994, pp 209–215.

Acute Central Cervical Spinal Cord Injuries S171



24. Tator CH, Rowed DW, Schwartz ML, Gertzbein SD, Bharatwal N,Barkin M, Edmonds VE: Management of acute spinal cord inju-ries. Can J Surg 27:289–296, 1984.

25. Tracy PT, Wright RM, Hanigan WC: Magnetic resonance imagingof spinal injury. Spine 14:292–301, 1989.


27. Turnbull IM: Blood supply of the spinal cord: Normal and patho-logical considerations. Clin Neurosurg 20:56–84, 1973.

28. Turnbull IM, Brieg A, Hassler O: Blood supply of cervical spinalcord in man: A microangiographic cadaver study. J Neurosurg24:951–965, 1966.

29. Vale FL, Burns J, Jackson AB, Hadley MN: Combined medical andsurgical treatment after acute spinal cord injury: Results of a prospectivepilot study to assess the merits of aggressive medical resuscitation andblood pressure management. J Neurosurg 87:239–246, 1997.


31. Zach GA, Seiler W, Dollfus P: Treatment results of spinal cord injuriesin the Swiss Paraplegic Centre of Basel. Paraplegia 14:58–65, 1976.

Dissection showing some of the conspicuous differences in passing from the vertebral column to the skull. From, Tilney F,Alsop Riley H: The Form and Functions of the Central Nervous System: An Introduction to the Study of Nervous Diseases.New York, Harper & Brothers, 1938, 3rd ed.


CHAPTER 22

Management of Vertebral Artery Injuries after NonpenetratingCervical Trauma

RECOMMENDATIONSDIAGNOSTIC:Standards: There is insufficient evidence to support diagnostic standards.Guidelines: There is insufficient evidence to support diagnostic guidelines.Options: Conventional angiography or magnetic resonance angiography is recommended for the diagnosis of

vertebral artery injury after nonpenetrating cervical trauma in patients who have complete cervical spinalcord injuries, fracture through the foramen transversarium, facet dislocation, and/or vertebral subluxation.

TREATMENT:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options:• Anticoagulation with intravenous heparin is recommended for patients with vertebral artery injury who

have evidence of posterior circulation stroke.• Either observation or treatment with anticoagulation in patients with vertebral artery injuries and evidence

of posterior circulation ischemia is recommended.• Observation in patients with vertebral artery injuries and no evidence of posterior circulation ischemia is

recommended.

RATIONALE

The association of cerebrovascular insufficiency and cer-vical fracture was first described by Suechting andFrench (13) in a patient with Wallenburg’s syndrome

occurring 4 days after a C5–C6 fracture-dislocation injury.Although Schneider et al. (8) implicated vertebral artery in-jury (VAI) at the site of cervical fracture-dislocation as a causeof posterior circulation cerebral ischemia, Gurdjian et al. (4)suggested that unilateral vertebral artery occlusions might beasymptomatic. Subsequent articles described larger series ofpatients with asymptomatic VAIs after blunt cervical spinaltrauma (2, 18). However, Biffl et al. (1), in the largest prospec-tive series to date, consisting of 38 patients with VAI diag-nosed by angiography, reported more frequent strokes inpatients not treated initially with intravenous heparin antico-agulation despite an asymptomatic VAI. Fractures throughthe foramen transversarium, facet fracture-dislocation, or ver-tebral subluxation is almost always seen in patients with VAI(1–3, 5, 17–19). A cadaveric study demonstrated progressivevertebral occlusion with higher degrees of flexion-distractioninjury, confirming this clinical observation (11). To developdiagnostic and treatment recommendations for VAI afterblunt cervical trauma, an analysis of the articles examining itsmanagement is undertaken in this report. Specific issues that

were addressed include the clinical and radiographic criteriaused to prompt diagnostic evaluation, the appropriate diag-nostic tests for identifying VAI, and the management of VAI(observation versus anticoagulation with heparin).

SEARCH CRITERIA

A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject headings “vertebral artery,”“cervical vertebrae,” “dislocation,” and “wounds and inju-ries” yielded 6,447, 15,667, 24,174, and 459,759 citations, re-spectively. Combination of the first two headings with thethird heading yielded a subset of 61 citations. Combination ofthe first two headings with the fourth heading yielded asubset of 239 citations. Abstracts were reviewed, and onlyarticles in English containing three or more human subjectswith VAI after blunt cervical trauma were included. Fourteenarticles, including two prospective studies that provide ClassII evidence (1–3, 5, 10, 16–18), met these selection criteria andprovided data on 122 patients for this report. These articlesare summarized in Table 22.1. A total of 19 referenced articlesprovide the foundation for this review.


TABLE 22.1. Summary of Reports on Vertebral Artery Injuriesa


Schellinger et al., 2001

(7)

Retrospective review of 4 patients with VA dissection

diagnosed with various imaging studies, among 27 patients

with cervical spine injuries.

III 2 with motor complete cord injuries awoke after cervical surgery with stupor

and posterior circulation infarcts; 1 with basilar occlusion by Doppler died,

and 1 with dissection by DSA partially recovered. 1 with C5 foramen

transversarium fracture became comatose 12 h after admission and had fatal

basilar thrombosis by CT hours later. 1 with treated C2 fractures had vertigo

and nystagmus 3 wk later; dissection by angio; symptoms resolved with

warfarin followed by aspirin.

Biffl et al., 2000 (1) Prospective angio screening in blunt craniovertebral injury

identified 38 patients with VA injuries.

III 38 patients with 47 VA injuries of approx. 350 angios performed. 27/38 had

cervical fractures. 6/27 had fractures through the foramen transversarium. 9/38

had a posterior circulation stroke. Stroke not related to occlusion versus

stenosis. 8-h to 12-d delay, most beyond 48 h. 3/21 asymptomatic patients

treated with heparin had stroke versus 6/17 without heparin had stroke (P �

0.13). 1/6 died with heparin. 2/3 died without heparin. 2 treated with heparin

had hemorrhagic strokes.

Weller et al., 1999 (17) Prospective MRA in 12 patients with foramen transversarium

fractures.

III 3/12 had VA occlusion; all remained asymptomatic on aspirin. 1/12 with

stenosis had delayed syncope on aspirin, resolved with brief intravenous

heparin followed by aspirin.

Vaccaro et al., 1998

(16)

Prospective study with follow-up MRA in 6/12 patients (1

excluded) previously reported by Giacobetti et al. (3).

III 1/6 treated without heparin reconstituted by 12 d. 5/6 remained occluded (2/5

with heparin) �1 yr later.

Giacobetti et al., 1997

(3)

Prospective study with MRA in 61 patients with cervical

injuries found 12 patients with VA occlusion.

III 1/4 with transverse foramen fractures had occlusion. 6/15 with facet

dislocation had occlusion. 3/12 with transient blurred vision resolved with 3

mo anticoagulation.

Thibodeaux et al., 1997

(14)

Retrospective review of 3 patients with vertebral dissection on

angio.

III 1 with blindness but no infarct improved with anticoagulation. 1 with C4–C5

facet fracture was asymptomatic without treatment. 1 with ataxia/dizziness 2 d

later with occipital infarct recovered with 6 mo anticoagulation.

Prahbu et al., 1996 (6) Retrospective review of 5 symptomatic patients with 4 VA

occlusions and 1 stenosis on MRA and/or angio.

III 3 with fractures had delayed coma at 3 h, 2 occlusions/1 narrowing, 1 died

with stroke, 2 improved with anticoagulation, but it was discontinued in 1

secondary to rectus hematoma. 1 with fracture had delayed confusion/aphasia

with multiple MCA strokes, improved without anticoagulation. 1 with fracture

was asymptomatic without anticoagulation.

Friedman et al., 1995

(2)

Prospective study of 37 patients with nonpenetrating cervical

trauma found 9 VA injuries by MRA.

II 50% patients with complete cord injuries had VA injury versus 12% patients

with incomplete cord injuries (P � 0.02). 5/13 patients with �3 mm

subluxation had VA injuries versus 4/24 patients with �3 mm subluxation. 1

with bilateral VA injuries died of large cerebellar infarct (bilateral foramen

transversarium fractures) 8 asymptomatic (1/8 with anticoagulation also had

carotid occlusion).

Tulyapronchote et al.,

1994 (15)

Retrospective study of 3 patients with VB ischemia �2 wk after

occult C2 fractures diagnosed by angio.

III 2 occluded, 1 narrowing. Symptoms included syncope, vertigo, dysphagia,

dysarthria, facial numbness, blurred vision. Treatment not reported.

Willis et al., 1994 (18) Prospective study of 26 patients with cervical facet dislocation

or facet fractures through foramen transversarium fractures

revealed 12 with VA injuries on angio.

II 9/14 with normal angio had foramen transversarium fractures versus 7/12 with

abnormal angio. 1/9 with occlusion without anticoagulation died from

unrelated injuries. 1 dissection became an occlusion on heparin. 1 intimal

flap/1 pseudoaneurysm healed with heparin in 7–10 d.

Sim et al., 1993 (10) Prospective delayed duplex sonography of 11 patients with

previously locked facets.

III 1/11 had occlusion (persistent locked facet). 1/11 had a narrow VA.

Woodring et al., 1993

(19)

Retrospective study of 216 patients with cervical fractures

showed 52 with TP fractures. 8 had angio.

III 78% of TP fractures extended into foramen transversarium. 4/8 had occlusion,

3/8 had dissection, 1 of each had stroke that improved with anticoagulation. 3

asymptomatic patients treated with anticoagulation.

Schwarz et al., 1991 (9) Retrospective review of 4 patients with symptomatic VA

injuries by angio.

III 4 had delayed symptoms (3 occlusions, 1 pseudoaneurysm). 1 with coma and

stroke 20 min after reduction of C4–C5 dislocation received streptokinase/

heparin for occlusion but died.

Louw et al., 1990 (5) Prospective study of 12 patients with facet dislocations with

DSA.

III 4/5 unilateral and 5/7 bilateral dislocations had occluded VAs (1 bilateral). 3

occluded at level, 5 within 2 cm. 2/9 symptomatic patients had bilateral C5–

C6 facet dislocation and improved without treatment.

a DSA, digital subtraction angiography; CT, computed tomography; angio, angiography; MRA, magnetic resonance angiography; VA, vertebralartery; MCA, middle cerebral artery; VB, vertebrobasilar artery; TP, transverse process.




Diagnosis

The diagnosis of VAI can be made with a variety of radio-graphic studies. Angiography is the traditional imaging tech-nique used to diagnose VAI. Angiography was used for allpatients in seven of the studies reviewed (1, 5, 9, 14, 15, 18, 19)and in combination with other modalities in two additionalstudies (6, 7). Three studies prospectively applied conven-tional angiography to patients who had sustained nonpen-etrating cervical trauma and who met certain inclusion crite-ria (1, 5, 18). Similarly, three studies prospectively appliedmagnetic resonance angiography (MRA) to patients who hadsustained nonpenetrating cervical trauma and who met cer-tain inclusion criteria (2, 16, 17).

Biffl et al. (1) reported the largest prospective study usingangiography. Study subjects were selected from among 7205blunt trauma patients by the use of clinical and radiographiccriteria. Patients underwent angiography if they had facialhemorrhage (bleeding from mouth, nose, ears), cervical bruit(in patients �50 yr), expanding cervical hematoma, cerebralinfarction by computed tomography (CT), or lateralizing neu-rological deficit. “Asymptomatic” patients were selected forangiography if they had cervical hyperextension/rotation orhyperflexion injuries, closed head injury with diffuse axonalinjury, near hanging, seat belt or other soft tissue injuries tothe neck, basilar cranial fractures extending into the carotidcanal, and cervical vertebral body fractures or distractioninjuries. Between 350 and 400 angiograms were performed,identifying 38 patients with VAI. However, neither the exactnumber of angiograms performed nor the number of patientswho met the various criteria without VAI were reported. As aresult, sensitivity, specificity, positive predictive value, andnegative predictive value of the selection criteria could not bedetermined. Cervical spine injuries were observed in 27 of 38patients with VAI, including fractures through the foramentransversarium in four, dislocations in six, vertebral subluxa-tions in two, and more than one of these injuries in two.Twenty-nine patients had unilateral VAI (18 left, 11 right);nine had bilateral VAIs. A vascular injury scale was used tostratify patients into five categories: Grade I, arterial dissec-tions with less than 25% luminal narrowing; Grade II, arterialdissections with more than 25% luminal narrowing; Grade III,pseudoaneurysm of the vertebral artery; Grade IV, occlusionof the vertebral artery; and Grade V, vertebral artery transec-tion. Seven patients died; five had bilateral VAIs (Grade I),and two had unilateral VAI (one Grade I, one Grade IV).Three patients with either no neurological deficit or milddeficit had bilateral VAIs. The authors concluded that strokeincidence and neurological outcome appeared to be indepen-dent of the grade of VAI.

Another prospective study by Willis et al. (18) identified 30patients with midcervical fractures and/or dislocation forangiography. However, only 26 patients who met the criteriaagreed to proceed with angiography. Twelve patients sus-tained VAIs demonstrated by angiography (six left occlu-sions, three right occlusions, one left intimal flap, one left

pseudoaneurysm, one left dissection). The authors providedsufficient data regarding the presence of foramen transver-sarium fracture, facet dislocation, and subluxation to deter-mine the usefulness of these radiographic findings in identi-fying patients with VAI. The calculated sensitivity of foramentransversarium fracture as a criterion for identifying VAI inthis study was 58%; the specificity was 36%. The positivepredictive value of foramen transversarium fracture was 44%;the negative predictive value was 50%. The calculated sensi-tivity of facet dislocation as a criterion for identifying VAI was42%; the specificity was 57%. The positive predictive value offacet dislocation was 45%; the negative predictive value was53%. The calculated sensitivity of subluxation as a criterionfor identifying VAI was 67%; the specificity was 29%. Thepositive predictive value of subluxation was 80%; the negativepredictive value was 50%. Any combination of foramen trans-versarium fracture, facet dislocation, and/or vertebral sublux-ation revealed a calculated sensitivity for identifying VAI of92% and a specificity of 0%. The positive predictive value ofthe presence of any of the three criteria and VAI was 44%; thenegative predictive value was 0%.

A prospective study by Louw et al. (5) examined 12 con-secutive patients with cervical spine facet dislocations withdigital subtraction angiography. Five of seven patients withbilateral facet dislocations had vertebral artery occlusion (onebilaterally), and four of five patients with unilateral facetdislocations had unilateral vertebral artery occlusion. Angiog-raphy was not performed in blunt cervical trauma patientswithout facet dislocation. In a retrospective study byWoodring et al. (19), seven of eight patients with transverseprocess fractures who underwent angiography had VAIs (tworight occlusions, two left occlusions, two right dissections, oneleft dissection). Seventy-eight percent of transverse processfractures extended into the foramen transversarium. Angiog-raphy was not performed in 44 other patients with transverseprocess fractures. Alternatively, MRA has been used to diag-nose VAI noninvasively. Weller et al. (17) prospectively ex-amined 12 patients who had experienced nonpenetrating cer-vical trauma and who sustained fractures through theforamen transversarium. Three patients had unilateral verte-bral artery occlusion, and one had focal narrowing, all at thesite of fracture. MRA was not performed on the 26 patientswithout these fractures. Giacobetti et al. (3) prospectivelyevaluated all patients admitted with cervical spine injurieswith MRA. Twelve of 61 patients had vertebral artery occlu-sion demonstrated by MRA; all injuries were unilateral (sixleft, six right). Although 7 of 12 patients with VAI had flexion-distraction injuries with facet dislocations, the types of cervi-cal spinal injuries sustained by the 49 patients with normalMRA results were not reported. Because none of these fourarticles (3, 5, 17, 19) provided sufficient information regardingthe types of injury and results of vertebral artery imaging inthe entire population of patients studied, sensitivity, specific-ity, positive predictive value, and negative predictive value ofthe injury types could not be determined.

Friedman et al. (2) prospectively examined 37 patients ad-mitted with “major” blunt cervical spine injuries identified byMRA and compared these patients with a size-matched con-

Vertebral Artery Injuries in Nonpenetrating Cervical Trauma S175


trol group of patients without a history of cervical trauma.Nine patients had VAIs (six unilateral occlusions, two nar-rowing, one bilateral injury). Both vertebral arteries werevisualized in all 37 control subjects. Complete spinal cordinjuries were observed in 12 of 37 patients with cervicaltrauma, 6 of whom had VAIs (P � 0.02; �2 test). More than 3mm of subluxation was observed in 13 of 37 patients, 5 ofwhom had VAIs (P � 0.14; �2 test). The calculated sensitivityof complete spinal cord injury as a criterion for identifyingVAI in this study was 67%; the specificity was 79%. Thepositive predictive value of complete spinal cord injury was50%; the negative predictive value was 88%. The calculatedsensitivity of subluxation as a criterion for identifying VAIwas 56%; the specificity was 71%. The positive predictivevalue of subluxation was 38%; the negative predictive valuewas 83%.

Other diagnostic modalities have also been used to identifyof VAI. CT with intravenous contrast demonstrated a unilat-eral vertebral artery occlusion in one patient with a Jeffersonfracture, which was subsequently confirmed by angiography(12). Duplex sonography has also been used to diagnose VAI(7, 10, 14). Angiography has occasionally been used to confirmthe results of MRA or ultrasonography, but there has not beena study comparing MRA or ultrasonography with angiogra-phy in the diagnosis of VAI.

Treatment

After diagnosis of VAI, treatment options examined by thevarious studies have included observation alone or anticoag-ulation with either intravenously administered heparin ororally administered antiplatelet agents. Some authors treatedasymptomatic patients (1, 2, 18, 19); others did not treat symp-tomatic patients (5, 7).

Several articles retrospectively identified patients with neu-rological complications of VAI. Schellinger et al. (7) describedfour patients with VAI who had delayed onset of neurologicaldysfunction. Two patients awoke after surgery with alteredconsciousness and posterior circulation stroke. One of thesepatients with vertebrobasilar occlusion diagnosed by ultra-sound died; the other with vertebral artery dissection andpseudoaneurysm improved. Neither patient was given anti-coagulation treatment. Two other patients developed delayedsymptoms at 12 hours and 3 weeks after injury. The firstpatient died from basilar artery thrombosis confirmed by CT.The second patient developed vertigo and nystagmus as aresult of a dissection diagnosed by ultrasound and angiogra-phy. This patient was successfully treated with intravenousheparin anticoagulation and recovered within several weeks.Thibodeaux et al. (14) reported one patient who developedblindness immediately after left vertebral artery dissection.No infarction was seen on CT. Treatment included 3 monthsof anticoagulation with sustained improvement at the 4-yearfollow-up examination. A second patient with ataxia anddizziness 2 days after VAI had occipital infarction by CT andbilateral vertebral artery dissections but recovered after 6months of anticoagulation therapy. One patient with dissec-tion remained asymptomatic without anticoagulation. Prabhu

et al. (6) reported three of five patients with VAI who expe-rienced sudden loss of consciousness 3 hours after injury. Allwere anticoagulated. One patient with vertebrobasilar throm-bosis died; one patient with bilateral vertebral occlusionsimproved. The third patient with vertebral artery stenosisimproved, although anticoagulation was stopped severaldays later secondary to a rectus sheath hematoma. Two pa-tients with asymptomatic vertebral artery occlusion were nottreated and remained asymptomatic. Tulyapronchote et al.(15) reported three patients with delayed onset of symptoms2 weeks to 3 months after VAI, including syncope, visualdisturbance, dysarthria, dysphagia, and vertigo. Two patientshad vertebral artery occlusions, and one had a dissection. Notreatment was reported for any of these patients. Woodring etal. (19) reported two strokes in nine patients with transverseprocess fractures after blunt cervical trauma. All nine werestudied with vertebral angiography; seven studies were ab-normal. One patient with vertebral artery occlusion wastreated with anticoagulation and improved. One with dissec-tion improved without treatment. Three of five asymptomaticpatients were anticoagulated (intravenous heparin convertedto warfarin; total treatment of 6 wk); two patients were nottreated. All five remained asymptomatic. Schwarz et al. (9)reported four patients with ischemic vertebrobasilar symp-toms after nonpenetrating cervical trauma. Two patients withatlantoaxial injuries recovered after atlantoaxial stabilization(one treated with halo immobilization, one operatively).Treatment of one patient with delayed symptoms after unrec-ognized facet dislocation was not reported. Streptokinase in-fusion was used in a patient with vertebral artery occlusionwho became comatose shortly after reduction of a facet dis-location injury. This strategy failed to achieve completethrombolysis, and the patient died days later.

Six prospective studies examining the diagnosis of VAIprovided the incidence of neurological complications relatedto VAI. Biffl et al. (1) reported the highest frequency of pos-terior circulation stroke: 24% (9 of 38 patients). Stroke oc-curred 8 hours to 12 days after injury in these patients, occur-ring more than 48 hours after injury in 78%. Three of 21asymptomatic patients treated with intravenous heparin sub-sequently developed stroke; one died and two had mild re-sidual deficits. In contrast, 6 of 17 asymptomatic patients notinitially treated with intravenous heparin had strokes; twodied, three had mild deficits, and one had a severe deficit.Neither of the two patients who died was treated with anti-coagulation when stroke occurred 9 to 12 days after VAI (oneGrade IV, one bilateral Grade I that progressed to right GradeII and left Grade III). All three asymptomatic patients whosubsequently developed stroke and were then treated withheparin had mild residual deficits (one Grade I, one Grade II,one Grade IV). A final patient who had a stroke and wastreated with antiplatelet agents developed severe residualdeficits (Grade I). Two of nine patients with stroke who weretreated with heparin developed hemorrhagic infarction. Twoof the 38 patients had dominant vertebral arteries; both hadinjuries to the dominant artery and neither developed astroke.



Weller et al. (17) reported on four of their patients withVAIs treated with aspirin. Three patients with vertebral arteryocclusion remained asymptomatic; one with vertebral ar-tery narrowing developed syncope 17 days after injury. Thispatient was treated with intravenous heparin followed byaspirin without recurrent symptoms. Giacobetti et al. (3) de-scribed 3 of 12 patients with vertebral artery occlusion whodeveloped transient blurred vision; all three were treated with3 months of anticoagulation (intravenous heparin followed bywarfarin), and none had recurrent symptoms. Friedman et al.(2) reported one patient with bilateral VAIs who died after amassive right cerebellar infarct. One patient with vertebralocclusion and concurrent carotid occlusion remained asymp-tomatic on heparin. The remaining seven patients remainedasymptomatic without treatment (five occlusions, two steno-sis). Willis et al. (18) reported no symptoms in 12 patients withVAIs, none of whom were treated with anticoagulation. Louwet al. (5) reported two of nine patients with vertebral arteryocclusion with symptoms, including blurring of vision, thatspontaneously resolved without anticoagulation.

The management and outcome of 106 patients with VAIafter nonpenetrating cervical injury could be determined fromthe articles reviewed. Twelve patients had radiographic evi-dence of a posterior circulation stroke as their first symptom.Four patients were treated with intravenous heparin; one diedand three improved (1, 19). The remaining eight patients werenot treated with intravenous heparin; five died, two im-proved, and one had a severe neurological deficit (1, 2, 7, 19).Fifteen patients developed symptoms of posterior circulationischemia without stroke before treatment was instituted (3,5–7, 9, 14, 17). Eleven of 15 patients were treated with intra-venous heparin; two died (both had strokes), and nine im-proved (one developed a stroke). The remaining four patientswere not treated with intravenous heparin; all improved with-out developing a posterior circulation stroke. Twenty-sevenasymptomatic patients were prophylactically treated with in-travenous heparin (1, 2, 18, 19). Three patients developedposterior circulation strokes; 27 patients remained asymptom-atic. Finally, 52 asymptomatic patients were not prophylacti-cally treated with heparin (8 were treated with aspirin and 1with embolization). Three patients died from injuries unre-lated to the VAI (1, 18). Four patients had severe deficits (threewere treated with aspirin and one with embolization), andthree had mild deficits (one was treated with aspirin); alldefects were related to the associated spinal cord injury (1).The remaining 42 patients (3 of whom were treated withaspirin) remained asymptomatic (1–3, 5, 6, 14, 17–19).

The articles reviewed did not specifically address the risk ofprogressive spinal cord hemorrhage worsening in patientswith VAI and an associated spinal cord injury treated withheparin. One patient with a complete cervical spinal cordinjury and hematomyelia was placed on intravenous heparinprophylactically for left carotid and vertebral artery occlusion;no neurological changes occurred with treatment (2).

At least 13 of 42 patients treated with intravenous heparinhad complications during their treatment; in 6 patients (14%),the complications were significant. Six patients (two of whomdied) developed posterior circulation strokes after treatment

with intravenous heparin was initiated (1, 6, 9, 14). Twopatients developed hemorrhagic posterior circulation strokes;the timing of intravenous heparin relative to the developmentof the posterior circulation stroke was not reported (1). Intra-venous heparin was discontinued in three patients. Onesymptomatic patient developed a rectus sheath hematoma (6);the patient’s symptoms stabilized after intravenous heparintreatment was discontinued. Two asymptomatic patients hadprogression in the grade of VAI; both remained asymptomaticafter intravenous heparin was discontinued (18). Four otherpatients had progression in the grade of VAI; intravenousheparin was continued, and none of the patients developeddeficits related to the progression of injury grade (1).

In several studies, patients were reimaged to determinewhether disease progression or resolution occurred after VAI.Biffl et al. (1) reported follow-up angiography on 21 patients.Of 16 patients treated with heparin, two improved to a lessergrade of vascular injury and four worsened to a higher grade.Of five patients not receiving heparin, one improved andthree had worse vascular injury grades. Vaccaro et al. (16),using MRA, found reconstitution in one of six patients withVAI 12 days after the original diagnosis; this patient was nottreated with anticoagulation. The other five patients still hadvertebral artery occlusion more than 1 year later, includingtwo treated with anticoagulation. Willis et al. (18) describedthe results of follow-up angiography in three patients withVAI. One patient with a pseudoaneurysm received 1 week ofintravenous heparin followed by aspirin; the pseudoaneu-rysm had slightly enlarged 7 days after treatment was begun,but it had disappeared on angiography performed 6 weekslater. One patient treated with intravenous heparin for avertebral artery dissection had an asymptomatic occlusion ofthe artery demonstrated by angiography 2 days later; heparinadministration was discontinued. The third patient wastreated with intravenous heparin for a vertebral artery intimalflap; the patient had a normal vertebral angiogram 10 dayslater. Thibodeaux et al. (14) found a patent vertebral artery 6months after dissection was diagnosed; this patient did notreceive anticoagulation treatment. Sim et al. (10) reporteddelayed duplex sonography in 11 patients with a history offacet dislocation but with unknown vertebral artery status atthe time of the original cervical spine injury. Two of thesestudies demonstrated VAI: one patient with persistent dislo-cation had a vertebral occlusion, and one patient with a re-duced injury had vertebral artery stenosis.

SUMMARY

The incidence of VAI may be as high as 11% after nonpen-etrating cervical spinal trauma in patients with specific clini-cal criteria, including facial hemorrhage (bleeding frommouth, nose, ears), cervical bruit in those younger than 50years, expanding cervical hematoma, cerebral infarction byCT, lateralizing neurological deficit, cervical hyperextension-rotation or hyperflexion injuries, closed head injury with dif-fuse axonal injury, near hanging, seat belt or other soft tissueinjuries to the neck, basilar cranial fractures extending into thecarotid canal, and cervical vertebral body fractures or distrac-

Vertebral Artery Injuries in Nonpenetrating Cervical Trauma S177


tion injuries. Many patients with VAI have complete spinalcord injuries, fracture through the foramen transversarium,facet dislocation, and/or vertebral subluxation, but many pa-tients with these injuries have normal vertebral arteries whenimaged, thus compromising the specificity of these injurycriteria. Many patients with VAI are asymptomatic, includingthose with vertebral artery occlusion or dissection. The liter-ature reviewed indicates that patients with posterior circula-tion stroke and VAI have a better outcome when treated withintravenous heparin than patients who do not receive thistreatment. However, others have reported improvementamong patients without anticoagulation (7, 19). The outcomeof patients who develop symptoms of posterior circulationischemia without stroke and are treated with intravenousheparin (3, 6, 7, 14, 17) is similar to that of patients receivingno treatment (5, 9). Although the largest prospective studysuggested a trend toward less frequent stroke in asymptom-atic patients treated with heparin (1), others have not reportedsimilar observations (2, 3, 5, 6, 14, 18, 19). Because the risk ofsignificant complications related to anticoagulation is approx-imately 14% in these studies, there is insufficient evidence torecommend anticoagulation in asymptomatic patients.


Although several prospective studies examined patients atrisk for VAI, most articles did not provide enough data on thecharacteristics of the patients with normal vertebral arteries tounderstand the clinical or radiographic characteristics thatwould predict which patients may have VAI. A prospectivestudy comparing MRA with conventional angiography innonpenetrating cervical spine trauma may define the role ofnoninvasive imaging studies in these patients. A multicenter,randomized, prospective study comparing anticoagulationwith intravenous heparin versus observation in asymptomaticpatients and in symptomatic patients with posterior circula-tion ischemia but without stroke is recommended to deter-mine whether anticoagulation of these patients is justified.


REFERENCES

1. Biffl WL, Moore EE, Elliott JP, Ray C, Offner PJ, Franciose RJ,Brega KE, Burch JM: The devastating potential of blunt vertebralarterial injuries. Ann Surg 231:672–681, 2000.

2. Friedman D, Flanders A, Thomas C, Millar W: Vertebral arteryinjury after acute cervical spine trauma: Rate of occurrence asdetected by MR angiography and assessment of clinical conse-quences. AJR Am J Roentgenol 164:443–447, 1995.

3. Giacobetti FB, Vaccaro AR, Bos-Giacobetti MA, Deeley DM,Albert TJ, Farmer JC, Cotler JM: Vertebral artery occlusion asso-ciated with cervical spine trauma: A prospective analysis. Spine22:188–192, 1997.

4. Gurdjian ES, Hardy WG, Lindner DW, Thomas LM: Closed cer-vical trauma associated with involvement of carotid and vertebralarteries. J Neurosurg 20:418–427, 1963.

5. Louw JA, Mafoyane NA, Small B, Neser CP: Occlusion of thevertebral artery in cervical spine dislocations. J Bone Joint SurgBr 72B:679–681, 1990.

6. Prabhu V, Kizer J, Patil A, Hellbusch L, Taylon C, Leibrock L:Vertebrobasilar thrombosis associated with nonpenetrating cervi-cal spine trauma. J Trauma 40:130–137, 1996.

7. Schellinger PD, Schwab S, Krieger D, Fiebach JB, Steiner T, HundEF, Hacke W, Meinck HM: Masking of vertebral artery dissectionby severe trauma to the cervical spine. Spine 26:314–319, 2001.

8. Schneider RC, Crosby EC, Russo RH, Gosch HH: Traumatic spi-nal cord syndromes and their management. Clin Neurosurg 20:424–492, 1973.

9. Schwarz N, Buchinger W, Gaudernak T, Russe F, Zechner W:Injuries to the cervical spine causing vertebral artery trauma: Casereports. J Trauma 31:127–133, 1991.

10. Sim E, Schwarz N, Biowski-Fasching I, Biowski P: Color-codedDuplex sonography of vertebral arteries: 11 cases of blunt cervicalspine injury. Acta Orthop Scand 64:133–137, 1993.

11. Sim E, Vaccaro AR, Berzlanovich A, Pienaar S: The effects ofstaged static cervical flexion-distraction deformities on the pa-tency of the vertebral arterial vasculature. Spine 25:2180–2186,2000.

12. Song WS, Chiang YH, Chen CY, Lin SZ, Liu MY: A simplemethod for diagnosing traumatic occlusion of the vertebral arteryat the craniovertebral junction. Spine 19:837–839, 1994.

13. Suechting RL, French LA: Posterior inferior cerebellar artery syn-drome following a fracture of the cervical vertebra. J Neurosurg12:187–189, 1955.

14. Thibodeaux LC, Hearn AT, Peschiera JL, Deshmukh RM,Kerlakian GM, Welling RE, Nyswonger GD: Extracranial verte-bral artery dissection after trauma: A five year review. Br J Surg84:94, 1997.

15. Tulyapronchote R, Selhorst JB, Malkoff MD, Gomez CR: Delayedsequelae of vertebral artery dissection and occult cervical frac-tures. Neurology 44:1397–1399, 1994.

16. Vaccaro AR, Klein GR, Flanders AE, Albert TJ, Balderston RA,Cotler JM: Long-term evaluation of vertebral artery injuries fol-lowing cervical spine trauma using magnetic resonance angiog-raphy. Spine 23:789–795, 1998.

17. Weller SJ, Rossitch E Jr, Malek AM: Detection of vertebral arteryinjury after cervical spine trauma using magnetic resonance an-giography. J Trauma 46:660–666, 1999.

18. Willis BK, Greiner F, Orrison WW, Benzel EC: The incidence ofvertebral artery injury after midcervical spine fracture or sublux-ation. Neurosurgery 34:435–442, 1994.

19. Woodring JH, Lee C, Duncan V: Transverse process fractures ofthe cervical vertebrae: Are they insignificant? J Trauma 34:797–802, 1993.


BIBLIOGRAPHY

1. Acland RH, Anthony A, Inglis GS, Walton DI, Xiong X: Meth-ylprednisolone use in acute spinal cord injury. N Z Med J114:99, 2001 (letter).

2. Aebi M, Zuber K, Marchesi D: Treatment of cervical spineinjuries with anterior plating: Indications, techniques, and re-sults. Spine 16[Suppl 3]:S38–S45, 1991.

3. Ahuja A, Glasauer FE, Alker GJ Jr, Klein DM: Radiology insurvivors of traumatic atlanto-occipital dislocation. SurgNeurol 41:112–118, 1994.

4. Ajani AE, Cooper DJ, Scheinkestel CD, Laidlaw J, Tuxen DV:Optimal assessment of cervical spine trauma in critically illpatients: A prospective evaluation. Anaesth Intensive Care26:487–491, 1998.

5. Aldrich EF, Crow WN, Weber PB, Spagnolia TN: Use of MRimaging-compatible Halifax interlaminar clamps for posteriorcervical fusion. J Neurosurg 74:185–189, 1991.

6. Alexander R, Proctor H: Advanced Trauma Life Support Course forPhysicians. Chicago, American College of Surgeons, 1993, pp21–22.

7. Alexander E Jr, Davis CH Jr, Forsyth HF: Reduction and fusionof fracture dislocation of the cervical spine. J Neurosurg 27:588–591, 1967.

8. Allen BL Jr, Ferguson RL, Lehmann TR, O’Brien RP: A mech-anistic classification of closed, indirect fractures and disloca-tions of the lower cervical spine. Spine 7:1–27, 1982.

9. Amar AP, Levy ML: Pathogenesis and pharmacological strate-gies for mitigating secondary damage in acute spinal cordinjury. Neurosurgery 44:1027–1040, 1999.

10. American Academy of Orthopedic Surgeons, Committee onInjuries: Emergency Care and Transportation of the Sick and Injured.Chicago, American Academy of Orthopedic Surgeons, 1971, pp111–115.

11. American College of Surgeons: Advanced Trauma Life Support.Chicago, American College of Surgeons, 1993, p 201.

12. American College of Surgeons: Advanced Trauma Life Support.Chicago, American College of Surgeons, 1993, pp 214–218.

13. American College of Surgeons: Spine and spinal cord trauma,in Advanced Trauma Life Support for Doctors: Student CourseManual. Chicago, American College of Surgeons, 1997, ed 6, pp215–242.

14. American Spinal Injury Association: Standards for NeurologicalClassification of Spinal Injury Patients. Chicago, American SpinalInjury Association, 1984.

15. American Spinal Injury Association: Standards for NeurologicalClassification of Spinal Injury Patients. Chicago, American SpinalInjury Association, 1989.

16. American Spinal Injury Association/International Medical So-ciety of Paraplegia: Standards for Neurological and FunctionalClassification of Spinal Cord Injury. Chicago, American SpinalInjury Association, 1992, rev ed.

17. American Spinal Injury Association/International Medical So-ciety of Paraplegia: International Standards for Neurological andFunctional Classification of Spinal Cord Injury. Chicago, AmericanSpinal Injury Association, 1996, rev ed.

18. Anderson LD, D’Alonzo RT: Fractures of the odontoid processof the axis. J Bone Joint Surg Am 56A:1663–1674, 1974.

19. Anderson PA, Montesano PX: Morphology and treatment ofoccipital condyle fractures. Spine 13:731–736, 1988.

20. Anderson PA, Henley MB, Grady MS, Montesano PX, WinnHR: Posterior cervical arthrodesis with AO reconstructionplates and bone graft. Spine 16[Suppl 3]:S72–S79, 1991.


22. Apfelbaum RI, Lonser RR, Veres R, Casey A: Direct anteriorscrew fixation for recent and remote odontoid fractures.J Neurosurg 93[Suppl 2]:227–236, 2000.

23. Apostolides PJ, Theodore N, Karahalios DG, Sonntag VKH:Triple anterior screw fixation of an acute combination atlas-axisfracture: Case report. J Neurosurg 87:96–99, 1997.

24. Argenson C, Lovet J, Sanouiller JL, de Peretti F: Traumaticrotatory displacement of the lower cervical spine. Spine 13:767–773, 1988.

25. Armitage JM, Pyne A, Williams SJ, Frankel H: Respiratoryproblems of air travel in patients with spinal cord injuries. BMJ300:1498–1499, 1990.

26. Augustine J: Spinal trauma, in Campbell JE (ed): Basic TraumaLife Support: Advanced Prehospital Care. Englewood Cliffs,Prentice-Hall, 1988, ed 2.

27. Augustine J: Spinal trauma, in Campbell JE (ed): Basic TraumaLife Support for Paramedics and Advanced EMS Providers. UpperSaddle River, Brady/Prentice Hall Health, 1998, ed 3.

28. Bachulis BL, Long WB, Hynes GD, Johnson MC: Clinical indi-cations for cervical spine radiographs in the traumatized pa-tient. Am J Surg 153:473–478, 1987.

29. Balshi JD, Cantelmo NL, Menzoian JO: Complications of cavalinterruption by Greenfield filter in quadriplegics. J Vasc Surg9:558–562, 1989.

30. Banit DM, Grau G, Fisher JR: Evaluation of the acute cervicalspine: A management algorithm. J Trauma 49:450–456, 2000.

31. Banna M, Stevenson GW, Tumiel A: Unilateral atlanto-occipitaldislocation complicating an anomaly of the atlas: A case report.J Bone Joint Surg Am 65A:685–687, 1983.

32. Barboriak JJ, Rooney CB, El Ghatit AZ, Spuda K, Anderson AJ:Nutrition in spinal cord injury patients. J Am Paraplegia Soc6:32–36, 1983.

33. Barker EG Jr, Krumpelman J, Long JM: Isolated fracture of themedial portion of the lateral mass of the atlas: A previouslyundescribed entity. AJR Am J Roentgenol 126:1053–1058, 1976.

34. Barney R, Cordell W: Pain associated with immobilization onrigid spine boards. Ann Emerg Med 18:918, 1989 (abstr).

35. Barros TE, Bohlman HH, Capen DA, Cotler J, Dons K, Biering-Sorensen F, Marchesi DG, Zigler JE: Traumatic spondylolisthesisof the axis: Analysis of management. Spinal Cord 37:166–171,1999.

36. Bauer D, Kowalski R: Effect of spinal immobilization devices onpulmonary function in the healthy, nonsmoking man. AnnEmerg Med 17:915–918, 1988.

37. Baum JA, Hanley EN Jr, Pullekines J: Comparison of halocomplications in adults and children. Spine 14:251–252, 1989.

38. Bayless P, Ray VG: Incidence of cervical spine injuries in associa-tion with blunt head trauma. Am J Emerg Med 7:139–142, 1989.

39. Beatson TR: Fractures and dislocation of the cervical spine.J Bone Joint Surg Br 45B:21–35, 1963.

40. Becker DM, Gonzalez M, Gentili A, Eismont F, Green BA:Prevention of deep venous thrombosis in patients with acutespinal cord injuries: Use of rotating treatment tables. Neuro-surgery 20:675–677, 1987.


41. Bednar DA, Parikh J, Hummel J: Management of type II odon-toid process fractures in geriatric patients: A prospective studyof sequential cohorts with attention to survivorship. J SpinalDisord 8:166–169, 1995.

42. Bednarczyk JH, Sanderson DJ: Comparison of functional andmedical assessment in the classification of persons with spinalcord injury. J Rehabil Res Dev 30:405–411, 1993.

43. Bell C: Surgical observations. Middlesex Hosp J 4:469, 1817.44. Belzberg AJ, Tranmer BI: Stabilization of traumatic atlanto-

occipital dislocation: Case report. J Neurosurg 75:478–482,1991.

45. Benzel EC, Kesterson L: Posterior cervical interspinous com-pression wiring and fusion for mid to low cervical spinal inju-ries. J Neurosurg 70:893–899, 1989.

46. Benzel EC, Larson SJ: Functional recovery after decompressivespine operation for cervical spine fractures. Neurosurgery 20:742–746, 1987.

47. Benzel EC, Hadden TA, Saulsbery CM: A comparison of theMinerva and halo jackets for stabilization of the cervical spine.J Neurosurg 70:411–414, 1989.

48. Benzel EC, Hart BL, Ball PA, Baldwin NG, Orrison WW,Espinosa MC: Fractures of the C-2 vertebral body. J Neurosurg81:206–212, 1994.

49. Benzel EC, Hart BL, Ball PA, Baldwin NG, Orrison WW,Espinosa MC: Magnetic resonance imaging for the evaluationof patients with occult cervical spine injury. J Neurosurg 85:824–829, 1996.


51. Berne JD, Velmahos GC, El-Tawil Q, Demetriades D, AsensioJA, Murray JA, Cornwell EE, Belzberg H, Berne TV: Value ofcomplete cervical helical computed tomographic scanning inidentifying cervical spine injury in the unevaluable blunttrauma patient with multiple injuries: A prospective study.J Trauma 47:896–903, 1999.

52. Bettini N, Malaguti MC, Sintini M, Monti C: Fractures of theoccipital condyles: Report of four cases and review of theliterature. Skeletal Radiol 22:187–190, 1993.

53. Beyer CA, Cabanela ME, Berquist TH: Unilateral facet disloca-tions and fracture-dislocations of the cervical spine. J BoneJoint Surg Br 73B:977–981, 1991.

54. Biffl WL, Moore EE, Elliott JP, Ray C, Offner PJ, Franciose RJ,Brega KE, Burch JM: The devastating potential of blunt verte-bral arterial injuries. Ann Surg 231:672–681, 2000.

54a. Birney TJ, Hanley EN: Traumatic cervical spine injuries inchildhood and adolescence. Spine 14:1277–1282, 1989.

55. Black CA, Buderer NM, Blaylock B, Hogan BJ: Comparativestudy of risk factors for skin breakdown with cervical orthoticdevices: Philadelphia and Aspen. J Trauma Nurs 5:62–66, 1998.

56. Blackmore CC, Emerson SS, Mann FA, Koepsell TD: Cervicalspine imaging in patients with trauma: Determination of frac-ture risk to optimize use. Radiology 211:759–765, 1999.

57. Blackmore CC, Ramsey SD, Mann FA, Deyo RA: Cervical spinescreening with CT in trauma patients: A cost-effectiveness anal-ysis. Radiology 212:117–125, 1999.

58. Blaylock B: Solving the problem of pressure ulcers resultingfrom cervical collars. Ostomy Wound Manage 42:26–33, 1996.

59. Bloom AI, Neeman Z, Floman Y, Gomori J, Bar-Ziv J: Occipitalcondyle fracture and ligament injury: Imaging by CT. PediatrRadiol 26:786–790, 1996.

60. Bloom AI, Neeman Z, Slasky BS, Floman Y, Milgrom M,Rivkind A, Bar-Ziv J: Fracture of the occipital condyles andassociated craniocervical ligament injury: Incidence, CT imag-ing and implications. Clin Radiol 52:198–202, 1997.

61. Bohay D, Gosselin RA, Contreras DM: The vertical axis frac-ture: A report on three cases. J Orthop Trauma 6:416–419, 1992.

62. Bohler J: An approach to non-union of fractures. Surg Annu14:299–315, 1982.


64. Bohn D, Armstrong D, Becker L, Humphreys R: Cervical spineinjuries in children. J Trauma 30:463–469, 1990.

65. Bolender N, Cromwell LD, Wendling L: Fracture of the occip-ital condyle. AJR Am J Roentgenol 131:729–731, 1978.

66. Bools JC, Rose BS: Traumatic atlantooccipital dislocation: Twocases with survival. AJNR Am J Neuroradiol 7:901–904, 1986.

67. Borne GM, Bedou GL, Pinaudeau M: Treatment of pedicularfractures of the axis: A clinical study and screw fixation tech-nique. J Neurosurg 60:88–93, 1984.

68. Borock EC, Gabram SG, Jacobs LM, Murphy MA: A prospectiveanalysis of a two-year experience using computed tomographyas an adjunct for cervical spine clearance. J Trauma 31:1001–1006, 1991.

69. Bosch A, Stauffer ES, Nickel VL: Incomplete traumatic quadri-plegia: A ten-year review. JAMA 216:473–478, 1971.


71. Botel U, Gläser E, Niedeggen A: The surgical treatment of acutespinal paralysed patients. Spinal Cord 35:420–428, 1997.

72. Botsford DJ, Esses SI: A new scale for the clinical assessment ofspinal cord function. Orthopedics 15:1309–1313, 1992.

73. Boyd CR, Corse KM, Campbell RC: Emergency interhospitaltransport of the major trauma patient: Air versus ground.J Trauma 29:789–794, 1989.

74. Bozboga M, Unal F, Hepgul K, Izgi N, Turantan MI, Turker K:Fracture of the occipital condyle: Case report. Spine 17:1119–1121, 1992.

75. Bracken MB: National Acute Spinal Cord Injury Study of meth-ylprednisolone or naloxone. Neurosurgery 28:628–629, 1991(letter).

76. Bracken MB: Methylprednisolone and spinal cord injury.J Neurosurg 93:175–177, 2000 (letter).

77. Bracken MB: The use of methylprednisolone. J Neurosurg 93:340–341, 2000 (letter).

78. Bracken MB: High dose methylprednisolone must be given for24 or 48 hours after acute spinal cord injury. BMJ 322:862–863,2001 (letter).

78a. Bracken MB: Methylprednisolone and acute spinal cord injury:An update of the randomized evidence. Spine 26(24 Suppl):S47–S54, 2001.

79. Bracken MB: Pharmacological interventions for acute spinalcord injury. Cochrane Database Syst Rev 1:1–32, 2001.

80. Bracken MB, Holford TR: Effects of timing of methylpred-nisolone or naloxone administration on recovery of segmentaland long-tract neurological function in NASCIS 2. J Neurosurg79:500–507, 1993.

81. Bracken MB, Holford TR: Response: Treatment of spinal cordinjury. J Neurosurg 80:954–955, 1994 (letter).



82. Bracken MB, Aldrich EF, Herr DL, Hitchon PW, Holford TR,Marshall LF, Nockels RP, Pascale V, Shepard MJ, Sonntag VKH,Winn HR, Young W: Clinical measurement, statistical analysis,and risk-benefit: Controversies from trials of spinal injury.J Trauma 48:558–561, 2000.

83. Bracken MB, Collins WF, Freeman DF, Shepard MJ, WagnerFW, Silten RM, Hellenbrand KG, Ransohoff J, Hunt WE, PerotPL Jr, Grossman RG, Green BA, Eisenberg HM, Rifkinson N,Goodman JH, Meagher JN, Fischer B, Clifton GL, Flamm ES,Rawe SE: Efficacy of methylprednisolone in acute spinal cordinjury. JAMA 251:45–52, 1984.

84. Bracken MB, Shepard MJ, Collins WF, Holford TR, Baskin DS,Flamm E, Eisenberg HM, Leo-Summers L, Maroon JC, MarshallLF, Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC Jr,Wilberger JL, Winn HR, Young W: Response: Methylpred-nisolone for spinal cord injury. J Neurosurg 77:325–327, 1992(letter).

85. Bracken MB, Shepard MJ, Collins WF Jr, Holford TR, BaskinDS, Eisenberg HM, Flamm E, Leo-Summers L, Maroon JC,Marshall LF, Perot PL Jr, Piepmeier J, Sonntag VKH, WagnerFC Jr, Wilberger JL, Winn HR, Young W: Methylprednisoloneor naloxone treatment after acute spinal cord injury: 1-yearfollow-up data—Results of the Second National Acute SpinalCord Injury Study. J Neurosurg 76:23–31, 1992.

86. Bracken MB, Shepard MJ, Collins WF, Holford TR, Young W,Baskin DS, Eisenberg HM, Flamm E, Leo-Summers L, MaroonJ, Marshall LF, Perot PL Jr, Piepmeier J, Sonntag VKH, WagnerFC, Wilberger JE, Winn HR: A randomized, controlled trial ofmethylprednisolone or naloxone in the treatment of acutespinal-cord injury: Results of the Second National Acute SpinalCord Injury Study. N Engl J Med 322:1405–1411, 1990.

87. Bracken MB, Shepard MJ, Hellenbrand KG, Collins WF, Leo LS,Freeman DF, Wagner FC, Flamm ES, Eisenberg HM, GoodmanJH, Perot PL Jr, Green BA, Grossman RG, Meagher JN, YoungW, Fischer B, Clifton GL, Hunt WE, Rifkinson N: Methylpred-nisolone and neurological function 1 year after spinal cordinjury: Results of the National Acute Spinal Cord Injury Study.J Neurosurg 63:704–713, 1985.

88. Bracken MB, Shepard MJ, Holford TR, Leo-Summers L, AldrichEF, Fazl M, Fehlings MG, Herr DL, Hitchon PW, Marshall LF,Nockels RP, Pascale V, Perot PL Jr, Piepmeier JM, SonntagVKH, Wagner F, Wilberger JE, Winn HR, Young W: Adminis-tration of methylprednisolone for 24 or 48 hours or tirilazadmesylate for 48 hours in the treatment of acute spinal cordinjury: Results of the Third National Acute Spinal Cord InjuryRandomized Controlled Trial—National Acute Spinal Cord In-jury Study. JAMA 277:1597–1604, 1997.

89. Bracken MB, Shepard MJ, Holford TR, Leo-Summers L, AldrichEF, Fazl M, Fehlings MG, Herr DL, Hitchon PW, Marshall LF,Nockels RP, Pascale V, Perot PL Jr, Piepmeier JM, SonntagVKH, Wagner F, Wilberger JE, Winn HR, Young W: Methyl-prednisolone or tirilazad mesylate administration after acutespinal cord injury: 1-year follow up—Results of the Third Na-tional Acute Spinal Cord Injury randomized controlled trial.J Neurosurg 89:699–706, 1998.

90. Bracken MB, Webb SB Jr, Wagner FC: Classification of theseverity of acute spinal cord injury: Implications for manage-ment. Paraplegia 15:319–326, 1978.

91. Brady WJ, Moghtader J, Cutcher D, Exline C, Young J: ED useof flexion-extension cervical spine radiography in the evalua-tion of blunt trauma. Am J Emerg Med 17:504–508, 1999.

92. Brain Trauma Foundation, AANS: Management and prognosisof severe traumatic brain injury: Part I—Guidelines for themanagement of severe traumatic brain injury. J Neurotrauma17:449–554, 2000.

93. Brashear R Jr, Venters G, Preston ET: Fractures of the neuralarch of the axis: A report of twenty-nine cases. J Bone JointSurg Am 57A:879–887, 1975.

94. Bridgman SA, McNab W: Traumatic occipital condyle fracture,multiple cranial nerve palsies, and torticollis: A case report andreview of the literature. Surg Neurol 38:152–156, 1992.

95. Bridle MJ, Lynch KB, Quesenberry CM: Long-term functionfollowing the central cord syndrome. Paraplegia 28:178–185,1990.

96. Brockmeyer DL, York JE, Apfelbaum RI: Anatomical suitabilityof C1-2 transarticular screw placement in pediatric patients.J Neurosurg 92[Suppl 1]:7–11, 2000.

97. Brodkey JS, Miller CF Jr, Harmody RM: The syndrome of acutecentral cervical spinal cord injury revisited. Surg Neurol 14:251–257, 1980.

98. Brohi K, Wilson-Macdonald J: Evaluation of unstable cervicalspine injury: A 6-year experience. J Trauma 49:76–80, 2000.

99. Brooks AL, Jenkins EB: Atlanto-axial arthrodesis by the wedgecompression method. J Bone Joint Surg Am 60A:279–284, 1978.

100. Brown LH, Gough JE, Simonds WB: Can EMS providers ade-quately assess trauma patients for cervical spinal injury?Prehosp Emerg Care 2:33–36, 1998.

101. Bruce D: Spinal cord injury without radiographic abnormalityin children. Pediatr Neurosci 15:175, 1989 (comment).

102. Brunette DD, Rockswold GL: Neurologic recovery followingrapid spinal realignment for complete cervical spinal cord in-jury. J Trauma 27:445–447, 1987.

103. Bucci MN, Dauser RC, Maynard FA, Hoff JT: Management ofpost-traumatic cervical spine instability: Operative fusion ver-sus halo vest immobilization—Analysis of 49 cases. J Trauma28:1001–1006, 1988.

104. Bucholz RW: Unstable hangman’s fractures. Clin Orthop 154:119–124, 1981.

105. Bucholz RD, Cheung KC: Halo vest versus spinal fusion forcervical injury: Evidence from an outcome study. J Neurosurg70:884–892, 1989.

106. Buhs C, Cullen M, Klein M, Farmer D: The pediatric traumaC-spine: Is the “odontoid” view necessary? J Pediatr Surg35:994–997, 2000.

107. Bulas DI, Fitz CR, Johnson DL: Traumatic atlanto-occipital dis-location in children. Radiology 188:155–158, 1993.

107a. Bullock R, Chesnut RM, Clifton GL, Ghajar J, Marion DW,Narayan RK, Newell DW, Pitts LH, Rosner MJ, Walters BC,Wilberger JE: Guidelines for the management of severe trau-matic brain injury: Intracranial pressure treatment threshold.J Neurotrauma 17:493–495, 2000.

108. Bullock R, Chesnut RM, Clifton G, Ghajar J, Marion DW,Narayan RK, Newell DW, Pitts LH, Rosner MJ, Wilberger JW:Guidelines for the management of severe head injury: BrainTrauma Foundation. Eur J Emerg Med 3:109–127, 1996.

109. Bundschuh CV, Alley JB, Ross M, Porter IS, Gudeman SK:Magnetic resonance imaging of suspected atlanto-occipital dis-location: Two case reports. Spine 17:245–248, 1992.

110. Burke DC, Berryman D: The place of closed manipulation in themanagement of flexion-rotation dislocations of the cervicalspine. J Bone Joint Surg Br 53B:165–182, 1971.

111. Burke JT, Harris JH Jr: Acute injuries of the axis vertebra.Skeletal Radiol 18:335–346, 1989.

Bibliography S181


112. Burke DC, Tiong TS: Stability of the cervical spine after conser-vative treatment. Paraplegia 13:191–202, 1975.


114. Burns GA, Cohn SM, Frumento RJ, Degutis LC, Hammers L:Prospective ultrasound evaluation of venous thrombosis inhigh-risk trauma patients. J Trauma 35:405–408, 1993.

115. Butman A, Vomacka R: Part 1: Spine immobilization. Emer-gency 23:48–51, 1991.

116. Cabanela ME, Ebersold MJ: Anterior plate stabilization forbursting teardrop fractures of the cervical spine. Spine 13:888–891, 1988.

117. Cahill DW, Bellegarrigue R, Ducker TB: Bilateral facet to spi-nous process fusion: A new technique for posterior spinal fu-sion after trauma. Neurosurgery 13:1–4, 1983.

118. Campanelli M, Kattner KA, Stroink A, Gupta K, West S: Pos-terior C1-C2 transarticular screw fixation in the treatment ofdisplaced type II odontoid fractures in the geriatric population:Review of seven cases. Surg Neurol 51:596–601, 1999.

119. Capaul M, Zollinger H, Satz N, Dietz V, Lehmann D, Schurch B:Analyses of 94 consecutive spinal cord injury patients usingASIA definition and modified Frankel score classification. Para-plegia 32:583–587, 1994.

120. Carey ME, Nance FC, Kirgis HD, Young HF, Megison LC Jr,Kline DG: Pancreatitis following spinal cord injury.J Neurosurg 47:917–922, 1977.

121. Carter VM, Fasen JA, Roman JM Jr, Hayes KW, Petersen CM:The effect of a soft collar, used as normally recommended orreversed, on three planes of cervical range of motion. J OrthopSports Phys Ther 23:209–215, 1996.

122. Casas ER, Sanchez MP, Arias CR, Masip JP: Prophylaxis ofvenous thrombosis and pulmonary embolism in patients withacute traumatic spinal cord lesions. Paraplegia 15:209–214,1977.

123. Castillo M, Mukherji SK: Vertical fractures of the dens. AJNRAm J Neuroradiol 17:1627–1630, 1996.

124. Castling B, Hicks K: Traumatic isolated unilateral hypoglossalnerve palsy: Case report and review of the literature. Br J OralMaxillofac Surg 33:171–173, 1995.

125. Cattell HS, Filtzer DL: Pseudosubluxation and other normalvariations in the cervical spine in children: A study of onehundred and sixty children. J Bone Joint Surg Am 47A:1295–1309, 1965.

126. Catz A, Itzkovich M, Agranov E, Ring H, Tamir A: SCIM:Spinal cord independence measure—A new disability scale forpatients with spinal cord lesions. Spinal Cord 35:850–856, 1997.

127. Chan D, Goldberg RM, Mason J, Chan L: Backboard versusmattress splint immobilization: A comparison of symptomsgenerated. J Emerg Med 14:293–298, 1996.

128. Chan D, Goldberg R, Tascone A, Harmon S, Chan L: The effectof spinal immobilization on healthy volunteers. Ann EmergMed 23:48–51, 1994.

129. Chan RC, Schweigel JF, Thompson GB: Halo-thoracic braceimmobilization in 188 patients with acute cervical spine inju-ries. J Neurosurg 58:508–515, 1983.

130. Chandler DR, Nemejc C, Adkins RH, Waters RL: Emergencycervical-spine immobilization. Ann Emerg Med 21:1185–1188,1992.

131. Chee SG: Review of the role of magnetic resonance imaging inacute cervical spine injuries. Ann Acad Med Singapore 22:757–761, 1993.

132. Chehrazi B, Wagner FC Jr, Collins WF Jr, Freeman DH Jr: Ascale for evaluation of spinal cord injury. J Neurosurg 54:310–315, 1981.

132a. Chen D, Apple DF, Hudson LM, Bode R: Medical complicationsduring acute rehabilitation following spinal cord injury: Cur-rent experience of the Model Systems. Arch Phys Med Rehabil80:1397–1401, 1999.

133. Chen TY, Dickman CA, Eleraky M, Sonntag VKH: The role ofdecompression for acute incomplete cervical spinal cord injuryin cervical spondylosis. Spine 23:2398–2403, 1998.

134. Chen TY, Lee ST, Lui TN, Wong CW, Yeh YS, Tzaan WC, HungSY: Efficacy of surgical treatment in traumatic central cordsyndrome. Surg Neurol 48:435–441, 1997.

135. Cheshire DJ: The stability of the cervical spine following theconservative treatment of fractures and fracture-dislocations.Paraplegia 7:193–203, 1969.

136. Cheshire DJ: A classification of the functional end-results ofinjury to the cervical spinal cord. Paraplegia 8:70–73, 1970.

137. Chesnut RM, Marshall LF, Klauber MR, Blunt BA, Baldwin N,Eisenberg HM, Jane JA, Marmarou A, Foulkes MA: The role ofsecondary brain injury in determining outcome from severehead injury. J Trauma 34:216–222, 1993.

138. Chiba K, Fujimura Y, Toyama Y, Fujii E, Nakanishi T,Hirabayashi K: Treatment protocol for fractures of the odontoidprocess. J Spinal Disord 9:267–276, 1996.

139. Chu DA, Ahn JH, Ragnarsson KT, Helt J, Folcarelli P, RamirezA: Deep venous thrombosis: Diagnosis in spinal cord injuredpatients. Arch Phys Med Rehabil 66:365–368, 1985.

140. Clancy MJ: Clearing the cervical spine of adult victims oftrauma. J Accid Emerg Med 16:208–214, 1999.

141. Clark CR, White AA III: Fractures of the dens: A multicenterstudy. J Bone Joint Surg Am 67A:1340–1348, 1985.

142. Clarke KS: Caloric costs of activity in paraplegic persons. ArchPhys Med Rehabil 47:427–435, 1966.

143. Clements WD, Mezue W, Mathew B: Os odontoideum: Congenital oracquired?—That’s not the question. Injury 26:640–642, 1995.

144. Cline JR, Scheidel E, Bigsby EF: A comparison of methods ofcervical immobilization used in patient extrication and trans-port. J Trauma 25:649–653, 1985.

145. Cloward RB: Reduction of traumatic dislocation of the cervicalspine with locked facets: Technical note. J Neurosurg 38:527–531, 1973.

146. Cohen J: A coefficient of agreement for nominal scales. EducPsychol Meas 20:37–46, 1960.

147. Cohen A, Bosshard R, Yeo JD: A new device for the care ofacute spinal injuries: The Russell Extrication Device (RED).Paraplegia 28:151–157, 1990.

148. Cohen ME, Ditunno JF Jr, Donovan WH, Maynard FM Jr: A testof the 1992 International Standards for Neurological and Func-tional Classification of Spinal Cord Injury. Spinal Cord 36:554–560, 1998.

149. Cohen ME, Sheehan TP, Herbison GJ: Content validity andreliability of the International Standards for Neurological Clas-sification of Spinal Cord Injury. Top Spinal Cord Inj Rehabil1:15–31, 1996.

150. Cohn SM, Lyle WG, Linden CH, Lancey RA: Exclusion of cervicalspine injury: A prospective study. J Trauma 31:570–574, 1991.

151. Coleman WP, Benzel D, Cahill DW, Ducker T, Geisler F, GreenB, Gropper MR, Goffin J, Madsen PW III, Maiman DJ, OndraSL, Rosner M, Sasso RC, Trost GR, Zeidman S: A critical ap-praisal of the reporting of the National Acute Spinal CordInjury Studies (II and III) of methylprednisolone in acute spinalcord injury. J Spinal Disord 13:185–199, 2000.



152. Collalto PM, DeMuth WW, Schwentker EP, Boal DK: Traumaticatlanto-occipital dislocation: Case report. J Bone Joint Surg Am68A:1106–1109, 1986.

153. Colnet G, Chabannes J, Commun C, Rigal MC, Alassaf M:Atlanto-occipital luxation and syringomyelia: 2 rare complica-tions of cervical injury—Diagnostic and therapeutic effects:Apropos of a case [in French]. Neurochirurgie 35:58–63, 1989.

154. Cone DC, Wydro GC, Mininger CM: Current practice in clinicalcervical spinal clearance: Implication for EMS. Prehosp EmergCare 3:42–46, 1999.

155. Consortium for Spinal Cord Medicine: Prevention of thrombo-embolism in spinal cord injury. J Spinal Cord Med 20:259–283,1997.

156. Contostavlos DL: Massive subarachnoid hemorrhage due tolaceration of the vertebral artery associated with fracture of thetransverse process of the atlas. J Forensic Sci 16:40–56, 1971.

157. Cooke M: Spinal boards. J Accid Emerg Med 13:433, 1996(letter).

158. Cooper PR, Maravilla KR, Sklar FH, Moody SF, Clark WK: Haloimmobilization of cervical spine fractures: Indications and re-sults. J Neurosurg 50:603–610, 1979.

159. Coric D, Wilson JA, Kelly DL Jr: Treatment of traumatic spon-dylolisthesis of the axis with nonrigid immobilization: A re-view of 64 cases. J Neurosurg 85:550–554, 1996.

160. Cotler HB, Cotler JM, Alden ME, Sparks G, Biggs CA: Themedical and economic impact of closed cervical spine disloca-tions. Spine 15:448–452, 1990.

161. Cotler JM, Herbison GJ, Nasuti JF, Ditunno JF Jr, An H, WolffBE: Closed reduction of traumatic cervical spine dislocationusing traction weights up to 140 pounds. Spine 18:386–390,1993.

162. Cotler HB, Miller LS, DeLucia FA, Cotler JM, Davne SH: Closedreduction of cervical spine dislocations. Clin Orthop 214:185–199, 1987.

163. Cottalorda J, Allard D, Dutour N: Fracture of the occipitalcondyle. J Pediatr Orthop B 5:61–63, 1996.

164. Cox SA, Weiss SM, Posuniak EA, Worthington P, Prioleau M,Heffley G: Energy expenditure after spinal cord injury: Anevaluation of stable rehabilitating patients. J Trauma 25:419–423, 1985.

165. Coyne TJ, Fehlings MG, Wallace MC, Bernstein M, Tator CH:C1-C2 posterior cervical fusion: Long-term evaluation of resultsand efficacy. Neurosurgery 37:688–693, 1995.

166. Craig JB, Hodgson BF: Superior facet fractures of the axisvertebra. Spine 16:875–877, 1991.

167. Cruse JM, Lewis RE, Dilioglou S, Roe DL, Wallace WF, ChenRS: Review of immune function, healing of pressure ulcers, andnutritional status in patients with spinal cord injury. J SpinalCord Med 23:129–135, 2000.

168. Cruse JM, Lewis RE, Roe DL, Dilioglou S, Blaine MC, WallaceWF, Chen RS: Facilitation of immune function, healing of pres-sure ulcers, and nutritional status in spinal cord injury patients.Exp Mol Pathol 68:38–54, 2000.

169. Crutchfield W: Skeletal traction in treatment of injuries to thecervical spine. JAMA 155:29–32, 1954.


171. Curri D, Cervellini P, Zanusso M, Benedetti A: Isolated fractureof occipital condyle: Case report. J Neurosurg Sci 32:157–159,1988.

172. Cybulski GR, Douglas RA, Meyer PR Jr, Rovin RA: Complica-tions in three-column cervical spine injuries requiring anterior-posterior stabilization. Spine 17:253–256, 1992.

173. Dai L, Jia L: Central cord injury complicating acute cervical discherniation in trauma. Spine 25:331–336, 2000.

174. Dai L, Yuan W, Ni B, Jia L: Os odontoideum: Etiology, diagno-sis, and management. Surg Neurol 53:106–109, 2000.

175. Dai LY, Yuan W, Ni B, Liu HK, Jia LS, Zhao DL, Xu YK:Surgical treatment of nonunited fractures of the odontoid pro-cess, with special reference to occipitocervical fusion forunreducible atlantoaxial subluxation or instability. Eur Spine J9:118–122, 2000.

176. D’Alise MD, Benzel EC, Hart BL: Magnetic resonance imagingevaluation of the cervical spine in the comatose or obtundedtrauma patient. J Neurosurg 91[Suppl 1]:54–59, 1999.

177. Davies G, Deakin C, Wilson A: The effect of a rigid collar onintracranial pressure. Injury 27:647–649, 1996.

178. Davis D, Bohlman H, Walker AE, Fisher R, Robinson R: Thepathological findings in fatal craniospinal injuries. J Neurosurg34:603–613, 1971.

179. Davis JW, Parks SN, Detlefs CL, Williams GG, Williams JL,Smith RW: Clearing the cervical spine in obtunded patients:The use of dynamic fluoroscopy. J Trauma 39:435–438, 1995.

180. Davis JW, Phreaner DL, Hoyt DB, Mackersie RC: The etiologyof missed cervical spine injuries. J Trauma 34:342–346, 1993.

181. Davis PC, Reisner A, Hudgins PA, Davis WE, O’Brien MS:Spinal injuries in children: Role of MR. AJNR Am JNeuroradiol 14:607–617, 1993.

182. Davis LA, Warren SA, Reid DC, Oberle K, Saboe LA, GraceMG: Incomplete neural deficits in thoracolumbar and lumbarspine fractures: Reliability of Frankel and Sunnybrookscales. Spine 18:257–263, 1993.

183. Deeb ZL, Rothfus WE, Goldberg AL, Daffner RH: Occult oc-cipital condyle fractures presenting as tumors. J ComputTomogr 12:261–263, 1988.

184. Della Torre P, Rinonapoli E: Halo-cast treatment of fracturesand dislocations of the cervical spine. Int Orthop 16:227–231,1992.

185. De Lorenzo RA: A review of spinal immobilization techniques.J Emerg Med 14:603–613, 1996.

186. De Lorenzo RA, Olson JE, Boska M, Johnston R, Hamilton GC,Augustine J, Barton R: Optimal positioning for cervical immo-bilization. Ann Emerg Med 28:301–308, 1996.

187. Demisch S, Lindner A, Beck R, Zierz S: The forgotten condyle:Delayed hypoglossal nerve palsy caused by fracture of theoccipital condyle. Clin Neurol Neurosurg 100:44–45, 1998.

188. Desai SS, Coumas JM, Danylevich A, Hayes E, Dunn EJ: Frac-ture of the occipital condyle: Case report and review of theliterature. J Trauma 30:240–241, 1990.

189. De Smet L, Vercauteren M, Verdonk R, Claessens H: Severeacute cervical spine injuries: Conservative treatment. ActaOrthop Belg 50:512–520, 1984.

190. DeVivo MJ, Kartus PL, Stover SL, Rutt RD, Fine PR: Cause ofdeath for patients with spinal cord injuries. Arch Intern Med149:1761–1766, 1989.

191. Dibenedetto T, Lee CK: Traumatic atlanto-occipital instability:A case report with follow-up and a new diagnostic technique.Spine 15:595–597, 1990.

192. Dick T: Comparing the short-board technique. Ann Emerg Med18:115–116, 1989 (letter).

193. Dick T, Land R: Spinal immobilization devices: Part 1—Cervi-cal extrication collars. J Emerg Med Serv JEMS 7:26–32, 1982.

Bibliography S183


194. Dick T, Land R: Full spinal immobilizers. J Emerg Med ServJEMS 8:34–36, 1983.

195. Dickman CA, Sonntag VKH: Surgical management of atlanto-axial nonunions. J Neurosurg 83:248–253, 1995.

196. Dickman C, Sonntag VKH: Injuries involving the transverseatlantal ligament: Classification and treatment guidelines basedupon experience with 39 injuries. Neurosurgery 40:886–887,1997 (letter).

197. Dickman CA, Sonntag VKH: Posterior C1–C2 transarticularscrew fixation for atlantoaxial arthrodesis. Neurosurgery 43:275–281, 1998.

197a. Dickman CA, Foley KT, Sonntag VKH, Smith MM: Cannulatedscrews for odontoid fixation and atlantoaxial transarticularscrew fixation. J Neurosurg 83:1095–1100, 1995.

198. Dickman CA, Greene KA, Sonntag VKH: Injuries involving thetransverse atlantal ligament: Classification and treatmentguidelines based upon experience with 39 injuries. Neuro-surgery 38:44–50, 1996.

199. Deleted in proof.200. Dickman CA, Papadopoulos SM, Sonntag VKH, Spetzler RF,

Rekate HL, Drabier J: Traumatic occipitoatlantal dislocations.J Spinal Disord 6:300–313, 1993.

201. Dickman CA, Rekate HL, Sonntag VKH, Zabramski JM: Pedi-atric spinal trauma: Vertebral column and spinal cord injuriesin children. Pediatr Neurosci 15:237–256, 1989.

202. Dickman CA, Sonntag VKH, Papadopoulos SM, Hadley MN:The interspinous method of posterior atlantoaxial arthrodesis.J Neurosurg 74:190–198, 1991.

203. Ditunno JF Jr: Functional assessment measures in CNS trauma.J Neurotrauma 9[Suppl 1]:S301–S305, 1992.

204. Ditunno JF Jr: New spinal cord injury standards, 1992. Paraple-gia 30:90–91, 1992.

205. Ditunno JF Jr: American spinal injury standards for neurolog-ical and functional classification of spinal cord injury: Past,present and future—1992 Heiner Sell Lecture of the AmericanSpinal Injury Association. J Am Paraplegia Soc 17:7–11, 1994.

206. Ditunno JF Jr, Ditunno PL, Graziani V, Scivoletto G, Bernardi M,Castellano V, Marchetti M, Barbeau H, Frankel HL, D’AndreaGreve JM, Ko H Y, Marshall R, Nance P: Walking Index for SpinalCord Injury (WISCI): An international multicenter validity andreliability study. Spinal Cord 38:234–243, 2000.

207. Ditunno JF Jr, Young W, Donovan WH, Creasey G: The Inter-national Standards Booklet for Neurological and FunctionalClassification of Spinal Cord Injury: American Spinal InjuryAssociation. Paraplegia 32:70–80, 1994.

208. Dodd FM, Simon E, McKeown D, Patrick MR: The effect of acervical collar on the tidal volume of anaesthetised adult pa-tients. Anaesthesia 50:961–963, 1995.

209. Dodds TA, Martin DP, Stolov WC, Deyo RA: A validation ofthe functional independence measurement and its performanceamong rehabilitation inpatients. Arch Phys Med Rehabil 74:531–536, 1993.


211. Domeier RM: Indications for prehospital spinal immobilization:National Association of EMS Physicians Standards and ClinicalPractice Committee. Prehosp Emerg Care 3:251–253, 1999.

212. Domeier RM, Evans RW, Swor RA, Hancock JB, Fales W,Krohmer J, Frederiksen SM, Shork MA: The reliability ofprehospital clinical evaluation for potential spinal injury is notaffected by the mechanism of injury. Prehosp Emerg Care3:332–337, 1999.

213. Domeier RM, Evans RW, Swor RA, Rivera-Rivera EJ,Frederiksen SM: High-risk criteria for performing pre-hospitalspinal immobilization in trauma. Ann Emerg Med 25:141–142,1995.

214. Domeier RM, Evans RW, Swor RA, Rivera-Rivera EJ,Frederiksen SM: Prehospital clinical findings associated withspinal injury. Prehosp Emerg Care 1:11–15, 1997.

215. Domeier RM, Evans RW, Swor RA, Rivera-Rivera EJ,Frederiksen SM: Prospective validation of out-of-hospital spi-nal clearance criteria: A preliminary report. Acad Emerg Med4:643–646, 1997.

216. Donahue DJ, Muhlbauer MS, Kaufman RA, Warner WC,Sanford RA: Childhood survival of atlantooccipital dislocation:Underdiagnosis, recognition, treatment, and review of the lit-erature. Pediatr Neurosurg 21:105–111, 1994.

217. Donovan WH, Kopaniky D, Stolzmann E, Carter RE: The neu-rological and skeletal outcome in patients with closed cervicalspinal cord injury. J Neurosurg 66:690–694, 1987.

218. Doran SE, Papadopoulos SM, Ducker TB, Lillehei KO: Magneticresonance imaging documentation of coexistent traumaticlocked facets of the cervical spine and disc herniation.J Neurosurg 79:341–345, 1993.

219. Dormans JP, Criscitiello AA, Drummond DS, Davidson RS:Complications in children managed with immobilization in ahalo vest. J Bone Joint Surg Am 77A:1370–1373, 1995.

220. Dorr LD, Harvey JP Jr, Nickel VL: Clinical review of the earlystability of spine injuries. Spine 7:545–550, 1982.

221. Dublin AB, Marks WM, Weinstock D, Newton TH: Traumaticdislocation of the atlanto-occipital articulation (AOA) withshort-term survival: With a radiographic method of measuringthe AOA. J Neurosurg 52:541–546, 1980.

222. Ducker TB, Zeidman SM: Spinal cord injury: Role of steroidtherapy. Spine 19:2281–2287, 1994.

223. Ducker TB, Kindt GW, Kempf LG: Pathological findings inacute experimental spinal cord trauma. J Neurosurg 35:700–708, 1971.

224. Duh MS, Shepard MJ, Wilberger JE, Bracken MB: The effective-ness of surgery on the treatment of acute spinal cord injury andits relation to pharmacological treatment. Neurosurgery 35:240–249, 1994.

225. Dunn ME, Seljeskog EL: Experience in the management ofodontoid process injuries: An analysis of 128 cases. Neuro-surgery 18:306–310, 1986.

226. Dyck P: Os odontoideum in children: Neurological manifesta-tions and surgical management. Neurosurgery 2:93–99, 1978.

227. Effendi B, Roy D, Cornish B, Dussault RG, Laurin CA: Fracturesof the ring of the axis: A classification based on the analysis of131 cases. J Bone Joint Surg Br 63B:319–327, 1981.

228. Eismont FJ, Bohlman HH: Posterior atlanto-occipital disloca-tion with fractures of the atlas and odontoid process. J BoneJoint Surg Am 60A:397–399, 1978.

229. Eismont FJ, Arena MJ, Green BA: Extrusion of an intervertebraldisc associated with traumatic subluxation or dislocation ofcervical facets: Case report. J Bone Joint Surg Am 73A:1555–1560.


231. Eleraky MA, Theodore N, Adams M, Rekate HL, Sonntag VKH:Pediatric cervical spine injuries: Report of 102 cases and reviewof the literature. J Neurosurg 92[Suppl 1]:12–17, 2000.



232. El-Khoury GY, Clark CR, Gravett AW: Acute traumatic rota-tory atlanto-axial dislocation in children: A report of threecases. J Bone Joint Surg Am 66A:774–777, 1984.

233. Elliott JM Jr, Rogers LF, Wissinger JP, Lee JF: The hangman’sfracture: Fractures of the neural arch of the axis. Radiology104:303–307, 1972.

234. El Masri WS, Silver JR: Prophylactic anticoagulant therapy inpatients with spinal cord injury. Paraplegia 19:334–342, 1981.

235. El Masry WS, Tsubo M, Katoh S, El Miligui YH, Khan A:Validation of the American Spinal Injury Association (ASIA)motor score and the National Acute Spinal Cord Injury Study(NASCIS) motor score. Spine 21:614–619, 1996.

236. Emery SE, Pathria MN, Wilber RG, Masaryk T, Bohlman HH:Magnetic resonance imaging of posttraumatic spinal ligamentinjury. J Spinal Disord 2:229–233, 1989.

237. Emery E, Saillant G, Ismail M, Fohanno D, Roy-Camille R:Fracture of the occipital condyle: Case report and review of theliterature. Eur Spine J 4:191–193, 1995.

238. Ersmark H, Kalen R: A consecutive series of 64 halo-vest-treated cervical spine injuries. Arch Orthop Trauma Surg 105:243–246, 1986.

239. Ersmark H, Lowenhielm P: Factors influencing the outcome ofcervical spine injuries. J Trauma 28:407–410, 1988.

240. Esses SI, Bednar DA: Screw fixation of odontoid fractures andnonunions. Spine 16[Suppl 10]:S483–S485, 1991.

241. Evans D: Reduction of cervical dislocations. J Bone Joint SurgBr 43B:552–555, 1961.

242. Evans DL, Bethem D: Cervical spine injuries in children.J Pediatr Orthop 9:563–568, 1989.

243. Evarts CM: Traumatic occipito-atlantal dislocation. J Bone JointSurg Am 52:1653–1660, 1970.

244. Farley FA, Graziano GP, Hensinger RN: Traumatic atlanto-occipital dislocation in a child. Spine 17:1539–1541, 1992.

245. Farmer J, Vaccaro A, Albert TJ, Malone S, Balderston RA, CotlerJM: Neurologic deterioration after cervical spinal cord injury.J Spinal Disord 11:192–196, 1998.

246. Farrington JD: Death in a ditch. Bull Am Coll Surg 52:121–130,1967.

247. Farrington JD: Extrication of victims: Surgical principles.J Trauma 8:493–512, 1968.

248. Farthing JW: Atlantooccipital dislocation with survival: A casereport. N C Med J 9:34–36, 1948.

249. Fehlings MG, Tator CH: An evidence-based review of decom-pressive surgery in acute spinal cord injury: Rationale, indica-tions, and timing based on experimental and clinical studies.J Neurosurg 91[Suppl 1]:1–11, 1999.

250. Fehlings MG, Cooper PR, Errico TJ: Posterior plates in themanagement of cervical instability: Long-term results in 44patients. J Neurosurg 81:341–349, 1994.

251. Fehlings MG, Rao SC, Tator CH, Skaf G, Arnold P, Benzel EC,Dickman C, Cuddy B, Green B, Hitchon P, Northrup B, SonntagVKH, Wagner F, Wilberger JE: The optimal radiologic methodfor assessing spinal canal compromise and cord compression inpatients with cervical spinal cord injury: Part II—Results of amulticenter study. Spine 24:605–613, 1999.

252. Fenstermaker RA: Acute neurologic management of the patientwith spinal cord injury. Urol Clin North Am 20:413–421, 1993.

253. Ferrera PC, Bartfield JM: Traumatic atlanto-occipital dislocation: Apotentially survivable injury. Am J Emerg Med 14:291–296, 1996.

254. Field MJ, Lohr KN (eds): Clinical Practice Guidelines: Directionsfor a New Program—Committee to Advise the Public Health Serviceon Clinical Practice Guidelines: Institute of Medicine. WashingtonDC, National Academy Press, 1990.

255. Field-Fote EC, Fluet GG, Schafer SD, Schneider EM, Smith R,Downey PA, Ruhl CD: The Spinal Cord Injury FunctionalAmbulation Inventory (SCI-FAI). J Rehabil Med 33:177–181,2001.

256. Fielding JW, Hawkins RJ: Atlanto-axial rotatory fixation: Fixedrotatory subluxation of the atlanto-axial joint. J Bone Joint SurgAm 59A:37–44, 1977.

257. Fielding JW, Cochran GV, Lawsing JF III, Hohl M: Tears of thetransverse ligament of the atlas: A clinical and biomechanicalstudy. J Bone Joint Surg Am 56A:1683–1691, 1974.


259. Fielding JW, Hensinger RN, Hawkins RJ: Os odontoideum.J Bone Joint Surg Am 62A:376–383, 1980.

260. Finch GD, Barnes MJ: Major cervical spine injuries in childrenand adolescents. J Pediatr Orthop 18:811–814, 1998.

261. Flanders AE, Schaefer DM, Doan HT, Mishkin MM, GonzalezCF, Northrup BE: Acute cervical spine trauma: Correlation ofMR imaging findings with degree of neurologic deficit. Radi-ology 177:25–33, 1990.

262. Flohr H, Pöll W, Brock M: Regulation of spinal cord blood flow,in Russell RWR (ed): Brain and Blood Flow: Proceedings of the 4thInternational Symposium on the Regulation of Cerebral Blood Flow.London, Pitman Medical, 1971, pp 406–409.

263. Fotter R, Sorantin E, Schneider U, Ranner G, Fast C, Schober P:Ultrasound diagnosis of birth-related spinal cord trauma: Neo-natal diagnosis and follow-up and correlation with MRI.Pediatr Radiol 24:241–244, 1994.

264. Fowler JL, Sandhu A, Fraser RD: A review of fractures of theatlas vertebra. J Spinal Disord 3:19–24, 1990.

265. Francis WR, Fielding JW, Hawkins RJ, Pepin J, Hensinger R:Traumatic spondylolisthesis of the axis. J Bone Joint Surg Br63B:313–318, 1981.

266. Frankel HL, Hancock DO, Hyslop G, Melzak J, Michaelis LS,Ungar GH, Vernon JD, Walsh JJ: The value of postural reduc-tion in the initial management of closed injuries of the spinewith paraplegia and tetraplegia: Part I. Paraplegia 7:179–192,1969.

267. Frankel H, Michaelis L, Paeslack V, Ungar G, Walsh JJ: Closedinjuries of the cervical spine and spinal cord: Results of conser-vative treatment of vertical compression injuries of the cervicalspine. Proc Veterans Adm Spinal Cord Inj Conf 19:28–32,1973.

268. Frankenfield DC, Smith JS, Cooney RN: Accelerated nitrogenloss after traumatic injury is not attenuated by achievement ofenergy balance. JPEN J Parenter Enteral Nutr 21:324–329, 1997.

269. Freemyer B, Knopp R, Piche J, Wales L, Williams J: Comparisonof five-view and three-view cervical spine series in the evalu-ation of patients with cervical trauma. Ann Emerg Med 18:818–821, 1989.

270. French HG, Burke SW, Roberts JM, Johnston CE II, WhitecloudT, Edmunds JO: Upper cervical ossicles in Down syndrome.J Pediatr Orthop 7:69–71, 1987.

271. Friedman D, Flanders A, Thomas C, Millar W: Vertebral arteryinjury after acute cervical spine trauma: Rate of occurrence asdetected by MR angiography and assessment of clinical conse-quences. AJR Am J Roentgenol 164:443–449, 1995.

272. Frisbie JH, Sasahara AA: Low dose heparin prophylaxis fordeep venous thrombosis in acute spinal cord injury patients: Acontrolled study. Paraplegia 19:343–346, 1981.

Bibliography S185


273. Frisbie JH, Sharma GV: Pulmonary embolism manifesting asacute disturbances of behavior in patients with spinal cordinjury. Paraplegia 32:570–572, 1994.

274. Frohna WJ: Emergency department evaluation and treatment ofthe neck and cervical spine injuries. Emerg Med Clin NorthAm 17:739–791, 1999.

275. Fruin AH, Pirotte TP: Traumatic atlantooccipital dislocation:Case report. J Neurosurg 46:663–666, 1977.

276. Fujii E, Kobayashi K, Hirabayashi K: Treatment in fractures ofthe odontoid process. Spine 13:604–609, 1988.

277. Fujimura Y, Nishi Y, Chiba K, Kobayashi K: Prognosis of neu-rological deficits associated with upper cervical spine injuries.Paraplegia 33:195–202, 1995.

278. Fujimura Y, Nishi Y, Kobayashi K: Classification and treatmentof axis body fractures. J Orthop Trauma 10:536–540, 1996.

279. Gabrielsen TO, Maxwell JA: Traumatic atlanto-occipital dislo-cation: With case report of a patient who survived. Am JRoentgenol Radium Ther Nucl Med 97:624–629, 1966.

280. Galandiuk S, Raque G, Appel S, Polk HC Jr: The two-edgedsword of large-dose steroids for spinal cord trauma. Ann Surg218:419–427, 1993.

281. Garber J: Abnormalities of the atlas and axis vertebrae: Con-genital and traumatic. J Bone Joint Surg Am 46A:1782–1791,1964.

282. Garfin SR, Shackford SR, Marshall LF, Drummond JC: Care ofthe multiply injured patient with cervical spine injury. ClinOrthop 239:19–29, 1989.

283. Garth G: Proposal for the establishment of minimum perfor-mance specifications for cervical extrication collars. Presentedat the 14th Annual Meeting of the American Society for Testingand Materials Skeletal Support Committee, WestConshohocken, PA, 1988.

284. Gaskill SJ, Marlin AE: Custom fitted thermoplastic Minervajackets in the treatment of cervical spine instability in preschoolage children. Pediatr Neurosurg 16:35–39, 1990.

285. Gaufin LM, Goodman SJ: Cervical spine injuries in infants:Problems in management. J Neurosurg 42:179–184, 1975.

286. Geerts WH, Code KI, Jay RM, Chen E, Szalai JP: A prospectivestudy of venous thromboembolism after major trauma. N EnglJ Med 331:1601–1606, 1994.

287. Geisler FH: Commentary on NASCIS-2. J Spinal Disord 5:132–133, 1992 (comment).

288. Geisler FH: GM-1 ganglioside and motor recovery followinghuman spinal cord injury. J Emerg Med 11[Suppl 1]:49–55,1993.

289. Geisler FH, Coleman WP, Grieco G, Dorsey FC, Poonian D, TheSygen Study Group: The GM-1 Ganglioside Multi-center AcuteSpinal Cord Injury Study. Spine 26(24 Suppl):S87–S98, 2001.

290. Geisler FH, Dorsey FC, Coleman WP: Correction: Recovery ofmotor function after spinal-cord injury—A randomized,placebo-controlled trial with GM-1 ganglioside. N Engl J Med325:1659–1660, 1991.

291. Geisler FH, Dorsey FC, Coleman WP: Recovery of motor func-tion after spinal-cord injury: A randomized, placebo-controlledtrial with GM-1 ganglioside. N Engl J Med 324:1829–1838, 1991.

292. Geisler FH, Dorsey FC, Coleman WP: GM-1 ganglioside inhuman spinal cord injury. J Neurotrauma 9[Suppl 2]:S517–S530, 1992.

293. Geisler FH, Dorsey FC, Coleman WP: GM-1 ganglioside forspinal cord injury. N Engl J Med 326:494, 1992 (letter).

294. Geisler WO, Wynne-Jones M, Jousse AT: Early management ofthe patient with trauma to the spinal cord. Med Serv J Can22:512–523, 1966.

295. George ER, Scholten DJ, Buechler CM, Jordan-Tibbs J, MatticeC, Albrecht RM: Failure of methylprednisolone to improve theoutcome of spinal cord injuries. Am Surg 61:659–664, 1995.

296. Georgopoulos G, Pizzutillo PD, Lee MS: Occipito-atlantal in-stability in children: A report of five cases and review of theliterature. J Bone Joint Surg Am 69A:429–436, 1987.

297. Gerhart KA, Johnson RL, Menconi J, Hoffman RE, LammertseDP: Utilization and effectiveness of methylprednisolone in apopulation-based sample of spinal cord injured persons. Para-plegia 33:316–321, 1995.

298. Gerndt SJ, Rodriguez JL, Pawlik JW, Taheri PA, Wahl WL,Micheals AJ, Papadopoulos SM: Consequences of high-dosesteroid therapy for acute spinal cord injury. J Trauma 42:279–284, 1997.

299. Gerrelts BD, Petersen EU, Mabry J, Petersen SR: Delayed diag-nosis of cervical spine injuries. J Trauma 31:1622–1626, 1991.

300. Giacobetti FB, Vaccaro AR, Bos-Giacobetti MA, Deeley DM,Albert TJ, Farmer JC, Cotler JM: Vertebral artery occlusionassociated with cervical spine trauma: A prospective analysis.Spine 22:188–192, 1997.

301. Givens TG, Polley KA, Smith GF, Hardin WD Jr: Pediatriccervical spine injury: A three-year experience. J Trauma 41:310–314, 1996.

302. Glaser JA, Whitehill R, Stamp WG, Jane JA: Complicationsassociated with the halo-vest: A review of 245 cases.J Neurosurg 65:762–769, 1986.

303. Gleizes V, Jacquot FP, Signoret F, Feron JM: Combined injuriesin the upper cervical spine: Clinical and epidemiological dataover a 14-year period. Eur Spine J 9:386–392, 2000.

304. Godard J, Hadji M, Raul JS: Odontoid fractures in the child withneurological injury: Direct anterior osteosynthesis with acortico-spongious screw and literature review. Childs NervSyst 13:105–107, 1997.

305. Goffin J, Plets C, Van den Bergh R: Anterior cervical fusion andosteosynthetic stabilization according to Caspar: A prospectivestudy of 41 patients with fractures and/or dislocations of thecervical spine. Neurosurgery 25:865–871, 1989.

306. Goldstein SJ, Woodring JH, Young AB: Occipital condyle frac-ture associated with cervical spine injury. Surg Neurol 17:350–352, 1982.

307. Gonzalez RP, Fried PO, Bukhalo M, Holevar MR, FalimirskiME: Role of clinical examination in screening for blunt cervicalspine injury. J Am Coll Surg 189:152–157, 1999.

308. Goth P: Spinal Injury: Clinical Criteria for Assessment and Man-agement. Augusta, Medical Care Development Publishing, 1994.

309. Govender S, Charles RW: Traumatic spondylolisthesis of theaxis. Injury 18:333–335, 1987.

310. Govender S, Grootboom M: Fractures of the dens: The results ofnon-rigid immobilization. Injury 19:165–167, 1988.

311. Grabb PA, Pang D: Magnetic resonance imaging in the evalu-ation of spinal cord injury without radiographic abnormality inchildren. Neurosurgery 35:406–414, 1994.

312. Grabb BC, Frye TA, Hedlund GL, Vaid YN, Grabb PA, RoyalSA: MRI diagnosis of suspected atlanto-occipital dissociation inchildhood. Pediatr Radiol 29:275–281, 1999.

313. Grady MS, Howard MA, Jane JA, Persing JA: Use of the Phil-adelphia collar as an alternative to the halo vest in patients withC-2, C-3 fractures. Neurosurgery 18:151–156, 1986.

314. Grahm TW, Zadrozny DB, Harrington T: The benefits of earlyjejunal hyperalimentation in the head-injured patient. Neuro-surgery 25:729–735, 1989.



315. Grant GA, Mirza SK, Chapman JR, Winn HR, Newell DW,Jones DT, Grady MS: Risk of early closed reduction in cervicalspine subluxation injuries. J Neurosurg 90[Suppl 1]:13–18,1999.

316. Graziano AF, Scheidel EA, Cline JR, Baer LJ: A radiographiccomparison of prehospital cervical immobilization methods.Ann Emerg Med 16:1127–1131, 1987.

317. Green D, Chen D, Chmiel JS, Olsen NK, Berkowitz M, NovickA, Alleva J, Steinberg D, Nussbaum S, Tolotta M: Prevention ofthromboembolism in spinal cord injury: Role of low molecularweight heparin. Arch Phys Med Rehabil 75:290–292, 1994.

318. Green BA, Eismont FJ, O’Heir JT: Spinal cord injury: A systemsapproach—Prevention, emergency medical services, and emer-gency room management. Crit Care Clin 3:471–493, 1987.

319. Green D, Lee MY, Ito VY, Cohn T, Press J, Filbrandt PR,VandenBerg WC, Yarkony GM, Meyer PR Jr: Fixed- vsadjusted-dose heparin in the prophylaxis of thromboembolismin spinal cord injury. JAMA 260:1255–1258, 1988.

320. Green D, Lee MY, Lim AC, Chmiel JS, Vetter M, Pang T, ChenD, Fenton L, Yarkony GM, Meyer PR Jr: Prevention of throm-boembolism after spinal cord injury using low-molecular-weight heparin. Ann Intern Med 113:571–574, 1990.

321. Greene KA, Dickman CA, Marciano FF, Drabier J, Drayer BP,Sonntag VKH: Transverse atlantal ligament disruption associ-ated with odontoid fractures. Spine 19:2307–2314, 1994.

322. Greene KA, Dickman CA, Marciano FF, Drabier JB, HadleyMN, Sonntag VKH: Acute axis fractures: Analysis of manage-ment and outcome in 340 consecutive cases. Spine 22:1843–1852, 1997.

323. Greenfield LJ: Does cervical spinal cord injury induce a higherincidence of complications after prophylactic Greenfield filterusage? J Vasc Interv Radiol 8:719–720, 1997.

324. Gresham GE, Labi ML, Dittmar SS, Hicks JT, Joyce SZ, StehlikMA: The Quadriplegia Index of Function (QIF): Sensitivity andreliability demonstrated in a study of thirty quadriplegic pa-tients. Paraplegia 24:38–44, 1986.

325. Griffiths SC: Fracture of odontoid process in children. J PediatrSurg 7:680–683, 1972.

326. Grisolia A, Bell RL, Peltier LF: Fractures and dislocations of thespine complicating ankylosing spondylitis: A report of sixcases. J Bone Joint Surg Am 49A:339–344, 1967.

327. Griswold DM, Albright JA, Schiffman E, Johnson R, SouthwickW: Atlanto-axial fusion for instability. J Bone Joint Surg Am60A:285–292, 1978.

328. Grogaard B, Dullerud R, Magnaes B: Acute torticollis in chil-dren due to atlanto-axial rotary fixation. Arch Orthop TraumaSurg 112:185–188, 1993.

329. Grossman MD, Reilly PM, Gillett T, Gillett D: National surveyof the incidence of cervical spine injury and approach to cervi-cal spine clearance in U.S. trauma centers. J Trauma 47:684–690,1999.

330. Gschaedler R, Dollfus P, Molé JP, Molé L, Loeb JP: Reflectionson the intensive care of acute cervical spinal cord injuries in ageneral traumatology centre. Paraplegia 17:58–61, 1979.

331. Guigui P, Milaire M, Morvan G, Lassale B, Deburge A: Trau-matic atlantooccipital dislocation with survival: Case reportand review of the literature. Eur Spine J 4:242–247, 1995.

332. Guiot B, Fessler RG: Complex atlantoaxial fractures.J Neurosurg 91[Suppl 2]:139–143, 1999.

333. Gunby P: New focus on spinal cord injury. JAMA 245:1201–1206, 1981.

334. Gunduz S, Ogur E, Mohur H, Somuncu I, Acjksoz E, UstunsozB: Deep vein thrombosis in spinal cord injured patients. Para-plegia 31:606–610, 1993.

335. Gunn BD, Eizenberg N, Silberstein M, McMeeken JM, Tully EA,Stillman BC, Brown DJ, Gutteridge GA: How should an uncon-scious person with a suspected neck injury be positioned?Prehospital Disaster Med 10:239–244, 1995.

336. Gurdjian ES, Hardy WG, Lindner DW, Thomas LM: Closedcervical trauma associated with involvement of carotid andvertebral arteries. J Neurosurg 20:418–427, 1963.

337. Guttman L: Initial treatment of traumatic paraplegia and tetra-plegia, in Spinal Injuries: Proceedings of the Symposium held in theRoyal College of Surgeons of Edinburgh, June 7–8, 1963. Edinburgh,The Royal College of Surgeons, 1967.

338. Hachen HJ: Anticoagulant therapy in patients with spinal cordinjury. Paraplegia 12:176–187, 1974.

339. Hachen HJ: Emergency transportation in the event of acutespinal cord lesion. Paraplegia 12:33–37, 1974.

340. Hachen HJ: Idealized care of the acutely injured spinal cord inSwitzerland. J Trauma 17:931–936, 1977.

341. Hadley MN: Hypermetabolism after CNS trauma: Arrestingthe “injury cascade.” Nutrition 5:143, 1989.

342. Hadley MN: Hypermetabolism following head trauma: Nutri-tional considerations, in Barrow DL (ed): Complications andSequelae of Head Injury (Neurosurgical Topics series). Park Ridge,AANS, 1992, pp 161–168.

343. Hadley MN, Argires PJ. The acute/emergent management ofvertebral column fracture dislocation injuries, in NeurologicalEmergencies. Park Ridge, AANS, 1994, pp 249–262.

344. Hadley MN, Browner CM, Liu SS, Sonntag VKH: New subtypeof acute odontoid fractures (type IIA). Neurosurgery 22:67–71,1988.

345. Hadley MN, Browner C, Sonntag VKH: Axis fractures: A com-prehensive review of management and treatment in 107 cases.Neurosurgery 17:281–290, 1985.

346. Hadley MN, Dickman CA, Browner CM, Sonntag VKH: Acutetraumatic atlas fractures: Management and long term outcome.Neurosurgery 23:31–35, 1988.

347. Hadley MN, Dickman CA, Browner CM, Sonntag VKH: Acuteaxis fractures: A review of 229 cases. J Neurosurg 71:642–647,1989.


349. Hadley MN, Grahm TW, Harrington T, Schiller WR,McDermott MK, Posillico DB: Nutritional support andneurotrauma: A critical review of early nutrition in forty-fiveacute head injury patients. Neurosurgery 19:367–373, 1986.

350. Hadley MN, Zabramski JM, Browner CM, Rekate H, SonntagVKH: Pediatric spinal trauma: Review of 122 cases of spinalcord and vertebral column injuries. J Neurosurg 68:18–24, 1988.

351. Hall ED, Wolf DL: A pharmacological analysis of the patho-physiological mechanisms of posttraumatic spinal cord isch-emia. J Neurosurg 64:951–961, 1986.

352. Hall AJ, Wagle VG, Raycroft J, Goldman RL, Butler AR: Mag-netic resonance imaging in cervical spine trauma. J Trauma34:21–26, 1993.

353. Halliday AL, Henderson BR, Hart BL, Benzel EC: The manage-ment of unilateral lateral mass/facet fractures of the subaxialcervical spine: The use of magnetic resonance imaging to pre-dict instability. Spine 22:2614–2621, 1997.


Bibliography S187


355. Hamilton RS, Pons PT: The efficacy and comfort of full-bodyvacuum splints for cervical-spine immobilization. J Emerg Med14:553–559, 1996.

356. Hamilton BB, Granger CV, Sherwin FS, Zielezny M, TashmanJS: A Uniform National Data System for Medical Rehabilitation,in Fuhrer MJ (ed): Rehabilitation Outcomes: Analysis and Measure-ment. Baltimore, Paul H. Brookes Publishing Co., 1987, pp137–147.

357. Hamilton BB, Laughlin JA, Fiedler RC, Granger CV: Interraterreliability of the 7-level Functional Independence Measure(FIM). Scand J Rehabil Med 26:115–119, 1994.

358. Hamilton BB, Laughlin JA, Granger CV, Kayton RM: Interrateragreement of the seven-level Functional Independence Mea-sure (FIM). Arch Phys Med Rehabil 72:790, 1991.

359. Han SY, Witten DM, Mussleman JP: Jefferson fracture of theatlas: Report of six cases. J Neurosurg 44:368–371, 1976.

360. Hanigan WC, Anderson RJ: Commentary on NASCIS-2. J Spi-nal Disord 5:125–133, 1992.


362. Hanson JA, Blackmore CC, Mann FA, Wilson AJ: Cervical spineinjury: A clinical decision rule to identify high-risk patients forhelical CT screening. AJR Am J Roentgenol 174:713–717, 2000.

363. Hanssen AD, Cabanela ME: Fractures of the dens in adultpatients. J Trauma 27:928–934, 1987.

364. Harding-Smith J, MacIntosh PK, Sherbon KJ: Fracture of theoccipital condyle: A case report and review of the literature.J Bone Joint Surg Am 63A:1170–1171, 1981.

365. Harmanli O, Koyfman Y: Traumatic atlanto-occipital disloca-tion with survival: A case report and review of the literature.Surg Neurol 39:324–330, 1993.

366. Harrington JF, Likavec MJ, Smith AS: Disc herniation in cervi-cal fracture subluxation. Neurosurgery 29:374–379, 1991.

367. Harris JH, Carson GC, Wagner LK: Radiologic diagnosis oftraumatic occipitovertebral dissociation: Part 1—Normal oc-cipitovertebral relationships on lateral radiographs of supinesubjects. AJR Am J Roentgenol 162:881–886, 1994.

368. Harris JH Jr, Carson GC, Wagner LK, Kerr N: Radiologic diag-nosis of traumatic occipitovertebral dissociation: Part 2—Com-parison of three methods of detecting occipitovertebral rela-tionships on lateral radiographs of supine subjects. AJR Am JRoentgenol 162:887–892, 1994.

369. Harris S, Chen D, Green D: Enoxaparin for thromboembolismprophylaxis in spinal injury: Preliminary report on experiencewith 105 patients. Am J Phys Med Rehabil 75:326–327, 1996.

370. Hashimoto T, Watanabe O, Takase M, Koniyama J, Kobota M:Collet-Sicard syndrome after minor head trauma. Neurosur-gery 23:367–370, 1988.

371. Hauswald M, Ong G, Tandberg D, Omar Z: Out-of-hospitalspinal immobilization: Its effect on neurologic injury. AcadEmerg Med 5:214–219, 1998.

372. Hays MB, Alker GJ Jr: Fractures of the atlas vertebra: Thetwo-part burst fracture of Jefferson. Spine 13:601–603, 1988.

373. Hays MB, Bernhang AM: Fractures of the atlas vertebra: Athree-part fracture not previously classified. Spine 17:240–242,1992.

374. Heary RF, Hunt CD, Krieger AJ, Antonio C, Livingston DH:Acute stabilization of the cervical spine by halo/vest applica-tion facilitates evaluation and treatment of multiple traumapatients. J Trauma 33:445–451, 1992.

375. Heller JG, Viroslav S, Hudson T: Jefferson fractures: The role ofmagnification artifact in assessing transverse ligament integ-rity. J Spinal Disord 6:392–396, 1993.

376. Henry AD, Bohly J, Grosse A: Fixation of odontoid fractures byan anterior screw. J Bone Joint Surg Br 81B:472–477, 1999.

377. Herr CH, Ball PA, Sargent SK, Quinton HB: Sensitivity ofprevertebral soft tissue measurement of C3 for detection ofcervical spine fractures and dislocations. Am J Emerg Med16:346–349, 1998.

378. Herzenberg JE, Hensinger RN, Dedrick DK, Phillips WA: Emer-gency transport and positioning of young children who have aninjury of the cervical spine: The standard backboard may behazardous. J Bone Joint Surg Am 71A:15–22, 1989.

379. Highland TR, Salciccioli GG: Is immobilization adequate treat-ment of unstable burst fractures of the atlas? A case report withlong-term follow-up evaluation. Clin Orthop 201:196–200,1985.

379a. Hinchey JJ, Bickel WH: Fracture of the atlas: Review and pre-sentation of data on eight cases. Ann Surg 121:826–830, 1945.

380. Hladky JP, Lejeune JP, Leclercq F, Dhellemmes P, Christiaens JL:Traumatic atlanto-occipital dislocation [in French]. Neurochirurgie37:312–317, 1991.

381. Hoffman JR, Mower WR, Wolfson AB, Todd KH, Zucker MI:Validity of a set of clinical criteria to rule out injury to thecervical spine in patients with blunt trauma: National Emer-gency X-Radiography Utilization Study Group. N Engl J Med343:94–99, 2000.

382. Hoffman JR, Schriger DL, Mower W, Luo JS, Zucker M: Low-risk criteria for cervical-spine radiography in blunt trauma: Aprospective study. Ann Emerg Med 21:1454–1460, 1992.

383. Hoffman JR, Wolfson AB, Todd K, Mower WR: Selective cer-vical spine radiography in blunt trauma: Methodology of theNational Emergency X-Radiography Utilization Study (NEX-US). Ann Emerg Med 32:461–469, 1998.

384. Holliman CJ, Mayer JS, Cook RT Jr, Smith JS Jr: Is the antero-posterior cervical spine radiograph necessary in initial traumascreening? Am J Emerg Med 9:421–425, 1991.

385. Hosono N, Yonenobu K, Ebara S, Ono K: Cineradiographicmotion analysis of atlantoaxial instability in os odontoideum.Spine 16[Suppl 10]:S480–S482, 1991.

386. Hosono N, Yonenobu K, Kawagoe K, Hirayama N, Ono K:Traumatic anterior atlanto-occipital dislocation: A case reportwith survival. Spine 18:786–790, 1993.

387. Huerta C, Griffith R, Joyce SM: Cervical spine stabilization inpediatric patients: Evaluation of current techniques. AnnEmerg Med 16:1121–1126, 1987.

388. Hulsewe KW, van Acker BA, von Meyenfeldt MF, Soeters PB:Nutritional depletion and dietary manipulation: Effects on theimmune response. World J Surg 23:536–544, 1999.

389. Hummel A, Plaue R: Diagnosis and treatment of atlanto-occipital ruptures [in German]. Unfallchirurgie 14:311–319,1988.

390. Hurlbert RJ: Methylprednisolone for acute spinal cord injury:An inappropriate standard of care. J Neurosurg 93[Suppl 1]:1–7, 2000.

391. Hurlbert RJ: The use of methylprednisolone. J Neurosurg93[Suppl 1]:340–341, 2000 (letter).

391a. Hurlbert RJ: The role of steroids in acute spinal cord injury: Anevidence-based analysis. Spine 26(24 Suppl):S39–S46, 2001.

392. Ide C, Nisolle JF, Misson N, Trigaux JP, Gustin T, De Coene B,Gilliard C: Unusual occipitoatlantal fracture dissociation withno neurological impairment: Case report. J Neurosurg 88:773–776, 1998.

393. Iida H, Tachibana S, Kitahara T, Horiike S, Ohwada T, Fujii K:Association of head trauma with cervical spine injury, spinalcord injury, or both. J Trauma 46:450–452, 1999.



394. Ireland AJ, Britton I, Forrester AW: Do supine oblique viewsprovide better imaging of the cervicothoracic junction thanswimmer’s views? J Accid Emerg Med 15:151–154, 1998.


396. Jacoby CG: Fracture of the occipital condyle. AJR Am JRoentgenol 132:500, 1979 (letter).

397. Jakim I, Sweet MB: Transverse fracture through the body of theaxis. J Bone Joint Surg Br 70B:728–729, 1988.

398. Jakim I, Sweet MB, Wisniewski T, Gantz ED: Isolated avulsionfracture of the anterior tubercle of the atlas. Arch OrthopTrauma Surg 108:377–379, 1989.

399. Jarrell BE, Posuniak E, Roberts J, Osterholm J, Cotler J, DitunnoJ: A new method of management using the Kim-Ray Greenfieldfilter for deep venous thrombosis and pulmonary embolism inspinal cord injury. Surg Gynecol Obstet 157:316–320, 1983.

400. Jeanneret B, Magerl F: Primary posterior fusion C1/2 in odon-toid fractures: Indications, technique, and results of transartic-ular screw fixation. J Spinal Disord 5:464–475, 1992.

401. Jeanneret B, Magerl F, Ward JC: Overdistraction: A hazard ofskull traction in the management of acute injuries of the cervicalspine. Arch Orthop Trauma Surg 110:242–245, 1991.

402. Jefferson G: Fractures of the atlas vertebra: Report of four casesand a review of those previously reported. Br J Surg 7:407–422,1920.

403. Jenkins JD, Coric D, Branch CL Jr: A clinical comparison of one-and two-screw odontoid fixation. J Neurosurg 89:366–370,1998.

404. Jevtich V: Traumatic lateral atlanto-occipital dislocation withspontaneous bony fusion: A case report. Spine 14:123–124,1989.

405. Johnson DR, Hauswald M, Stockhoff C: Comparison of a vac-uum splint device to a rigid backboard for spinal immobiliza-tion. Am J Emerg Med 14:369–372, 1996.

406. Jones SL: Spinal trauma board. Phys Ther 57:921–922, 1977.407. Jones DN, Knox AM, Sage MR: Traumatic avulsion fracture of

the occipital condyles and clivus with associated unilateralatlantooccipital distraction. AJNR Am J Neuroradiol 11:1181–1183, 1990.

408. Jonsson M, Tollback A, Gonzales H, Borg J: Inter-rater reliabil-ity of the 1992 international standards for neurological andfunctional classification of incomplete spinal cord injury. Spi-nal Cord 38:675–679, 2000.

409. Julien TD, Frankel B, Traynelis VC, Ryken TC: Evidence-basedanalysis of odontoid fracture management. Neurosurg Focus8:Article 1, 2000.

410. Kalfas I, Wilberger JE, Goldberg A, Prostko ER: Magnetic res-onance imaging in acute spinal cord trauma. Neurosurgery23:295–299, 1988.

411. Kalff R, Kocks W, Grote W, Schmit-Neuerburg KP: Operativespondylodesis in injuries of the lower cervical spine.Neurosurg Rev 16:211–220, 1993.

412. Kaneriya PP, Schweitzer ME, Spettell C, Cohen MJ, Karasick D:The cost-effectiveness of oblique radiography in the exclusionof C7-T1 injury in trauma patients. AJR Am J Roentgenol171:959–962, 1998.

413. Katzberg RW, Benedetti PF, Drake CM, Ivanovic M, Levine RA,Beatty CS, Nemzek WR, McFall RA, Ontell FK, Bishop DM,Poirier VC, Chong BW: Acute cervical spine injuries: Prospec-tive MR imaging assessment at a level 1 trauma center. Radi-ology 213:203–212, 1999.

414. Kaufman RA, Dunbar JS, Botsford JA, McLaurin RL: Traumaticlongitudinal atlanto-occipital distraction injuries in children.AJNR Am J Neuroradiol 3:415–419, 1982.

415. Kaufman HH, Rowlands BJ, Stein DK, Kopaniky DR,Gildenberg PL: General metabolism in patients with acute para-plegia and quadriplegia. Neurosurgery 16:309–313, 1985.

416. Kaups K, Davis J: Patients with gunshot wounds to the head donot require cervical spine immobilization and evaluation.J Trauma 44:865–867, 1998.

417. Kawabe N, Hirotani H, Tanaka O: Pathomechanism of atlanto-axial rotatory fixation in children. J Pediatr Orthop 9:569–574,1989.

418. Kearns PJ, Thompson JD, Werner PC, Pipp TL, Wilmot CB:Nutritional and metabolic response to acute spinal-cord injury.JPEN J Parenter Enteral Nutr 16:11–15, 1992.

419. Keiper MD, Zimmerman RA, Bilaniuk LT: MRI in the assess-ment of the supportive soft tissues of the cervical spine in acutetrauma in children. Neuroradiology 40:359–363, 1998.


421. Key A: Cervical spine dislocations with unilateral facet inter-locking. Paraplegia 13:208–215, 1975.

422. Khansarinia S, Dennis JW, Veldenz HC, Butcher JL, Hartland L:Prophylactic Greenfield filter placement in selected high-risktrauma patients. J Vasc Surg 22:231–236, 1995.

423. Kilburn MP, Smith DP, Hadley MN: The initial evaluation andtreatment of the patient with spinal trauma, in Batjer HH,Loftus CM (eds): Textbook of Neurological Surgery: Principles andPractice. Philadelphia, Lippincott Williams & Wilkins (in press).

424. Kindt GW, Ducker TB, Huddlestone J: Regulation of spinalcord blood flow, in Russell RWR (ed): Brain and Blood Flow:Proceedings of the 4th International Symposium on the Regulation ofCerebral Blood Flow. London: Pitman Medical, 1971, pp 401–405.

425. King BS, Gupta R, Narayan RK: The early assessment andintensive care unit management of patients with severe trau-matic brain and spinal cord injuries. Surg Clin North Am80:855–870, viii–ix, 2000.

426. Kinney TB, Rose SC, Valji K, Oglevie SB, Roberts AC: Doescervical spinal cord injury induce a higher incidence of com-plications after prophylactic Greenfield inferior vena cava filterusage? J Vasc Interv Radiol 7:907–915, 1996.

427. Kiwerski JE: Early anterior decompression and fusion for crushfractures of cervical vertebrae. Int Orthop 17:166–168, 1993.

428. Klein GR, Vaccaro AR, Albert TJ, Schweitzer M, Deely D,Karasick D, Cotler JM: Efficacy of magnetic resonance imagingin the evaluation of posterior cervical spine fractures. Spine24:771–774, 1999.

429. Kleyn PJ: Dislocations of the cervical spine: Closed reductionunder anaesthesia. Paraplegia 22:271–281, 1984.

430. Klose KJ, Green BA, Smith RS, Adkins RH, MacDonald AM:University of Miami Neuro-Spinal Index (UMNI): A quantita-tive method for determining spinal cord function. Paraplegia18:331–336, 1980.

431. Kobrine AI, Doyle TF, Martins AN: Autoregulation of spinalcord blood flow. Clin Neurosurg 22:573–581, 1975.

432. Kobrine AI, Doyle TF, Rizzoli HV: Spinal cord blood flow asaffected by changes in systemic arterial blood pressure.J Neurosurg 44:12–15, 1976.

433. Kolb JC, Summers RL, Galli RL: Cervical collar-inducedchanges in intracranial pressure. Am J Emerg Med 17:135–137,1999.

Bibliography S189


434. Koop SE, Winter RB, Lonstein JE: The surgical treatment ofinstability of the upper part of the cervical spine in children andadolescents. J Bone Joint Surg Am 66A:403–411, 1984.

435. Kornberg M: Atypical unstable burst fracture of the atlas:Treated by primary atlantoaxial fusion. Orthop Rev 15:727–729,1986.

436. Korres DS, Zoubos AB, Kavadias K, Babis GC, Balalis K: The“tear drop” (or avulsed) fracture of the anterior inferior angle ofthe axis. Eur Spine J 3:151–154, 1994.

437. Koskinen EV, Nieminen R: Fractures and dislocations of thecervical spine: Treatment and results of 159 cases. Int Surg47:472–485, 1967.

438. Kossuth LC: The removal of injured personnel from wreckedvehicles. J Trauma 5:703–708, 1965.

439. Kossuth LC: The initial movement of the injured. Mil Med132:18–21, 1967.

440. Kowalski HM, Cohen WA, Cooper P, Wisoff JH: Pitfalls in theCT diagnosis of atlantoaxial rotary subluxation. AJR Am JRoentgenol 149:595–600, 1987.


442. Kucukdeveci AA, Yavuzer G, Tennant A, Suldur N, Sonel B,Arasil T: Adaptation of the modified Barthel Index for use inphysical medicine and rehabilitation in Turkey. Scand JRehabil Med 32:87–92, 2000.

443. Kuhns LR, Loder RT, Farley FA, Hensinger RN: Nuchal cordchanges in children with os odontoideum: Evidence for associ-ated trauma. J Pediatr Orthop 18:815–819, 1998.

444. Kulkarni JR, Burt AA, Tromans AT, Constable PD: Prophylacticlow dose heparin anticoagulant therapy in patients with spinalcord injuries: A retrospective study. Paraplegia 30:169–172,1992.

445. Laham JL, Cotcamp DH, Gibbons PA, Kahana MD, Crone KR:Isolated head injuries versus multiple trauma in pediatric pa-tients: Do the same indications for cervical spine evaluationapply? Pediatr Neurosurg 21:221–226, 1994.

446. Lam CH, Stratford J: Bilateral hypoglossal nerve injury withoccipital condylar fracture. Can J Neurol Sci 23:145–148, 1996.

447. Lamb GC, Tomski MA, Kaufman J, Maiman DJ: Is chronicspinal cord injury associated with increased risk of venousthromboembolism? J Am Paraplegia Soc 16:153–156, 1993.

448. Landells CD, Van Peteghem PK: Fractures of the atlas: Classi-fication, treatment and morbidity. Spine 13:450–452, 1988.

449. Landi G, Ciccone A: GM-1 ganglioside for spinal-cord injury.N Engl J Med 326:493–494, 1992.

450. Landis JR, Koch GG: The measurement of observer agreementfor categorical data. Biometrics 33:159–174, 1977.

451. Lazar RB, Yarkony GM, Ortolano D, Heinemann AW, PerlowE, Lovell L, Meyer PR: Prediction of functional outcome bymotor capability after spinal cord injury. Arch Phys MedRehabil 70:819–822, 1989.

452. Ledsome JR, Sharp JM: Pulmonary function in acute cervicalcord injury. Am Rev Respir Dis 124:41–44, 1981.

453. Lee C, Woodring JH: Unstable Jefferson variant atlas fractures:An unrecognized cervical injury. AJNR Am J Neuroradiol12:1105–1110, 1991.

454. Lee TT, Green BA, Petrin DR: Treatment of stable burst fractureof the atlas (Jefferson fracture) with rigid cervical collar. Spine23:1963–1967, 1998.

455. Lee AS, MacLean JC, Newton DA: Rapid traction for reductionof cervical spine dislocations. J Bone Joint Surg Br 76B:352–356,1994.

456. Lee C, Woodring JH, Goldstein SJ, Daniel TL, Young AB, TibbsPA: Evaluation of traumatic atlantooccipital dislocations. AJNRAm J Neuroradiol 8:19–26, 1987.

457. Lee C, Woodring JH, Walsh JW: Carotid and vertebral injury insurvivors of atlanto-occipital dislocation: Case reports and lit-erature review. J Trauma 31:401–407, 1991.

458. Legros B, Fournier P, Chiaroni P, Ritz O, Fusciardi J: Basalfracture of the skull and lower (IX, X, XI, XII) cranial nervespalsy: Four case reports including two fractures of the occipitalcondyle—A literature review. J Trauma 48:342–348, 2000.

459. Lehmann KG, Lane JG, Piepmeier JM, Batsford WP: Cardiovas-cular abnormalities accompanying acute spinal cord injury inhumans: Incidence, time course and severity. J Am CollCardiol 10:46–52, 1987.

460. Lemons VR, Wagner FC Jr: Stabilization of subaxial cervicalspinal injuries. Surg Neurol 39:511–518, 1993.

461. Lennarson PJ, Mostafavi H, Traynelis VC, Walters BC: Manage-ment of type II dens fractures: A case-control study. Spine25:1234–1237, 2000.

462. Leone A, Cerase A, Colosimo C, Lauro L, Puca A, Marano P:Occipital condylar fractures: A review. Radiology 216:635–644,2000.

463. Lerner EB, Billittier AJ IV, Moscati RM: The effects of neutralpositioning with and without padding on spinal immobiliza-tion of healthy subjects. Prehosp Emerg Care 2:112–116, 1998.

464. Leventhal MR, Boydston WR, Sebes JI, Pinstein ML, WatridgeCB, Lowrey R: The diagnosis and treatment of fractures of theoccipital condyle. Orthopedics 15:944–947, 1992.

465. Levi L, Wolf A, Belzberg H: Hemodynamic parameters in pa-tients with acute cervical cord trauma: Description, interven-tion, and prediction of outcome. Neurosurgery 33:1007–1017,1993.

466. Levi L, Wolf A, Rigamonti D, Ragheb J, Mirvis S, Robinson WL:Anterior decompression in cervical spine trauma: Does thetiming of surgery affect the outcome? Neurosurgery 29:216–222, 1991.

467. Levine AM, Edwards CC: The management of traumatic spon-dylolisthesis of the axis. J Bone Joint Surg Am 67A:217–226,1985.

468. Levine AM, Edwards CC: Treatment of injuries in the C1-C2complex. Orthop Clin North Am 17:31–44, 1986.

469. Levine AM, Edwards CC: Fractures of the atlas. J Bone JointSurg Am 73A:680–691, 1991.

470. Levine AM, Mazel C, Roy-Camille R: Management of fractureseparations of the articular mass using posterior cervical plat-ing. Spine 17[Suppl 10]:S447–S454, 1992.

471. Levine AM, Nash MS, Green BA, Shea JD, Aronica MJ: Anexamination of dietary intakes and nutritional status of chronichealthy spinal cord injured individuals. Paraplegia 30:880–889,1992.

472. Lewelt W, Jenkins LW, Miller JD: Autoregulation of cerebralblood flow after experimental fluid percussion injury of thebrain. J Neurosurg 53:500–511, 1980.

473. Lewis LM, Docherty M, Ruoff BE, Fortney JP, Keltner RA Jr,Britton P: Flexion-extension views in the evaluation of cervical-spine injuries. Ann Emerg Med 20:117–121, 1991.

474. Lieberman IH, Webb JK: Cervical spine injuries in the elderly.J Bone Joint Surg Br 76B:877–881, 1994.

475. Liew SC, Hill DA: Complication of hard cervical collars inmulti-trauma patients. Aust N Z J Surg 64:139–140, 1994.

476. Lifeso RM, Colucci MA: Anterior fusion for rotationally unsta-ble cervical spine fractures. Spine 25:2028–2034, 2000.



477. Linares HA, Mawson AR, Suarez E, Biundo JJ: Associationbetween pressure sores and immobilization in the immediatepost-injury period. Orthopedics 10:571–573, 1987.

478. Lind B, Nordwall A, Sihlbom H: Odontoid fractures treatedwith halo-vest. Spine 12:173–177, 1987.

479. Lind B, Sihlbom H, Nordwall A: Halo-vest treatment of unsta-ble traumatic cervical spine injuries. Spine 13:425–432, 1988.

480. Lindsey RW, Diliberti TC, Doherty BJ, Watson AB: Efficacy ofradiographic evaluation of the cervical spine in emergencysituations. South Med J 86:1253–1255, 1993.

481. Link TM, Schuierer G, Hufendiek A, Horch C, Peters PE: Sub-stantial head trauma: Value of routine CT examination of thecervicocranium. Radiology 196:741–745, 1995.

482. Lipson SJ: Fractures of the atlas associated with fractures of theodontoid process and transverse ligament ruptures. J BoneJoint Surg Am 59A:940–943, 1977.

483. Lockey AS, Handley R, Willett K: “Clearance” of cervical spineinjury in the obtunded patient. Injury 29:493–497, 1998.

484. Louw JA, Mafoyane NA, Small B, Neser CP: Occlusion of thevertebral artery in cervical spine dislocations. J Bone Joint SurgBr 72B:679–681, 1990.

485. Lowry DW, Pollack IF, Clyde B, Albright AL, Adelson PD:Upper cervical spine fusion in the pediatric population.J Neurosurg 87:671–676, 1997.

486. Lu K, Lee TC, Chen HJ: Closed reduction of bilateral lockedfacets of the cervical spine under general anaesthesia. ActaNeurochir (Wien) 140:1055–1061, 1998.

487. Lu K, Lee TC, Liang CL, Chen HJ: Delayed apnea in patientswith mid- to lower cervical spinal cord injury. Spine 25:1332–1338, 2000.

488. Lucas JT, Ducker TB: Motor classification of spinal cord injurieswith mobility, morbidity and recovery indices. Am Surg 45:151–158, 1979.

489. Ludwig SC, Vaccaro AR, Balderston RA, Cotler JM: Immediatequadriparesis after manipulation for bilateral cervical facet sub-luxation: A case report. J Bone Joint Surg Am 79A:587–590,1997.

490. Lui TN, Lee ST, Wong CW, Yeh YS, Tzaan WC, Chen TY, HungSY: C1-C2 fracture dislocations in children and adolescents.J Trauma 40:408–411, 1996.

491. Lukhele M: Fractures of the vertebral lamina associated withunifacet and bifacet cervical spine dislocations. S Afr J Surg32:112–114, 1994.

492. Lyons MK, Partington MD, Meyer FB: A randomized, con-trolled trial of methylprednisolone or naloxone in the treatmentof acute spinal-cord injury. N Engl J Med 323:1207–1208, 1990(letter).

493. Macdonald RL, Schwartz ML, Mirich D, Sharkey PW, NelsonWR: Diagnosis of cervical spine injury in motor vehicle crashvictims: How many x-rays are enough? J Trauma 30:392–397,1990.

494. Mace SE: Emergency evaluation of cervical spine injuries: CTversus plain radiographs. Ann Emerg Med 14:973–975, 1985.

495. Mace SE: Unstable occult cervical-spine fracture. Ann EmergMed 20:1373–1375, 1991.

496. MacKinnon JA, Perlman M, Kirpalani H, Rehan V, Sauve R,Kovacs L: Spinal cord injury at birth: Diagnostic and prognosticdata in twenty-two patients. J Pediatr 122:431–437, 1993.

497. Mahadevan S, Mower WR, Hoffman JR, Peeples N, GoldbergW, Sonner R: Interrater reliability of cervical spine injury crite-ria in patients with blunt trauma. Ann Emerg Med 31:197–201,1998.


499. Mahale YJ, Silver JR, Henderson NJ: Neurological complica-tions of the reduction of cervical spine dislocations. J BoneJoint Surg Br 75B:403–409, 1993.

500. Mahoney FI, Barthel DW: Functional evaluation: The BarthelIndex. Md State Med J 14:61–65, 1965.

501. Maiman DJ, Larson SJ: Management of odontoid fractures.Neurosurgery 11:820, 1982 (letter).

502. Maiman DJ, Barolat G, Larson SJ: Management of bilaterallocked facets of the cervical spine. Neurosurgery 18:542–547,1986.

503. Mandabach M, Ruge JR, Hahn YS, McLone DG: Pediatric axisfractures: Early halo immobilization, management and out-come. Pediatr Neurosurg 19:225–232, 1993.

504. Mann FA, Cohen W: Occipital condyle fracture: Significance inthe assessment of occipitoatlantal stability. AJR Am JRoentgenol 163:193–194, 1994.

505. Mansel JK, Norman JR: Respiratory complications and manage-ment of spinal cord injuries. Chest 97:1446–1452, 1990.

506. Marar BC, Tay CK: Fracture of the odontoid process. Aust N ZJ Surg 46:231–236, 1976.

507. Mariani PJ: Occipital condyle fracture presenting asretropharyngeal hematoma. Ann Emerg Med 19:1447–1449,1990.

508. Marino RJ, Huang M, Knight P, Herbison GJ, Ditunno JF Jr,Segal M: Assessing selfcare status in quadriplegia: Comparisonof the Quadriplegia Index of Function (QIF) and the FunctionalIndependence Measure (FIM). Paraplegia 31:225–233, 1993.

509. Marino RJ, Rider-Foster D, Maissel G, Ditunno JF: Superiorityof motor level over single neurological level in categorizingtetraplegia. Paraplegia 33:510–513, 1995.

510. Markenson D, Foltin G, Tunik M, Cooper A, Giordano L, FittonA, Lanotte T: The Kendrick extrication device used for pediatricspinal immobilization. Prehosp Emerg Care 3:66–69, 1999.

511. Marks DS, Roberts P, Wilton PJ, Burns LA, Thompson AG: Ahalo jacket for stabilisation of the paediatric cervical spine.Arch Orthop Trauma Surg 112:134–135, 1993.

512. Maroon JC: “Burning hands” in football spinal cord injuries.JAMA 238:2049–2051, 1977.

513. Marshall LF, Knowlton S, Garfin SR, Klauber MR, EisenbergHM, Kopaniky D, Miner ME, Tabbador K, Clifton GL: Deteri-oration following spinal cord injury: A multicenter study.J Neurosurg 66:400–404, 1987.

514. Massaro F, Lanotte M: Fracture of the occipital condyle. Injury24:419–420, 1993.

515. Matava MJ, Whitesides TE Jr, Davis PC: Traumatic atlanto-occipital dislocation with survival: Serial computerized tomog-raphy as an aid to diagnosis and reduction—A report of threecases. Spine 18:1897–1903, 1993.

516. Matsui H, Imada K, Tsuji H: Radiographic classification of osodontoideum and its clinical significance. Spine 22:1706–1709,1997.

517. Matsumoto T, Tamaki T, Kawakami M, Yoshida M, Ando M,Yamada H: Early complications of high-dose methylpred-nisolone sodium succinate treatment in the follow-up of acutecervical spinal cord injury. Spine 26:426–430, 2001.

518. Maves CK, Souza A, Prenger EC, Kirks DR: Traumatic atlanto-occipital disruption in children. Pediatr Radiol 21:504–507,1991.

Bibliography S191


519. Mawson AR, Biundo JJ Jr, Neville P, Linares HA, Winchester Y,Lopez A: Risk factors for early occurring pressure ulcers fol-lowing spinal cord injury. Am J Phys Med Rehabil 67:123–127,1988.

520. Maynard FM Jr, Bracken MB, Creasey G, Ditunno JF Jr,Donovan WH, Ducker TB, Garber SL, Marino RJ, Stover SL,Tator CH, Waters RL, Wilberger JE, Young W: InternationalStandards for Neurological and Functional Classification ofSpinal Cord Injury: American Spinal Injury Association. SpinalCord 35:266–274, 1997.

521. Maynard FM Jr, Reynolds GG, Fountain S, Wilmot C, HamiltonR: Neurological prognosis after traumatic quadriplegia: Three-year experience of California Regional Spinal Cord Injury CareSystem. J Neurosurg 50:611–616, 1979.

522. Mazolewski P, Manix TH: The effectiveness of strapping tech-niques in spinal immobilization. Ann Emerg Med 23:1290–1295, 1994.

523. McCabe JB, Nolan DJ: Comparison of the effectiveness of dif-ferent cervical immobilization collars. Ann Emerg Med 15:50–53, 1986.

524. McGuire RA Jr: Protection of the unstable spine during trans-port and early hospitalization. J Miss State Med Assoc 32:305–308, 1991.

525. McGuire RA Jr, Harkey HL: Primary treatment of unstableJefferson’s fractures. J Spinal Disord 8:233–236, 1995.

526. McGuire RA Jr, Degnan G, Amundson GM: Evaluation ofcurrent extrication orthoses in immobilization of the unstablecervical spine. Spine 15:1064–1067, 1990.

527. McGuire RA Jr, Neville S, Green BA, Watts C: Spinal instabilityand the log-rolling maneuver. J Trauma 27:525–531, 1987.

528. McHugh TP, Taylor JP: Unnecessary out-of-hospital use of fullspinal immobilization. Acad Emerg Med 5:278–280, 1998.

529. McKinley WO, Jackson AB, Cardenas DD, DeVivo MJ: Long-term medical complications after traumatic spinal cord injury:A regional model systems analysis. Arch Phys Med Rehabil80:1402–1410, 1999.

530. McLaurin RL, Vernal R, Salmon JH: Treatment of fractures ofthe atlas and axis by wiring without fusion. J Neurosurg 36:773–780, 1972.

531. McMichan JC, Michel L, Westbrook PR: Pulmonary dysfunc-tion following traumatic quadriplegia: Recognition, prevention,and treatment. JAMA 243:528–531, 1980.

532. McNamara RM, Heine E, Esposito B: Cervical spine injury andradiography in alert, high-risk patients. J Emerg Med 8:177–182, 1990.

533. McSwain NE Jr: Spine management skills, in Pre-HospitalTrauma Life Support. Akron, Educational Direction, 1990, ed 2,pp 225–256.

534. McSwain NE Jr: Proper C-spine immobilization. Emerg Med27:120–121, 1995.

535. Meldon SW, Brant TA, Cydulka RK, Collins TE, Shade BR:Out-of-hospital cervical spine clearance: Agreement betweenemergency medical technicians and emergency physicians.J Trauma 45:1058–1061, 1998.

536. Menezes AH, Ryken TC: Craniovertebral abnormalities inDown’s syndrome. Pediatr Neurosurg 18:24–33, 1992.

537. Menticoglou SM, Perlman M, Manning FA: High cervical spinalcord injury in neonates delivered with forceps: Report of 15cases. Obstet Gynecol 86:589–594, 1995.

538. Merianos P, Tsekouras G, Koskinas A: An unusual fracture ofthe atlas. Injury 22:489–490, 1991.

539. Merli GJ, Crabbe S, Doyle L, Ditunno JF, Herbison GJ: Mechan-ical plus pharmacological prophylaxis for deep vein thrombosisin acute spinal cord injury. Paraplegia 30:558–562, 1992.

540. Merli GJ, Herbison GJ, Ditunno JF, Weitz HH, Henzes JH, ParkCH, Jaweed MM: Deep vein thrombosis: Prophylaxis in acutespinal cord injured patients. Arch Phys Med Rehabil 69:661–664, 1988.

541. Merriam WF, Taylor TK, Ruff SJ, McPhail MJ: A reappraisal ofacute traumatic central cord syndrome. J Bone Joint Surg Br68B:708–713, 1986.

542. Mestdagh H, Letendart J, Sensey JJ, Duquennoy A: Treatmentof fractures of the posterior axial arch: Results of 41 cases [inFrench]. Rev Chir Orthop Reparatrice Appar Mot 70:21–28,1984.

543. Mirvis SE, Diaconis JN, Chirico PA, Reiner BI, Joslyn JN,Militello P: Protocol-driven radiologic evaluation of suspectedcervical spine injury: Efficacy study. Radiology 170:831–834,1989.

544. Mirvis SE, Geisler FH, Jelinek JJ, Joslyn JN, Gellad F: Acutecervical spine trauma: Evaluation with 1.5-T MR imaging. Ra-diology 166:807–816, 1988.

545. Mirvis SE, Young JW, Lim C, Greenberg J: Hangman’s fracture:Radiologic assessment in 27 cases. Radiology 163:713–717,1987.

546. Mody BS, Morris EW: Fracture of the occipital condyle: Casereport and review of the world literature. Injury 23:350–352,1992.

547. Mollan RA, Watt PC: Hangman’s fracture. Injury 14:265–267,1982.

548. Montane I, Eismont FJ, Green BA: Traumatic occipitoatlantaldislocation. Spine 16:112–116, 1991.

549. Montesano PX, Anderson PA, Schlehr F, Thalgott JS, Lowrey G:Odontoid fractures treated by anterior odontoid screw fixation.Spine 16[Suppl 3]:S33–S37, 1991.


551. Morgan MK, Onofrio BM, Bender CE: Familial osodontoideum: Case report. J Neurosurg 70:636–639, 1989.

552. Deleted in proof.553. Mubarak SJ, Camp JF, Vuletich W, Wenger DR, Garfin SR: Halo

application in the infant. J Pediatr Orthop 9:612–614, 1989.554. Muhr MD, Seabrook DL, Wittwer LK: Paramedic use of a spinal

injury clearance algorithm reduces spinal immobilization in theout-of-hospital setting. Prehosp Emerg Care 3:1–6, 1999.

555. Muller EJ, Wick M, Muhr G: Traumatic spondylolisthesis of theaxis: Treatment rationale based on the stability of the differentfracture types. Eur Spine J 9:123–128, 2000.

556. Muller EJ, Wick M, Russe O, Muhr G: Management of odontoidfractures in the elderly. Eur Spine J 8:360–365, 1999.

557. Myllynen P, Kammonen M, Rokkanen P, Bostman O, Lalla M,Laasonen E: Deep venous thrombosis and pulmonary embo-lism in patients with acute spinal cord injury: A comparisonwith nonparalyzed patients immobilized due to spinal frac-tures. J Trauma 25:541–543, 1985.

558. Naso F: Pulmonary embolism in acute spinal cord injury. ArchPhys Med Rehabil 55:275–278, 1974.

559. Naso WB, Cure J, Cuddy BG: Retropharyngeal pseudomeningoceleafter atlanto-occipital dislocation: Report of two cases. Neurosurgery40:1288–1291, 1997.

560. Nazarian SM, Louis RP: Posterior internal fixation with screwplates in traumatic lesions of the cervical spine. Spine 16[Suppl3]:S64–S71, 1991.



561. Neifeld GL, Keene JG, Hevesy G, Leikin J, Proust A, ThistedRA: Cervical injury in head trauma. J Emerg Med 6:203–207,1988.

562. Nesathurai S: Steroids and spinal cord injury: Revisiting theNASCIS 2 and NASCIS 3 trials. J Trauma 45:1088–1093, 1998.

563. Neville S, Watts C: Management of the unstable cervical spinein transport: A re-evaluation. Aeromed J Sept/Oct:32, 1987.

564. Newey ML, Sen PK, Fraser RD: The long-term outcome aftercentral cord syndrome: A study of the natural history. J BoneJoint Surg Br 82B:851–855, 2000.

565. Nischal K, Chumas P, Sparrow O: Prolonged survival afteratlanto-occipital dislocation: Two case reports and review. Br JNeurosurg 7:677–682, 1993.

566. Noble ER, Smoker WR: The forgotten condyle: The appearance,morphology, and classification of occipital condyle fractures.AJNR Am J Neuroradiol 17:507–513, 1996.

567. Nunez DB Jr, Quencer RM: The role of helical CT in the assess-ment of cervical spine injuries. AJR Am J Roentgenol 171:951–957, 1998.


569. Oda T, Panjabi MM, Crisco JJ III, Oxland TR: Multidirectionalinstabilities of experimental burst fractures of the atlas. Spine17:1285–1290, 1992.

570. Oda T, Panjabi MM, Crisco JJ III, Oxland TR, Katz L, Nolte LP:Experimental study of atlas injuries: Part II—Relevance to clin-ical diagnosis and treatment. Spine 16[Suppl 10]:S466–S473,1991.

571. Odent T, Langlais J, Glorion C, Kassis B, Bataille J, PouliquenJC: Fractures of the odontoid process: A report of 15 cases inchildren younger than 6 years. J Pediatr Orthop 19:51–54, 1999.

572. Olerud C, Jonsson H Jr: Compression of the cervical spine cordafter reduction of fracture dislocations: Report of 2 cases. ActaOrthop Scand 62:599–601, 1991.

573. Olson CM, Jastremski MS, Vilogi JP, Madden CM, Beney KM:Stabilization of patients prior to interhospital transfer. Am JEmerg Med 5:33–39, 1987.

574. Olsson R, Kunz R: Fracture of the occipital condyle as anincidental finding during CT-evaluation of a maxillary fracture.Acta Radiol 35:90–91, 1994.

575. Orbay T, Aykol S, Seckin Z, Ergun R: Late hypoglossal nervepalsy following fracture of the occipital condyle. Surg Neurol31:402–404, 1989.

576. Ordonez BJ, Benzel EC, Naderi S, Weller SJ: Cervical facetdislocation: Techniques for ventral reduction and stabilization.J Neurosurg 92[Suppl 1]:18–23, 2000.

577. Orledge JD, Pepe PE: Out-of-hospital spinal immobilization: Isit really necessary? Acad Emerg Med 5:203–204, 1998.

578. Osenbach RK, Menezes AH: Spinal cord injury without radio-graphic abnormality in children. Pediatr Neurosci 15:168–175,1989.


580. Osterholm JL: The pathophysiological response to spinal cordinjury: The current status of related research. J Neurosurg40:5–33, 1974.

581. Osti OL, Fraser RD, Griffiths ER: Reduction and stabilisation ofcervical dislocations: An analysis of 167 cases. J Bone JointSurg Br 71B:275–282, 1989.

582. Ota T, Akaboshi K, Nagata M, Sonoda S, Domen K, Seki M,Chino N: Functional assessment of patients with spinal cordinjury: Measured by the motor score and the Functional Inde-pendence Measure. Spinal Cord 34:531–535, 1996.

583. Paeslack V, Frankel H, Michaelis L: Closed injuries of thecervical spine and spinal cord: Results of conservative treat-ment of flexion fractures and flexion rotation fracture disloca-tion of the cervical spine with tetraplegia. Proc Veterans AdmSpinal Cord Inj Conf 19:39–42, 1973.

584. Page CP, Story JL, Wissinger JP, Branch CL: Traumaticatlantooccipital dislocation: Case report. J Neurosurg 39:394–397, 1973.

585. Paley MD, Wood GA: Traumatic bilateral hypoglossal nervepalsy. Br J Oral Maxillofac Surg 33:239–241, 1995.

586. Palmer MT, Turney SZ: Tracheal rupture and atlanto-occipitaldislocation: Case report. J Trauma 37:314–317, 1994.

587. Pang D, Hanley EN: Special problems of spinal stabilization inchildren, in Cooper PR (ed): Management of Posttraumatic SpinalInstability (Neurosurgical Topics No. 3). Park Ridge, AANS, 1990,pp 181–206.

588. Pang D, Pollack IF: Spinal cord injury without radiographicabnormality in children: The SCIWORA syndrome. J Trauma29:654–664, 1989.

589. Pang D, Wilberger JE Jr: Traumatic atlanto-occipital dislocationwith survival: Case report and review. Neurosurgery 7:503–508, 1980.

590. Pang D, Wilberger JE Jr: Spinal cord injury without radio-graphic abnormalities in children. J Neurosurg 57:114–129,1982.

591. Panjabi MM, Oda T, Crisco JJ III, Oxland TR, Katz L, Nolte LP:Experimental study of atlas injuries: Part I—Biomechanicalanalysis of their mechanisms and fracture patterns. Spine16[Suppl 10]:S460–S465, 1991.

592. Papadopoulos SM, Dickman CA, Sonntag VKH, Rekate HL,Spetzler RF: Traumatic atlantooccipital dislocation with sur-vival. Neurosurgery 28:574–579, 1991.

593. Paramore CG, Dickman CA, Sonntag VKH: The anatomicalsuitability of the C1-2 complex for transarticular screw fixation.J Neurosurg 85:221–224, 1996.

594. Pasciak M, Doniec J: Results of conservative treatment of uni-lateral cervical spine dislocations. Arch Orthop Trauma Surg112:226–227, 1993.

595. Pasquale M, Fabian TC: Practice Management Guidelines forTrauma from the Eastern Association for the Surgery ofTrauma. J Trauma 44:941–957, 1998.

596. Patton JH, Kralovich KA, Cuschieri J, Gasparri M: Clearing thecervical spine in victims of blunt assault to the head and neck:What is necessary? Am Surg 66:326–331, 2000.

597. Pedersen AK, Kostuik JP: Complete fracture-dislocation of theatlantoaxial complex: Case report and recommendations for anew classification of dens fractures. J Spinal Disord 7:350–355,1994.

598. Peiffer SC, Blust P, Leyson JF: Nutritional assessment of thespinal cord injured patient. J Am Diet Assoc 78:501–505, 1981.

599. Pennecot GF, Leonard P, Peyrot Des Gachons S, Hardy JR,Pouliquen JC: Traumatic ligamentous instability of the cervicalspine in children. J Pediatr Orthop 4:339–345, 1984.

600. Pepin JW, Hawkins RJ: Traumatic spondylolisthesis of the axis:Hangman’s fracture. Clin Orthop 157:133–138, 1981.

601. Pepin JW, Bourne RB, Hawkins RJ: Odontoid fractures, withspecial reference to the elderly patient. Clin Orthop 193:178–183, 1985.

601a. Perkash A: Experience with the management of deep veinthrombosis in patients with spinal cord injury: Part II—A crit-ical evaluation of the anticoagulant therapy. Paraplegia 18:2–14, 1980.

Bibliography S193


602. Perkash A, Prakash V, Perkash I: Experience with the manage-ment of thromboembolism in patients with spinal cord injury:Part I—Incidence, diagnosis and role of some risk factors. Para-plegia 16:322–331, 1978.

603. Perry SD, McLellan B, McIlroy WE, Maki BE, Schwartz M,Fernie GR: The efficacy of head immobilization techniquesduring simulated vehicle motion. Spine 24:1839–1844, 1999.

604. Phillips WA, Hensinger RN: The management of rotatoryatlanto-axial subluxation in children. J Bone Joint Surg Am71A:664–668, 1989.

605. Piepmeier JM, Lehmann KB, Lane JG: Cardiovascular instabil-ity following acute cervical spinal cord trauma. Cent Nerv SystTrauma 2:153–160, 1985.

606. Pitzen T, Caspar W, Steudel WI, Barbier D: Dens fracture inelderly patients and surgical management [in German].Aktuelle Traumatol 24:56–59, 1994.

607. Plaisier B, Gabram SG, Schwartz RJ, Jacobs LM: Prospectiveevaluation of craniofacial pressure in four different cervicalorthoses. J Trauma 37:714–720, 1994.

608. Podolsky S, Baraff LJ, Simon RR, Hoffman JR, Larmon B, AblonW: Efficacy of cervical spine immobilization methods. J Trauma23:461–465, 1983.

609. Deleted in proof.610. Pointillart V, Petitjean ME, Wiart L, Vital JM, Lassie P, Thicoipe

M, Dabadie P: Pharmacological therapy of spinal cord injuryduring the acute phase. Spinal Cord 38:71–76, 2000.

611. Polin RS, Szabo T, Bogaev CA, Replogle RE, Jane JA: Nonopera-tive management of types II and III odontoid fractures: ThePhiladelphia collar versus the halo vest. Neurosurgery 38:450–457, 1996.

612. Pollack IF, Pang D, Sclabassi R: Recurrent spinal cord injurywithout radiographic abnormalities in children. J Neurosurg69:177–182, 1988.

613. Powell M, Kirshblum S, O’Connor KC: Duplex ultrasoundscreening for deep vein thrombosis in spinal cord injured pa-tients at rehabilitation admission. Arch Phys Med Rehabil80:1044–1046, 1999.

614. Powers B, Miller MD, Kramer RS, Martinez S, Gehweiler JA Jr:Traumatic anterior atlanto-occipital dislocation. Neurosurgery4:12–17, 1979.

615. Poynton AR, O’Farrell DA, Shannon F, Murray P, McManus F,Walsh MG: An evaluation of the factors affecting neurologi-cal recovery following spinal cord injury. Injury 28:545–548,1997.

616. Prabhu V, Kizer J, Patil A, Hellbusch L, Taylon C, Leibrock L:Vertebrobasilar thrombosis associated with nonpenetrating cer-vical spine trauma. J Trauma 40:130–137, 1996.

617. Prasad VS, Schwartz A, Bhutani R, Sharkey PW, Schwartz ML:Characteristics of injuries to the cervical spine and spinal cordin polytrauma patient population: Experience from a regionaltrauma unit. Spinal Cord 37:560–568, 1999.

618. Priebe MM, Waring WP: The interobserver reliability of therevised American Spinal Injury Association standards for neu-rological classification of spinal injury patients. Am J Phys MedRehabil 70:268–270, 1991.

619. Przybylski GJ, Clyde BL, Fitz CR: Craniocervical junction sub-arachnoid hemorrhage associated with atlanto-occipital dislo-cation. Spine 21:1761–1768, 1996.

620. Putnam WE, Stratton FT, Rohr RJ, Stitzell W, Roat G: Traumaticatlanto-occipital dislocations: Value of the Powers ratio in di-agnosis. J Am Osteopath Assoc 86:798–804, 1986.

621. Quencer RM, Bunge RP, Egnor M, Green BA, Puckett W,Naidich TP, Post MJ, Norenberg M: Acute traumatic centralcord syndrome: MRI-pathological correlations. Neuroradiol-ogy 34:85–94, 1992.

622. Quirke TE, Ritota PC, Swan KG: Inferior vena caval filter use inU.S. trauma centers: A practitioner survey. J Trauma 43:333–337, 1997.

623. Raila FA, Aitken AT, Vickers GN: Computed tomography andthree-dimensional reconstruction in the evaluation of occipitalcondyle fracture. Skeletal Radiol 22:269–271, 1993.

624. Ralston ME, Chung K, Barnes PD, Emans JB, Schutzman SA:Role of flexion-extension radiographs in blunt pediatric cervicalspine injury. Acad Emerg Med 8:237–245, 2001.

625. Ramon S, Dominguez R, Ramirez L, Paraira M, Olona M,Castello T, Garcia Fernandez L: Clinical and magnetic reso-nance imaging correlation in acute spinal cord injury. SpinalCord 35:664–673, 1997.

626. Ramsay AH, Waxman BP, O’Brien JF: A case of traumaticatlanto-occipital dislocation with survival. Injury 17:412–413,1986.

627. Raphael JH, Chotai R: Effects of the cervical collar on cerebro-spinal fluid pressure. Anaesthesia 49:437–439, 1994.

628. Rapp RP, Young B, Twyman D, Bivins BA, Haack D, Tibbs PA,Bean JR: The favorable effect of early parenteral feeding onsurvival in head-injured patients. J Neurosurg 58:906–912,1983.

629. Rathbone D, Johnson G, Letts M: Spinal cord concussion inpediatric athletes. J Pediatr Orthop 12:616–620, 1992.

630. Reid DC, Henderson R, Saboe L, Miller JD: Etiology and clinicalcourse of missed spine fractures. J Trauma 27:980–986, 1987.

631. Reines HD, Harris RC: Pulmonary complications of acute spi-nal cord injuries. Neurosurgery 21:193–196, 1987.

632. Reinges MH, Mayfrank L, Rohde V, Spetzger U, Gilsbach JM:Surgically treated traumatic synchondrotic disruption of theodontoid process in a 15-month-old girl. Childs Nerv Syst14:85–87, 1998.

633. Riley KO, May AK, Hadley MN: Neurological injury and nu-tritional support, in Batjer HH, Loftus CM (eds): Textbook ofNeurological Surgery: Principles and Practice. Philadelphia, Lip-pincott Williams & Wilkins (in press).

634. Rimel RW, Jane JA, Edlich RF: An educational training programfor the care at the site of injury of trauma to the central nervoussystem. Resuscitation 9:23–28, 1981.

635. Ripa DR, Kowall MG, Meyer PR Jr, Rusin JJ: Series of ninety-two traumatic cervical spine injuries stabilized with anteriorASIF plate fusion technique. Spine 16[Suppl 3]:S46–S55, 1991.

636. Rizzolo SJ, Piazza MR, Cotler JM, Balderston RA, Schaefer D,Flanders A: Intervertebral disc injury complicating cervicalspine trauma. Spine 16(6 Suppl):S187–S189, 1991.

637. Roberge RJ, Wears RC, Kelly M, Evans TC, Kenny MA, DaffnerRD, Kremen R, Murray K, Cottington EC: Selective applicationof cervical spine radiography in alert victims of blunt trauma: Aprospective study. J Trauma 28:784–748, 1988.

638. Robertson PA, Ryan MD: Neurological deterioration after re-duction of cervical subluxation: Mechanical compression bydisc tissue. J Bone Joint Surg Br 74B:224–227, 1992.

639. Rockswold GL, Seljeskog EL: Traumatic atlantocranial disloca-tion with survival. Minn Med 62:151–154, 1979.

640. Rockswold GL, Bergman TA, Ford SE: Halo immobilizationand surgical fusion: Relative indications and effectiveness inthe treatment of 140 cervical spine injuries. J Trauma 30:893–898, 1990.



641. Rodgers JA, Rodgers WB: Marginal mandibular nerve palsydue to compression by a cervical hard collar. J Orthop Trauma9:177–179, 1995.

642. Rodriguez DJ, Benzel EC: Nutritional support, in Benzel EC(ed): Spine Surgery: Techniques, Complication Avoidance, and Man-agement. New York, Churchill Livingstone, 1999, vol 2, pp1321–1331.

643. Rodriguez DJ, Benzel EC, Clevenger FW: The metabolic re-sponse to spinal cord injury. Spinal Cord 35:599–604, 1997.

644. Roozmon P, Gracovetsky SA, Gouw GJ, Newman N: Examin-ing motion in the cervical spine: Part I—Imaging systems andmeasurement techniques. J Biomed Eng 15:5–12, 1993.

645. Rorabeck CH, Rock MG, Hawkins RJ, Bourne RB: Unilateralfacet dislocation of the cervical spine. An analysis of the resultsof treatment in 26 patients. Spine 12:23–27, 1987.

646. Deleted in proof.647. Rosen PB, McSwain NE Jr, Arata M, Stahl S, Mercer D: Com-

parison of two new immobilization collars. Ann Emerg Med21:1189–1195, 1992.

648. Rosenberg J, Greenberg MK: Practice parameters: Strategies forsurvival into the nineties. Neurology 42:1110–1115, 1992.

649. Rosenfeld JF, Vaccaro AR, Albert TJ, Klein GR, Cotler JM: Thebenefits of early decompression in cervical spinal cord injury.Am J Orthop 27:23–28, 1998.

650. Rosner MJ: National acute spinal cord injury study of methyl-prednisolone or naloxone. Neurosurgery 28:628, 1991 (letter).

651. Rosner MJ: Methylprednisolone for spinal cord injury.J Neurosurg 77:324–327, 1992.

652. Rosner MJ: Treatment of spinal cord injury. J Neurosurg 80:954–955, 1994.

653. Ross SE, O’Malley KF, DeLong WG, Born CT, Schwab CW:Clinical predictors of unstable cervical spinal injury in multiplyinjured patients. Injury 23:317–319, 1992.

654. Rossitch E Jr, Oakes WJ: Perinatal spinal cord injury: Clinical,radiographic and pathologic features. Pediatr Neurosurg 18:149–152, 1992.

655. Roth EJ, Lawler MH, Yarkony GM: Traumatic central cordsyndrome: Clinical features and functional outcomes. ArchPhys Med Rehabil 71:18–23, 1990.

656. Roth BJ, Martin RR, Foley K, Barcia PJ, Kennedy P: Roentgen-ographic evaluation of the cervical spine: A selective approach.Arch Surg 129:643–645, 1994.

657. Roussi J, Bentolila S, Boudaoud L, Casadevall N, Vallee C,Carlier R, Lortat-Jacob S, Dizien O, Bussel B: Contribution ofd-dimer determination in the exclusion of deep venous throm-bosis in spinal cord injury patients. Spinal Cord 37:548–552,1999.

658. Roy-Camille R, Benazet JP, Saillant G, Henry P, Mamoudy P,Leonard P: Traumatic atlanto-occipital luxation: Value of newradiologic signs (apropos of 2 cases) [in French]. Rev ChirOrthop Reparatrice Appar Mot 72:303–309, 1986.

659. Roy-Camille R, Saillant G, Laville C, Benazet JP: Treatment oflower cervical spinal injuries: C3 to C7. Spine 17[Suppl 10]:S442–S446, 1992.


661. Rutledge G, Sumchai A: A safe method for transportation ofpatients with cervical spine injuries. Aeromed J Sept/Oct:33,1987.

662. Ryan MD, Henderson JJ: The epidemiology of fractures andfracture-dislocations of the cervical spine. Injury 23:38–40,1992.

663. Ryan MD, Taylor TK: Odontoid fractures: A rational approachto treatment. J Bone Joint Surg Br 64B:416–421, 1982.

664. Ryan MD, Taylor TK: Odontoid fractures in the elderly. J Spi-nal Disord 6:397–401, 1993.

665. Sabiston CP, Wing PC, Schweigel JF, Van Peteghem PK, Yu W:Closed reduction of dislocations of the lower cervical spine.J Trauma 28:832–835, 1988.

666. Salmon JH: Fractures of the second cervical vertebra: Internalfixation by interlaminar wiring. Neurosurgery 1:125–127, 1977.

667. Sandler AN, Tator CH: Effect of acute spinal cord compressioninjury on regional spinal cord blood flow in primates.J Neurosurg 45:660–676, 1976.

668. Sandler AN, Tator CH: Review of the effect of spinal cordtrauma on the vessels and blood flow in the spinal cord.J Neurosurg 45:638–646, 1976.

669. San Mateo County, California: EMS System Policy Memorandum#F-3A. 1991.

670. Savini R, Parisini P, Cervellati S: The surgical treatment of lateinstability of flexion-rotation injuries in the lower cervicalspine. Spine 12:178–182, 1987.

671. Savolaine ER, Ebraheim NA, Jackson WT, Rusin JJ: Three-dimensional computed tomography in evaluation of occipitalcondyle fracture. J Orthop Trauma 3:71–75, 1989.

672. Scarrow AM, Levy EI, Resnick DK, Adelson PD, Sclabassi RJ:Cervical spine evaluation in obtunded or comatose pediatrictrauma patients: A pilot study. Pediatr Neurosurg 30:169–175,1999.

672a. Schaefer DM, Flanders A, Northrup BE, Doan HT, OsterholmJL: Magnetic resonance imaging of acute cervical spine trauma.Spine 14:1090–1095, 1989.


674. Schellinger PD, Schwab S, Krieger D, Fiebach JB, Steiner T,Hund EF, Hacke W, Meinck HM: Masking of vertebral arterydissection by severe trauma to the cervical spine. Spine 26:314–319, 2001.

675. Schleehauf K, Ross SE, Civil ID, Schwab CW: Computed to-mography in the initial evaluation of the cervical spine. AnnEmerg Med 18:815–817, 1989.

676. Schlicke LH, Callahan RA: A rational approach to burst frac-tures of the atlas. Clin Orthop 154:18–21, 1981.

677. Schneider RC: A syndrome in acute cervical spine injuries forwhich early operation is indicated. J Neurosurg 8:360–367,1951.

678. Schneider R: “Hangman’s fracture” of the cervical spine.J Neurosurg 22:141–154, 1965.

679. Schneider RC, Cherry G, Pantek H: The syndrome of acutecentral cervical spinal cord injury. J Neurosurg 546–577, 1954.

680. Schneider RC, Crosby EC, Russo RH, Gosch HH: Traumaticspinal cord syndromes and their management. Clin Neurosurg20:424–492, 1973.

681. Schneider RC, Thompson JC, Bebin J: The syndrome of acutecentral cervical spinal cord injury. J Neurol Neurosurg Psychi-atry 21:216–227, 1958.

682. Schönhöfer PS: GM-1 ganglioside for spinal-cord injury. N EnglJ Med 326:493–494, 1992.

683. Schriger DL: Immobilizing the cervical spine in trauma: Shouldwe seek an optimal position or an adequate one? Ann EmergMed 28:351–353, 1996.

684. Schriger DL, Larmon B, LeGassick T, Blinman T: Spinal immo-bilization on a flat backboard: Does it result in neutral positionof the cervical spine? Ann Emerg Med 20:878–881, 1991.

Bibliography S195


685. Schwarz N: The fate of missed atlanto-axial rotatory subluxa-tion in children. Arch Orthop Trauma Surg 117:288–289, 1998.

686. Schwarz N, Buchinger W, Gaudernak T, Russe F, Zechner W:Injuries to the cervical spine causing vertebral artery trauma:Case reports. J Trauma 31:127–133, 1991.

687. Schwarz N, Genelin F, Schwarz AF: Post-traumatic cervicalkyphosis in children cannot be prevented by non-operativemethods. Injury 25:173–175, 1994.

688. Sears W, Fazl M: Prediction of stability of cervical spine fracturemanaged in the halo vest and indications for surgical interven-tion. J Neurosurg 72:426–432, 1990.

689. Sedlock DA, Laventure SJ: Body composition and resting en-ergy expenditure in long term spinal cord injury. Paraplegia28:448–454, 1990.

690. Sees DW, Rodriguez Cruz LR, Flaherty SF, Ciceri DP: The useof bedside fluoroscopy to evaluate the cervical spine inobtunded trauma patients. J Trauma 45:768–771, 1998.


692. Seljeskog EL: Non-operative management of acute upper cer-vical injuries. Acta Neurochir (Wien) 41:87–100, 1978.

693. Seljeskog EL, Chou SN: Spectrum of the hangman’s fracture.J Neurosurg 45:3–8, 1976.

694. Senter HJ, Venes JL: Loss of autoregulation and posttraumaticischemia following experimental spinal cord trauma.J Neurosurg 50:198–206, 1979.


696. Shacked I, Ram Z, Hadani M: The anterior cervical approach fortraumatic injuries to the cervical spine in children. Clin Orthop292:144–150, 1993.

697. Shaffer MA, Doris PE: Limitation of the cross table lateral viewin detecting cervical spine injuries: A retrospective analysis.Ann Emerg Med 10:508–513, 1981.

698. Shah S, Vanclay F, Cooper B: Improving the sensitivity of theBarthel Index for stroke rehabilitation. J Clin Epidemiol 42:703–709, 1989.

699. Shapiro SA: Methylprednisolone for spinal cord injury.J Neurosurg 77:324–327, 1992.

700. Shapiro SA: Management of unilateral locked facet of the cer-vical spine. Neurosurgery 33:832–837, 1993.

701. Shapiro S, Snyder W, Kaufman K, Abel T: Outcome of 51 casesof unilateral locked cervical facets: Interspinous braided cablefor lateral mass plate fusion compared with interspinous wireand facet wiring with iliac crest. J Neurosurg 91[Suppl 1]:19–24,1999.

702. Sharma BS, Mahajan RK, Bhatia S, Khosla VK: Collet-Sicardsyndrome after closed head injury. Clin Neurol Neurosurg96:197–198, 1994.

703. Shaw M, Burnett H, Wilson A, Chan O: Pseudosubluxation ofC2 on C3 in polytraumatized children: Prevalence and signifi-cance. Clin Radiol 54:377–380, 1999.

704. Sherk HH, Nicholson JT: Fractures of the atlas. J Bone JointSurg Am 52A:1017–1024, 1970.

705. Sherk HH, Nicholson JT, Chung SM: Fractures of the odontoidprocess in young children. J Bone Joint Surg Am 60A:921–924,1978.

706. Shirasaki N, Okada K, Oka S, Hosono N, Yonenobu K, Ono K:Os odontoideum with posterior atlantoaxial instability. Spine16:706–715, 1991.

707. Short D: Use of steroids for acute spinal cord injury must bereassessed. BMJ 321:1224, 2000 (letter).

708. Short DJ, El Masry WS, Jones PW: High dose methylpred-nisolone in the management of acute spinal cord injury: Asystematic review from a clinical perspective. Spinal Cord38:273–286, 2000.

709. Shoung HM, Lee LS: Anterior metal plate fixation in the treat-ment of unstable lower cervical spine injuries. Acta Neurochir(Wien) 98:55–59, 1989.

710. Shrosbree RD: Acute central cervical spinal cord syndrome:Aetiology, age incidence and relationship to the orthopaedicinjury. Paraplegia 14:251–258, 1977.

711. Shrosbree RD: Neurological sequelae of reduction of fracturedislocations of the cervical spine. Paraplegia 17:212–221, 1979.

712. Sim E, Schwarz N, Biowski-Fasching I, Biowski P: Color-codedDuplex sonography of vertebral arteries: 11 cases of blunt cer-vical spine injury. Acta Orthop Scand 64:133–137, 1993.

713. Sim E, Vaccaro AR, Berzlanovich A, Pienaar S: The effects of stagedstatic cervical flexion-distraction deformities on the patency of thevertebral arterial vasculature. Spine 25:2180–2186, 2000.

714. Smith MG, Bourn S, et al: Ties that bind: Immobilizing theinjured spine. J Emerg Med Serv JEMS 14:28–35, 1989.

715. Smith MD, Phillips WA, Hensinger RN: Fusion of the uppercervical spine in children and adolescents: An analysis of 17patients. Spine 16:695–701, 1991.

716. Song WS, Chiang YH, Chen CY, Lin SZ, Liu MY: A simplemethod for diagnosing traumatic occlusion of the vertebralartery at the craniovertebral junction. Spine 19:837–839, 1994.

717. Sonntag VKH: Management of bilateral locked facets of thecervical spine. Neurosurgery 8:150–152, 1981.

718. Spence KF Jr, Decker S, Sell KW: Bursting atlantal fractureassociated with rupture of the transverse ligament. J Bone JointSurg Am 52A:543–549, 1970.

719. Spencer JA, Yeakley JW, Kaufman HH: Fracture of the occipitalcondyle. Neurosurgery 15:101–103, 1984.

720. Spierings EL, Braakman R: The management of osodontoideum: Analysis of 37 cases. J Bone Joint Surg Br 64B:422–428, 1982.

721. Sponseller PD, Cass JR: Atlanto-occipital fusion for dislocationin children with neurologic preservation: A case report. Spine22:344–347, 1997.

722. Star AM, Jones AA, Cotler JM, Balderston RA, Sinha R: Imme-diate closed reduction of cervical spine dislocations using trac-tion. Spine 15:1068–1072, 1990.

723. Starr JK, Eismont FJ: Atypical hangman’s fractures. Spine 18:1954–1957, 1993.

724. Stauffer ES: Orthotics for spinal cord injuries. Clin Orthop102:92–99, 1974.

725. Stauffer ES, Kelly EG: Fracture-dislocations of the cervicalspine: Instability and recurrent deformity following treatmentby anterior interbody fusion. J Bone Joint Surg Am 59A:45–48,1977.

726. Stevens JM, Chong WK, Barber C, Kendall BE, Crockard HA: Anew appraisal of abnormalities of the odontoid process associ-ated with atlanto-axial subluxation and neurological disability.Brain 117:133–148, 1994.

727. Stineman MG, Marino RJ, Deutsch A, Granger CV, Maislin G: Afunctional strategy for classifying patients after traumatic spi-nal cord injury. Spinal Cord 37:717–725, 1999.

728. Stover S, Fine PR: Spinal Cord Injury: The Facts & Figures. Bir-mingham, University of Alabama, 1986.

729. Stroobants J, Fidlers L, Storms JL, Klaes R, Dua G, Van Hoye M:High cervical pain and impairment of skull mobility as the onlysymptoms of an occipital condyle fracture: Case report.J Neurosurg 81:137–138, 1994.



730. Stroobants J, Seynaeve P, Fidlers L, Klaes R, Brabants K, VanHoye M: Occipital condyle fracture must be considered in thepediatric population: Case report. J Trauma 36:440–441, 1994.

731. Subach BR, McLaughlin MR, Albright AL, Pollack IF: Currentmanagement of pediatric atlantoaxial rotatory subluxation.Spine 23:2174–2179, 1998.

732. Subach BR, Morone MA, Haid RW Jr, McLaughlin MR, RodtsGR, Comey CH: Management of acute odontoid fractures withsingle-screw anterior fixation. Neurosurgery 45:812–820, 1999.

733. Suechting RL, French LA: Posterior inferior cerebellar arterysyndrome following a fracture of the cervical vertebra.J Neurosurg 12:187–189, 1955.

734. Deleted in proof.735. Suter R, Tighe T, et al: Thoracolumbar spinal instability during

variations of the log-roll maneuver. Prehospital Disaster Med7:133–138, 1992.

736. Swain A, Dove J, Baker H: ABCs of major trauma: Trauma ofthe spine and spinal cord—Part I. BMJ 301:34–38, 1990.

737. Sweeney JF, Rosemurgy AS, Gill S, Albrink MH: Is the cervicalspine clear? Undetected cervical fractures diagnosed only atautopsy. Ann Emerg Med 21:1288–1290, 1992.

738. Swischuk LE, John SD, Hendrick EP: Is the open-mouth odon-toid view necessary in children under 5 years? Pediatr Radiol30:186–189, 2000.

738a. Sypert GW: Hangman’s fracture: Arguments in favor of surgi-cal therapy for type II and III according to Edwards and Levine.Surg Neurol 49:262, 1998 (comment).

739. Taggard DA, Menezes AH, Ryken TC: Treatment of Downsyndrome-associated craniovertebral junction abnormalities.J Neurosurg 93[Suppl 2]:205–213, 2000.

740. Tan ES, Balachandran N: Hangman’s fracture in Singapore(1975–1988). Paraplegia 30:160–164, 1992.

741. Tan E, Schweitzer ME, Vaccaro L, Spetell AC: Is computedtomography of nonvisualized C7-T1 cost-effective? J SpinalDisord 12:472–476, 1999.


743. Tator CH: Hemodynamic issues and vascular factors in acuteexperimental spinal cord injury. J Neurotrauma 9:139–141,1992.

744. Tator CH: Ischemia as a secondary neural injury, in SalzmanSK, Faden AI (eds): Neurobiology of Central Nervous SystemTrauma. New York, Oxford University Press, 1994, pp 209–215.

745. Tator CH: Experimental and clinical studies of the pathophys-iology and management of acute spinal cord injury. J SpinalCord Med 19:206–214, 1996.

746. Tator CH: Biology of neurological recovery and functional res-toration after spinal cord injury. Neurosurgery 42:696–708,1998.

747. Tator CH, Fehlings MG: Review of the secondary injury theoryof acute spinal cord trauma with emphasis on vascular mech-anisms. J Neurosurg 75:15–26, 1991.

748. Tator CH, Duncan EG, Edmonds VE, Lapczak LI, Andrews DF:Comparison of surgical and conservative management in 208patients with acute spinal cord injury. Can J Neurol Sci 14:60–69, 1987.

749. Tator CH, Duncan EG, Edmonds VE, Lapczak LI, Andrews DF:Changes in epidemiology of acute spinal cord injury from 1947to 1981. Surg Neurol 40:207–215, 1993.

750. Tator CH, Rowed DW, Schwartz MI: Sunnybrook Cord InjuryScales for assessing neurological injury and neurologic recov-ery, in Tator CH (ed): Early Management of Acute Spinal CordInjury. New York, Raven Press, 1982, pp 17–24.

751. Tator CH, Rowed DW, Schwartz ML, Gertzbein SD, BharatwalN, Barkin M, Edmonds VE: Management of acute spinal cordinjuries. Can J Surg 27:289–296, 1984.

752. Tehranzadeh J, Bonk RT, Ansari A, Mesgarzadeh M: Efficacy oflimited CT for nonvisualized lower cervical spine in patientswith blunt trauma. Skeletal Radiol 23:349–352, 1994.

753. Thibodeaux LC, Hearn AT, Peschiera JL, Deshmukh RM,Kerlakian GM, Welling RE, Nyswonger GD: Extracranial ver-tebral artery dissection after trauma: A 5-year review. Br J Surg84:94, 1997.

754. Todd JW, Frisbie JH, Rossier AB, Adams DF, Als AV, ArmeniaRJ, Sasahara AA, Tow DE: Deep venous thrombosis in acutespinal cord injury: A comparison of 125I fibrinogen leg scan-ning, impedance plethysmography and venography. Paraple-gia 14:50–57, 1976.

755. Toh E, Arima T, Mochida J, Omata M, Matsui S: Functionalevaluation using motor scores after cervical spinal cord injuries.Spinal Cord 36:491–496, 1998.

756. Tola JC, Holtzman R, Lottenberg L: Bedside placement of inferiorvena cava filters in the intensive care unit. Am Surg 65:833–838, 1999.

756a. Tomaio A, Kirshblum SC, O’Connor KC, Johnston M: Treat-ment of acute deep vein thrombosis in spinal cord injuredpatients with enoxaparin: A cost analysis. J Spinal Cord Med21:205–210, 1998.

757. Torreman M: Long-term prognosis of the hangman’s fracture[in Dutch]. Ned Tijdschr Geneeskd 134:1173–1176, 1990.

758. Toscano J: Prevention of neurological deterioration before admissionto a spinal cord injury unit. Paraplegia 26:143–150, 1988.

759. Totten VY, Sugarman DB: Respiratory effects of spinal immo-bilization. Prehosp Emerg Care 3:347–352, 1999.

760. Tracy PT, Wright RM, Hanigan WC: Magnetic resonance im-aging of spinal injury. Spine 14:292–301, 1989.

761. Traynelis VC: Evidence-based management of type II odontoidfractures. Clin Neurosurg 44:41–49, 1997.

762. Traynelis VC, Marano GD, Dunker RO, Kaufman HH: Trau-matic atlanto-occipital dislocation: Case report. J Neurosurg65:863–870, 1986.

763. Treloar DJ, Nypaver M: Angulation of the pediatric cervicalspine with and without cervical collar. Pediatr Emerg Care13:5–8, 1997.

764. Tuite GF, Veres R, Crockard HA, Peterson D, Hayward RD: Useof an adjustable, transportable, radiolucent spinal immobiliza-tion device in the comprehensive management of cervical spineinstability: Technical note. J Neurosurg 85:1177–1180, 1996.

765. Tuli S, Tator CH, Fehlings MG, Mackay M: Occipital condylefractures. Neurosurgery 41:368–377, 1997.

766. Tulyapronchote R, Selhorst JB, Malkoff MD, Gomez CR: De-layed sequelae of vertebral artery dissection and occult cervicalfractures. Neurology 44:1397–1399, 1994.

767. Turgut M, Akpinar G, Akalan N, Ozcan OE: Spinal injuries inthe pediatric age group: A review of 82 cases of spinal cord andvertebral column injuries. Eur Spine J 5:148–152, 1996.


769. Turnbull IM: Blood supply of the spinal cord: Normal andpathological considerations. Clin Neurosurg 20:56–84, 1973.

770. Turnbull IM, Brieg A, Hassler O: Blood supply of cervicalspinal cord in man: A microangiographic cadaver study.J Neurosurg 24:951–965, 1966.

771. Urculo E, Arrazola M, Arrazola M Jr, Riu I, Moyua A: Delayedglossopharyngeal and vagus nerve paralysis following occipitalcondyle fracture: Case report. J Neurosurg 84:522–525, 1996.

Bibliography S197



773. Vaccaro AR, Falatyn SP, Flanders AE, Balderston RA, NorthrupBE, Cotler JM: Magnetic resonance evaluation of the interver-tebral disc, spinal ligaments, and spinal cord before and afterclosed traction reduction of cervical spine dislocations. Spine24:1210–1217, 1999.

774. Vaccaro AR, Klein GR, Flanders AE, Albert TJ, Balderston RA,Cotler JM: Long-term evaluation of vertebral artery injuriesfollowing cervical spine trauma using magnetic resonance an-giography. Spine 23:789–795, 1998.

775. Valaskatzis EP, Hammer AJ: Fracture of the occipital condyle:A case report. S Afr Med J 77:47–48, 1990.

776. Vale FL, Burns J, Jackson AB, Hadley MN: Combined medicaland surgical treatment after acute spinal cord injury: Results ofa prospective pilot study to assess the merits of aggressivemedical resuscitation and blood pressure management.J Neurosurg 87:239–246, 1997.

777. Verbiest H: Anterolateral operations for fractures and disloca-tions in the middle and lower parts of the cervical spine: Reportof a series of forty-seven cases. J Bone Joint Surg Am 51A:1489–1530, 1969.

778. Verheggen R, Jansen J: Hangman’s fracture: Arguments in fa-vor of surgical therapy for type II and III according to Edwardsand Levine. Surg Neurol 49:253–262, 1998.

779. Verska JM, Anderson PA: Os odontoideum: A case report ofone identical twin. Spine 22:706–709, 1997.

780. Viccellio P, Simon H, Pressman BD, Shah MN, Mower WR,Hoffman JR: A prospective multicenter study of cervical spineinjury in children. Pediatrics 108:E20, 2001.

781. Vital JM, Gille O, Senegas J, Pointillart V: Reduction techniquefor uni- and biarticular dislocations of the lower cervical spine.Spine 23:949–955, 1998.

782. Vogel LC: Unique management needs of pediatric spinal cordinjury patients: Etiology and pathophysiology. J Spinal CordMed 20:10–13, 1997.

783. Waddell JP, Reardon GP: Atlantoaxial arthrodesis to treatodontoid fractures. Can J Surg 26:255–260, 1983.

784. Wagner FC Jr, Johnson RM: Cervical bracing after trauma. MedInstrum 16:287–288, 1982.

785. Wallace MC, Tator CH: Successful improvement of blood pres-sure, cardiac output, and spinal cord blood flow after experi-mental spinal cord injury. Neurosurgery 20:710–715, 1987.

786. Walsh M, Grant T, Mickey S: Lung function compromised byspinal immobilization. Ann Emerg Med 19:615–616, 1990.

787. Walters BC: Clinical practice parameter development in neuro-surgery, in Bean JR (ed): Neurosurgery in Transition: The Socio-economic Transformation of Neurological Surgery. Baltimore, Wil-liams & Wilkins, 1998, pp 99–111.

788. Walton G: A new method of reducing dislocation of cervicalvertebrae. J Nerv Ment Disord 20:609, 1893.

789. Walton R, DeSalvo JF, Ernst AA, Shahane A: Padded vs unpad-ded spine board for cervical spine immobilization. Acad EmergMed 2:725–728, 1995.

790. Wang GJ, Mabie KN, Whitehill R, Stamp WG: The nonsurgicalmanagement of odontoid fractures in adults. Spine 9:229–230,1984.


792. Wani MA, Tandon PN, Banerji AK, Bhatia R: Collet-Sicardsyndrome resulting from closed head injury: Case report.J Trauma 31:1437–1439, 1991.

793. Waring WP, Karunas RS: Acute spinal cord injuries and theincidence of clinically occurring thromboembolic disease. Para-plegia 29:8–16, 1991.

794. Washtenaw/Livingston County Medical Control Authority:Spinal Injury Assessment and Immobilization: EMS Protocols. AnnArbor, Washtenaw/Livingston County Medical Control Au-thority, 1997.

795. Wasserberg J, Bartlett RJV: Occipital condyle fractures diag-nosed by high-definition CT and coronal reconstructions. Neu-roradiology 37:370–373, 1995.

796. Watanabe M, Toyama Y, Fujimura Y: Atlantoaxial instability inos odontoideum with myelopathy. Spine 21:1435–1439, 1996.

797. Waters RL, Adkins RH, Yakura JS: Definition of complete spi-nal cord injury. Paraplegia 29:573–581, 1991.

798. Waters RL, Adkins RH, Yakura JS, Vigil D: Prediction of am-bulatory performance based on motor scores derived fromstandards of the American Spinal Injury Association. ArchPhys Med Rehabil 75:756–760, 1994.

799. Waters RL, Meyer PR Jr, Adkins RH, Felton D: Emergency,acute, and surgical management of spine trauma. Arch PhysMed Rehabil 80:1383–1390, 1999.

800. Watridge CB, Orrison WW, Arnold H, Woods GA: Lateralatlantooccipital dislocation: Case report. Neurosurgery 17:345–347, 1985.

801. Watson N: Venous thrombosis and pulmonary embolism inspinal cord injury. Paraplegia 6:113–121, 1968.

802. Watson N: Anti-coagulant therapy in the prevention of venousthrombosis and pulmonary embolism in the spinal cord injury.Paraplegia 16:265–269, 1978.

802a. Weingarden SI, Weingarden DS, Belen J: Fever and thrombo-embolic disease in acute spinal cord injury. Paraplegia 26:35–42, 1988.

803. Weinstein PR, Karpman RR, Gall EP, Pitt M: Spinal cord injury,spinal fracture, and spinal stenosis in ankylosing spondylitis.J Neurosurg 57:609–616, 1982.

804. Weller SJ, Rossitch E Jr, Malek AM: Detection of vertebralartery injury after cervical spine trauma using magnetic reso-nance angiography. J Trauma 46:660–666, 1999.

805. Wells JD, Nicosia S: Scoring acute spinal cord injury: A study ofthe utility and limitations of five different grading systems.J Spinal Cord Med 18:33–41, 1995.

806. Wessels LS: Fracture of the occipital condyle: A report of 3cases. S Afr Med J 28:155–156, 1990.

807. White AA, Panjabi MM: Clinical Biomechanics of the Spine. Phil-adelphia, J.B. Lippincott, 1990, ed 2.

808. White P, Seymour R, Powell N: MRI assessment of the pre-vertebral soft tissues in acute cervical spine trauma. Br J Radiol72:818–823, 1999.

809. Wholey MH, Bruwer AJ, Baker HL: The lateral roentgenogramof the neck (with comments on the atlanto-odontoid-basionrelationship). Radiology 71:350–356, 1958.

810. Wilberger JE: Immobilization and traction, in Tator CH, BenzelEC (eds): Contemporary Management of Spinal Cord Injuries: FromImpact to Rehabilitation (Neurosurgical Topics). Park Ridge,AANS, 2000, ed 2, pp 91–98.

811. Williams TG: Hangman’s fracture. J Bone Joint Surg Br 57B:82–88, 1975.

812. Willis BK, Greiner F, Orrison WW, Benzel EC: The incidence ofvertebral artery injury after midcervical spine fracture or sub-luxation. Neurosurgery 34:435–442, 1994.



813. Wilson JT, Rogers FB, Wald SL, Shackford SR, Ricci MA: Pro-phylactic vena cava filter insertion in patients with traumaticspinal cord injury: Preliminary results. Neurosurgery 35:234–239, 1994.

814. Winemiller MH, Stolp-Smith KA, Silverstein MD, TherneauTM: Prevention of venous thromboembolism in patients withspinal cord injury: Effects of sequential pneumatic compressionand heparin. J Spinal Cord Med 22:182–191, 1999.

815. Wing PC, Nance P, Connell DG, Gagnon F: Risk of avascularnecrosis following short term megadose methylprednisolonetreatment. Spinal Cord 36:633–636, 1998.

816. Wolf A, Levi L, Mirvis S, Ragheb J, Huhn S, Rigamonti D,Robinson WL: Operative management of bilateral facet dislo-cation. J Neurosurg 75:883–890, 1991.

817. Wood-Jones F: The ideal lesion produced by judicial hanging.Lancet 1:53, 1913.


819. Woodring JH, Lee C, Duncan V: Transverse process fractures ofthe cervical vertebrae: Are they insignificant? J Trauma 34:797–802, 1993.

820. Woodring JH, Selke AC Jr, Duff DE: Traumatic atlantooccipitaldislocation with survival. AJR Am J Roentgenol 137:21–24,1981.

821. Worsing RA Jr: Principles of prehospital care of musculoskel-etal injuries. Emerg Med Clin North Am 2:205–217, 1984.

822. Xiong XH, Bean A, Anthony A, Inglis G, Walton D: Manipula-tion for cervical spinal dislocation under general anaesthesia:Serial review for 4 years. Spinal Cord 36:21–24, 1998.

823. Yamaguchi N, Ikeda K, Ishise J, Yamashita J: Traumatic atlanto-occipital dislocation with long-term survival. Neurol Med Chir(Tokyo) 36:36–39, 1996.

824. Yamashita Y, Takahashi M, Sakamoto Y, Kojima R: Atlantoaxialsubluxation: Radiography and magnetic resonance imagingcorrelated to myelopathy. Acta Radiol 30:135–140, 1989.

825. Yarkony GM, Roth E, Lovell L, Heinemann A, Katz RT, Wu Y:Rehabilitation outcomes in complete C5 quadriplegia. Am JPhys Med Rehabil 67:73–76, 1988.

826. Yashon D, Tyson G, Vise WM: Rapid closed reduction of cer-vical fracture dislocations. Surg Neurol 4:513–514, 1975.

827. Yavuz N, Tezyurek M, Akyuz M: A comparison of two func-tional tests in quadriplegia: The Quadriplegia Index of Functionand the Functional Independence Measure. Spinal Cord 36:832–837, 1998.

828. Yelnik A, Dizien O, Bussel B, Schouman-Claeys E, Frija G,Pannier S, Held JP: Systematic lower limb phlebography inacute spinal cord injury in 147 patients. Paraplegia 29:253–260,1991.

829. Young W, Bracken MB: The Second National Acute Spinal CordInjury Study. J Neurotrauma 9[Suppl 1]:S397–S405, 1992.

830. Young B, Ott L, Rapp RP, Norton J: The patient with criticalneurological disease. Crit Care Clin 3:217–233, 1987.

831. Young B, Ott L, Twyman D, Norton J, Rapp R, Tibbs P, HaackD, Brivins B, Dempsey R: The effect of nutritional support onoutcome from severe head injury. J Neurosurg 67:668–676,1987.

832. Young WF, Rosenwasser RH, Getch C, Jallo J: Diagnosis andmanagement of occipital condyle fractures. Neurosurgery 34:257–261, 1994.

833. Zach GA, Seiler W, Dollfus P: Treatment results of spinal cordinjuries in the Swiss Paraplegic Centre of Basel. Paraplegia14:58–65, 1976.

834. Zampella EJ, Duvall ER, Langford KH: Computed tomographyand magnetic resonance imaging in traumatic locked-in syn-drome. Neurosurgery 22:591–593, 1988.

835. Zavanone M, Guerra P, Rampini P, Crotti F, Vaccari U: Trau-matic fractures of the craniovertebral junction: Management of23 cases. J Neurosurg Sci 35:17–22, 1991.

836. Zeidman SM, Ling GS, Ducker TB, Ellenbogen RG: Clinicalapplications of pharmacologic therapies for spinal cord injury.J Spinal Disord 9:367–380, 1996.

837. Zigler JE, Waters RL, Nelson RW, Capen DA, Perry J: Occipito-cervico-thoracic spine fusion in a patient with occipito-cervicaldislocation and survival. Spine 11:645–646, 1986.

838. Zimmerman E, Grant J, Vise WM, Yashon D, Hunt WE: Treat-ment of Jefferson fracture with a halo apparatus: Report of twocases. J Neurosurg 44:372–375, 1976.

Early representations of spinal vertebrae. Courtesy, Rare Book Room, NorrisMedical Library, Keck School of Medicine, University of Southern California,Los Angeles, California.

Bibliography S199

guidelines for management of acute cervical spinal injuries

Documents

public health service

palsy resolved uneventfully

san mateo county

magnetic resonance imaging

ammerman halo orthosis

milla mitchell

perot pl jr

leonardo da vinci