upper cervical spine traumas9ab564d99fcaaf1d.jimcontent.com/download/version/1435337413/… ·...

Review Article

Upper Cervical Spine Trauma

Abstract

Injuries to the upper cervical spine are potentially lethal; thus, fullcharacterization of the injuries requires an accurate history andphysical examination, and management requires an in-depthunderstanding of the radiographic projection of the craniocervicalcomplex. Occipital condyle fractures may represent major ligamentavulsions and may be highly unstable, requiring surgery.Craniocervical dissociation results from disruption of the primaryosseoligamentous stabilizers between the occiput and C2. Dynamicfluoroscopy can differentiate the subtypes of craniocervicaldissociation and help guide treatment. Management of atlas fracturesis dictated by transverse alar ligament integrity. Atlantoaxialdislocations are rotated, translated, or distracted and are treated witha rigid cervical orthosis or fusion. Treatment of odontoid fractures iscontroversial and dictated by fracture characteristics, patientcomorbidities, and radiographic findings. Hangman’s fractures of theaxis are rarely treated surgically, but atypical patterns and displacedfractures may cause neurologic injury and should be reduced andfused. Management of injuries to the craniocervical junction remainschallenging, but good outcomes can be achieved witha comprehensive plan that consists of accurate and timely diagnosisand stabilization of the craniocervical junction.

The craniocervical junction (CCJ)represents the complex transi-

tion between the cranium and theupper cervical spine. The CCJ pro-tects the brainstem, cranial nerves,and cranial blood supply while al-lowing complicated motion. Theanatomy in this area consists of twoprimary joints, the atlanto-occipital(AO) joint and the atlantoaxial (AA)joint, and the intrinsic and extrinsicstabilizing ligaments (Figure 1).Injury to the CCJ must be suspectedin all patients exposed to high-energy trauma; therefore, a thor-ough understanding of the CCJ’sanatomy and radiographic pro-jection is essential for proper diag-nosis and management of patientswith upper cervical spine traumainjuries (Table 1).

Anatomy

The occiput and atlas rotate throughmultiple osseous articulations. Theocciput-C1 joints are shallow condy-loid joints that provide some osseousstability. The dens extends craniallyfrom the axis to form a synovialarticulation with the posterior aspectof the anterior atlas. Laterally, pairedarthrodial synovial joints completethe C1 and C2 osseous articulation.Intrinsic ligaments, including the

joint capsules, provide most of thestability of the CCJ. From dorsal toventral, these structures consist ofthe tectorial membrane, the cruciateligament, and the alar ligament. Thetectorial membrane is the cranial ex-tension of the posterior longitudinal

718 Journal of the American Academy of Orthopaedic Surgeons

Richard J. Bransford, MD

Timothy B. Alton, MD

Amit R. Patel, MD

Carlo Bellabarba, MD

From the University of Washington/Harborview Medical Center, Seattle,WA (Dr. Bransford, Dr. Alton, andDr. Bellabarba), and the OSS HealthSpine Center, York, PA (Dr. Patel).

Dr. Bransford or an immediate familymember has received research orinstitutional support from DePuy.None of the following authors or anyimmediate family member hasreceived anything of value from or hasstock or stock options held ina commercial company or institutionrelated directly or indirectly to thesubject of this article: Dr. Alton,Dr. Patel, and Dr. Bellabarba.

J Am Acad Orthop Surg 2014;22:718-729

http://dx.doi.org/10.5435/JAAOS-22-11-718

Copyright 2014 by the AmericanAcademy of Orthopaedic Surgeons.

Copyright � the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited.

Figure 1

A, Illustration demonstrating sagittal view of the occipitocervical articulation. Posterior (B) and anterior (C) illustrations of theatlantoaxial articulation. AC = accessory ligament, AL = alar ligament, AP = apical ligament, TR = transverse atlantal ligament

Richard J. Bransford, MD, et al

November 2014, Vol 22, No 11 719


ligament, connecting the posterior axisto the anterior foramen magnum. Thecruciate ligament constrains the poste-rior dens to the anterior atlas archthrough its strongest component, thetransverse atlantal ligament (TAL), andvertical fibers extend from the foramenmagnum to the axis. Obliquely ori-entedalar ligaments connect thedens tothe lateral occipital condyles.Extrinsic ligaments also contribute

to the stability of the CCJ. They

include the ligamentum nuchae,connecting the external occipitalprotuberance to the posterior atlasand cervical spinous process, and thefibroelastic continuations of theanterior longitudinal ligaments andthe ligamentum flavum.The motion of the AO junction is

dictated by constraints, whereas sta-bility of theAAarticulation is providedprimarily by ligament restraints.1 Sag-ittal motion occurs primarily through

the AO joints, and rotary motion oc-curs primarily through the AA joints.Panjabi et al2 reported 27.1� of flexionand 24.9� of extension with only 8� ofrotation at the AO joint. Steinmetzet al3 noted that AO flexion waslimited by osseous contact betweenthe dens and foramen magnum,whereas extension was limited by thetectorial membrane. The AA articu-lation provides rotation, ranging from23.8� to 38.9�.3

Table 1

Summary of Craniocervical Junction Injury Type, Characteristics, and Treatment Based on Anatomic Location

Injury Type Characteristics Treatment

Occipital condylefracture

I Comminuted fracture of occipital condyle,axial load. Stable

Cervical collar, significant collapse treatedwith halo vest, CCD treated as such

II Extension of basilar skull fracture, shearinjury. Stable

Cervical collar, CCD treated as such

III Transverse fracture, avulsion of alarligament, consider CCD

Originally labeled unstable, commonlytreated with halo vest, commonlyassociated with CCD and treated as such

Craniocervicaldissociation

I MRI evidence of craniocervicalosseoligamentous injury, #2 mmdisplacement with traction test

Stable, no intervention required, cervicalcollar for comfort

II MRI evidence of craniocervicalosseoligamentous injury, .2 mmdisplacement with traction test

Rigid posterior segmental stabilization,hardware from occiput to C2 at least

III Craniocervical malalignment of.2 mmon static radiography

Rigid posterior segmental stabilization,hardware from occiput to C2 at least

Atlas fracture Stable Posterior arch fracture Soft versus rigid collar for comfort

Anterior arch avulsion fracture Rigid collar. If associated with CCD, treatas such

C1 ring fracture, LMD ,7 mm Rigid collar or halo vest

C1 ring fracture, LMD $7 mm Posterior C1-C2 arthrodesis comparedwith traction followed by halo vest

Unstable Anterior arch fracture with posteriordisplacement relative to the dens (ploughfracture)

Halo vest compared with posterior C1-C2arthrodesis

Unilateral sagittal split lateral mass fracture Recumbent traction ($3 wk) followed byhalo vest compared with C1 ring internalfixation

Atlantoaxial dislocation A Rotation centered on the dens, TALintact

Closed reduction and immobilization.Beware of associated fractures

B Translation betweenC1-C2, TAL disrupted Mid-substance TAL tears (type I): C1-C2arthrodesis, bony avulsions (type II):Posterior C1-C2 arthrodesis comparedwith halo vest after recumbent traction

C Distraction indicating CCD C1-C2 arthrodesis, consider occiput-C2 ifO-C1 joint involved. Akin to CCD

(continued)

ACDF = anterior cervical diskectomy and fusion, CCD = craniocervical dissociation, LMD = lateral mass displacement, TAL = transverse atlantalligament




Evaluation

Because injury and displacement atthe CCJ can have devastating con-sequences, instability should besuspected in all patients with high-energy injuries. Routine neurologicassessment should be performedaccording to the American SpinalInjury Association’s guidelines (ie,motor, sensory, reflexes, and spi-nal shock) and Advanced TraumaLife Support guidelines.

Diagnostic ImagingFor the past decade, routine traumascreening for cervical spine injury haschanged from the use of plain radi-ography to increased use of fine-cutCT with sagittal and coronal re-constructions because of its highersensitivity. However, a single cross-table lateral plain radiograph remainsan acceptable first-line radiographicscreening tool, assuming that theentire cervical spine can be visual-ized. Either of these studies allows for

the evaluation of the relevant anat-omy, including the presence ofa fracture or an altered spatial rela-tionship of the basion, opisthion,occipital condyles, dens, atlas, andaxis, or for evaluation of wideningor incongruence of the occipitocer-vical and AA joints (Figure 2). Inpatients with severe ligamentousinjury, supine imaging obtainedimmediately postinjury may showfalsely maintained alignment becauseof gravitational forces and muscle

Table 1 (continued )

Injury Type Characteristics Treatment

Dens fracture I Alar ligament insertion avulsion (cranialto TAL)

External immobilization compared withposterior stabilization if associatedwith CCD

II Waist of dens fracture at the level ofthe TAL

Controversial. Halo vest (if tolerable)compared with surgery. Surgicalindications: spinal cord injury, distracted,irreducible. Relative indications:displaced$5 mm, 10� angulation,delayed presentation (.2 wk), manynonunion risk factors, inability to tolerateexternal immobilization. High risk ofnonunion. Surgery: anterior screw (1versus 2) compared with posterior C1-C2arthrodesis. Elderly, compromisedconsider external immobilization iftolerable

III Fracture extending into cancellous bone ofthe C2 vertebral body

Halo vest compared with brace,uncommonly surgery (displaced.5 mm,spinal cord injury, external immobilizationintolerable, high nonunion risk)

Traumaticspondylolisthesis

I C2 arch fracture, displacement ,2 mm Rigid cervical orthosis versus halo vest

IA Atypical. Unilateral arch fracture withcontralateral vertebral body fractures.Vertebral foramen commonly involved.

Displacement may cause considerablecanal compromise/spinal cord injury.Treat with a halo vest compared withreduction/fixation if displaced or spinalcord injured (C2-C3ACDFcomparedwithposterior C1-C3 or C2-C3 arthrodesis)

II C2 arch fracture, displacement .2 mm Halo vest

IIA C2 arch fracture with C2-C3 intervertebraldisk disruption (angulation of C2-C3 endplates or anterior translation of C2 bodyon C3)

Halo vest. If markedly displaced, directfixationof fractured arch throughposteriorapproach versus posterior C2-C3 or C1-C3 arthrodesis versus C2-C3 ACDF

III C2 arch fracture with C2-C3 facet jointdislocation

Open reduction and posterior C2-C3 or C1-C3 arthrodesis

ACDF = anterior cervical diskectomy and fusion, CCD = craniocervical dissociation, LMD = lateral mass displacement, TAL = transverse atlantalligament




spasm, thus leading to confounding,false-negative findings. Prevertebralsoft-tissue widening can indicateunderlying trauma.Whereas CT allows for direct

visualization of the AO and AAjoints, AA joint congruity may bemore difficult to discern becauseof variations in normal rotationalalignment. However, AO joint con-gruity should be within 2 mm ofnormal. Because of the difficulty indirectly visualizing the AO and AAjoints with plain radiography, sev-eral indirect methods for assessingcraniocervical alignment have been

developed; these methods also maybe used with midsagittal CT imagesfor additional clarity. Harris linesprovide the most sensitive measure-ment of craniocervical instabilityand are often the most clinically rel-evant.4 The basion-dens interval(BDI) (ie, distance from the basionto the upper odontoid tip) should be#12 mm in 95% of patients. Thebasion-axis interval (BAI) (ie, dis-tance between the line projectingcranially from the posterior cortexof the C2 body to the basion) shouldbe #12 mm in 98% of the pop-ulation4 (Figure 2). The Wack-

enheim line is a straight projectionfrom the caudal posterior projectionof the clivus toward the upper cer-vical spine; it should be within 1 to 2mm of the tip of the odontoid. Theanterior spinal laminar line is drawnbetween the opisthion and theanterior cortex of the posterior archof the atlas and the C2 and C3laminae. With no fracture, this dis-tance should be 1 to 2 mm.The distance between the anterior

odontoid cortex and the posteriorcortex of the anterior arch of C1 (ie,the atlanto-dens interval [ADI])should be #3 mm; measurements of.3 to 5 mm indicate TAL insuffi-ciency and C1-C2 instability. Theposterior cortex of the atlas shouldparallel the anterior cortex of theaxis. V-shaped deformity between theanterior atlas and the odontoid withasymmetric widening of the 1-2 in-terspinous space may indicate cra-niocervical instability. The spaceavailable for the cord at this levelshould be $13 mm in adults.An open-mouth odontoid view or

coronal CT reformats of the uppercervical spine provide an AP pro-jection inwhich theoccipital condyles,lateralmasses ofC1, and the odontoidprocess are visualized. The lateral ADIand joint articulations shouldhave,2mm of asymmetry. Combined lateralmass overhang should be ,7 mm.5

Asymmetry or diastasis between theC1 lateral masses indicates a TALinjury (Figure 3) (Table 2).Any deviation from these initial

screening measures should promptfurther evaluation. Flexion-extensionradiographs to evaluate for instabilityprovide little additional informationandmay increase the risk of neurologicinjury.6

Dynamic FluoroscopyDynamic motion fluoroscopy undercontrolled conditions can help dif-ferentiate borderline cases of cranio-cervical instability and help guide

Figure 2

Illustration demonstrating radiographic projections of the craniocervicaljunction and useful measurements for evaluating lateral cervical spineradiographs in the setting of trauma. * = distance between the odontoid andthe Wackenheim line, BAI/BDI = Harris lines rule of 12. ADI = atlanto-densinterval, BAI = basion-axis interval, BDI = basion-dens interval, PADI =posterior atlanto-dens interval




treatment when advanced imagingresults are inconclusive, but it hasa minimal role in routine cervicalspine screening.7

Other ImagingIn patients with cervical trauma, cer-vical CT plays an integral role indiagnosis and surgical planning andhas supplanted plain radiography asthe first-line study for screening of thecervical spine in many institutions.Myelographic CT may be helpfulwhen MRI is contraindicated orunavailable, particularly in the pres-ence of neurologic deficits. Use ofMRI for evaluating the CCJ is indi-cated in patients with spinal cordinjury and for assessment of uppercervical spine ligament integrity.

Occipital CondyleFractures

Rapid, thin-slice CT scanning tech-niques are frequently used as primaryscreening tools for occipital condylefractures; open-mouth odontoid viewradiographs are also used, but may beless reliable (Figure 4). Managementis dictated by ligamentous injury andcraniocervical stability, which can beassessed with dynamic fluoroscopy toclarify the extent of the patient’sinjury. Saternus8 first proposed sixfracture patterns, a schema latercondensed by Anderson and Mon-tesano9 to three types: I, comminuted;II, basilar skull fracture; and III,avulsed.Collapsed type I and stable type III

injuries are most commonly man-aged nonsurgically with externalimmobilization. Surgery is reservedfor unstable injuries associated withcraniocervical dissociation (CCD).When injuries are unstable, surgicaltreatment requires rigid posteriorsegmental stabilization with instru-mentation from the occiput to atleast C2.

Type I (ie, impaction) injuries areroutinely stable and can be treatednonsurgically. Conversely, in type III(ie, avulsion) fractures it is imperativeto ensure the absence of associatedCCD; this injury combination requiresa posterior occipitocervical fusion. Ifa type III fracture is not associatedwithCCD, the injury may be treated non-surgically with external immobiliza-tion,usuallywitha rigid cervical collar.A type III fracture is not itself an indi-cation for surgery unless it is associatedwith a more significant ligamentousinjury.Nonsurgical management results in

mild neck disability regardless of theamount of displacement, patient gen-der, bilaterality, andpresenceof aheadinjury; greater disability is expected inpatients aged 40 to 60 years.10 Anoccipital condyle fracture associatedwith CCD is a predictor of a pooroutcome.11 Cranial nerve (ie, IX, X,XI, XII) injuries may occur.

Craniocervical Dissociation

CCD is commonly fatal,12 but ad-vancements in the prehospital care oftraumatized patients have improvedsurvival rates.7 Postmortem dis-sections, in which complete or nearcomplete disruption of all intrinsicand extrinsic CCJ ligamentousstructures were observed, have con-tributed to an initial understanding ofCCD and injury to the CCJ;13 rarecases of survival began to be reportedin the 1960s.14 As CCD survivorshipimproves, so too has awareness of thecondition and the necessity for accu-rate and timely diagnosis and stabi-lization of the CCJ.Diagnosis of CCJ disruption is

challenging, requiring a thoroughunderstanding of theCCJ osseous andligamentous anatomy, the radio-graphic projections of injury, and therecognition that plain radiographs orCT may produce false-negative find-ings. Bellabarba et al7 reported 17

survivors of CCD, of whom only twopatients (12%) were accuratelydiagnosed with initial lateral cervicalspine radiographs, despite retrospec-tive image review identifying injury in94% of patients using Harris linemeasurements. The average BDI andBAI at presentation were 17 mm and14 mm, respectively. The averagedelay in diagnosis was 2 days (13 of17 patients [76%]). Five of theseundetected injuries (38%) were rec-ognized only after the delayed onsetof neurologic deficit. Chaput et al15

reported 16 patients with CCD afterhigh-energy mechanisms, identifyingsix deaths with high cervical spinalcord transection. Thus, emergencydepartments and first-line providersof spine care must evaluate allhigh-energy polytraumatized patientsfor CCD.Providers may decrease morbidity

and mortality with a few simple in-terventions, consisting of placingsand bags around the patient’s head,taping the head and sand bags toa backboard, and, if tolerable, usingreverse Trendelenburg positioning.The positioning alerts all caregiversof the presence of a significantinjury, prevents further injury of the

Figure 3

Illustration demonstrating coronalview of the C1-C2 articulation,highlighting lateral mass overhang inthe setting of transverse alarligament rupture where total lateralmass displacement = a 1 b. Lateralmass displacement$7 mm indicatestransverse alar ligament rupture.




patient from movement, and coun-teracts dangerous distractive forces.A standard rigid cervical orthosisand full spine precautions alone maybe inadequate and even distractacross the injured CCJ.16 In patientswith severe CCJ displacement withneurologic deficits, halo vest con-structs can be applied, after manual,fluoroscopically guided reduction, toprovisionally stabilize the CCJ untildefinitive surgery can be performed.The halo vest must be used withcaution because of its potential dis-

tractive effect, and its use is generallylimited to provisional stabilizationafter closed reduction in patientswith severe displacement and neu-rologic deficits, pending urgentsurgical intervention. MRI of thecervical spine is recommended toevaluate for injury of the spinal cordand ligamentous injury.If CCJ alignment is maintained (ie,

,2 mm displacement) on plain ra-diographs and CT, but concern existsfor significant injury to the osseoliga-mentous structures of the CCJ on

MRI, provocative traction fluoroscopymay be performed to differentiatestable injuries from unstable injuries(ie, type I compared with type II) andto determine if occiput-to-cervicalposterior fusion is needed. Supervi-sion by qualified spine physicianswith expertise in management ofCCJ injuries is essential to minimizepatient risk while performing dynamicfluoroscopy.Baseline images are obtained with

the patient supine and the lateral pro-jectionC-arm centered onC1.Weight,

Table 2

Summary of Radiographic Parameters Used to Evaluate Alignment of the Craniocervical Junction onPlain Radiographs

RadiographicParameters How toPerformMeasurements What is Being Evaluated Normal Range

Prevertebral soft-tissue swelling

AP distance of the soft tissuesanterior to the cervical spine atC2/3 level

Increased swelling mayindicate upper cervical spineinjury

,7 mm at C2/3

Wackenheim line Straight projection from thecaudal posterior projection ofthe clivus toward the uppercervical spine

Reduction of the CCJ Line within 1 to 2 mm of odontoidtip

Anterior spinallaminar line

Straight projection from theopisthion and the anteriorcortex of the posterior arch ofthe atlas, C2 and C3 laminae

Alignment of the posteriormargin of the spinal canal

Linewithin 1 to2mmof each level

Atlanto-dens interval Distance between the anteriorodontoid cortex and posteriorcortex of the anterior arch of C1

.3 to 5 mm indicates TALdisruption and C1-2 instability

,3 mm

Basion-dens interval Distance from basion to upperodontoid tip

Alignment of the CCJ ,12 mm

Basion-axis interval Distance between line projectingcranially from the posteriorcortex of C2 body to the basion

Alignment of the CCJ 4 to 12 mm

Space available forcord

APdistance fromposterior cortexof the dens to theanterior cortexof the posterior arch of C1

Space available for the spinalcord at the level of C1

.13 mm

Lateral atlanto-densinterval

Distance between the lateralsurface of the dens and themedial surface of the lateralmass of C1

Atlantoaxial relationship ,2 mm of asymmetry

Combined lateralmass overhang

Combined horizontal distancefrom lateral border of C1 tolateral border of C2 on openmouth radiographs or coronalCT (Figure 3)

.7 mm indicates TALdisruption and C1-2 instability

,7 mm

Harris lines Measure the BAI and BDI Alignment of the CCJ Both lines,12 mm in adultsindicate normal CCJ alignment

AP = anteroposterior, BAI = basion-axis interval, BDI = basion-dens interval, CCJ = craniocervical junction, TAL = transverse atlantal ligament




starting at 5 lb, is added to Gardner-Wells tong traction and imaging isrepeated, evaluating for distractivechanges between the occiput, atlas,and axis. If no changes are identified,theweight is increased to10 lb, and theimaging is repeated. Fracture dis-placement .2 mm, atlanto-occipitaldistraction .2 mm, or atlantoaxialdistraction .3 mm indicates CCJinstability (Figure 5). Alternatively,manual traction may be applied undercontinuous fluoroscopy, which pro-vides for the above radiographic

findings combined with the tactilesensation of a firm end point or lackthereof. Despite the specific numericparameters described previously, thedifference between a positive test anda negative test is generally not subtle.In most patients, this test serves toavoid themorbidity of occipitocervicalfusion in those with MRI findings ofcraniocervical ligamentous injury.Traynelis et al17 described three

injury patterns based on the direc-tion of occiput displacement relativeto the upper cervical spine. Given

the complete ligamentous disrup-tion of the CCJ, the location ofdisplacement is of little clinical rel-evance and depends on forces beingapplied to the head and neck at thetime of imaging.The Harborview classification sys-

tem focuses on the degree of instabilityrather than direction of displacementand describes three injury patternswith therapeutic implications.7 Stage 1is stable, with minimally or non-displaced, often unilateral injury to thecraniocervical ligaments. Sufficient

Figure 4

Illustrations of occipital condyle fracture patterns. A, Type I, comminuted impaction fracture. B, Type 2, condyle fracture withassociated basilar skull fracture. C, Type 3, avulsion of the alar ligament attachment.

Figure 5

Lateral cervical spine fluoroscopic views demonstrating the positive provocative traction test of the craniocervical junction.A, Pre-traction occiput-C1 alignment. B, Traction view showing separation between occiput and C1 (double arrow).C, Postoperative radiograph of occiput to C2 posterior fusion in patient with craniocervical dissociation. (Reproduced withpermission from Bellabarba C, Mirza SK, Chapman JR: Injuries of the craniocervical junction, in Bucholz RW, ed: Rockwoodand Green’s Fractures in Adults, ed 6. Philadelphia, PA, Lippincott Williams & Wilkins, 2006, pp 1436-1496.)




ligamentous integrity remains tomaintain craniocervical stability, al-lowing for treatment by externalimmobilization alone. Stage 2 injuriesare minimally displaced on initialimaging and may resemble stage 1injuries. MRI may demonstrate sig-nificant soft-tissue injury and aid indiagnosis but will not indicate insta-bility. However, the unstable natureof stage 2 injuries can be demon-strated through a positive tractiontest, which indicates a partially orcompletely reduced, but highlyunstable, injury to the CCJ that re-quires surgical stabilization. Thisinjury type is the most problematicbecause its relatively nondisplacednature may create the dangerouscombination of a neurologically intactpatient with a deceptively well-aligned but highly unstable injury;the patient is at the highest risk forneurologic deterioration if notappropriately treated. It is likely thatthe increased number of survivors ofCCD injuries is largely accountedfor by this injury type because ofimprovements in prehospital andposthospital diagnosis and treat-ment. Harborview stage 3 injuriesare highly unstable, with gross cra-niocervical misalignment (ie, a BAIor BDI .2 mm beyond the upperlimits of normal). These injuries areusually fatal, and survivors tend tohave profound neurologic deficits.16

Stage 3 injuries generally do not pres-ent the same diagnostic and treatmentchallenges as do stage 2 injuries. Adiagnosis of CCD implies significantinstability; therefore, it is reserved forstage 2 and stage 3 injuries.In a study by Kazemi et al,18 14 of

28 patients with CCD were screenedfor blunt cerebrovascular injuries; 25patients had vascular injuries, 12patients had vertebral injuries, and13 patients had carotid artery in-juries. It is important to ensure thatall patients with CCD are screenedfor blunt cerebrovascular injuriesbecause perioperative management

and surgical plans may be altered inthe presence of vascular injury.Surgical stabilizationofCCDrequires

rigid posterior segmental stabilizationwith instrumentation from the occiputto at least C2, even when the primarydistraction is between the occiput andC1, because ligamentous stabilizersextend from the occiput to C2, essen-tially bypassing C1. Moreover, it is rel-atively common to encounter injuries inwhich there is distractive compromiseof both the occiput-C1 and C1-C2joints.7 Bellabarba et al16 reported on48 patients, none of whom had post-operative craniocervical pseudarthrosisor hardware failure. Fifty-five percentof patients with CCD had usefulpostoperative motor function (ie,American Spinal Injury Associationgrade D or E), compared with 26% ofpatients preoperatively.Patients who are diagnosed with

CCD and a cervical cord injury, whorequire cardiopulmonary resuscitation,and who have a Glasgow Coma Scalescore of 3 are not expected to survivethe injury.19

Atlas Fractures

Axial loading causes most C1 ringfractures, possibly injuring the TAL. Itis essential to obtain AP open-mouthodontoid views or coronal CT re-formats to determine the total lateraldisplacement of the C1 lateral massesrelative to the C2 lateral masses. Com-promise of the TAL generally occurswith total C1 lateralmass displacement(LMD) $7 mm.5 However, Dickmanet al20 reported that using the 7-mmthreshold in patients would have de-tected only 39% of atlas fractures withTAL disruption because of reboundafter the initial injury (Figure 3).The stability of isolated atlas frac-

tures is based on TAL integrity; thismay be determined directly by MRIvisualizationof the disrupted ligament,directly byCTdemonstrating ligamentavulsion fragments, or indirectly by

measuring the lateral mass overhangor evaluating the ADI (maximumADIfor adults with an intact TAL is 3 mm;in children, it is 5 mm).Patients with isolated arch fractures

and ring fractures with an LMD ,7mm are treated with a rigid collar.When the LMD is $7 mm or ifadvanced imaging studies suggestdisruption of the TAL, surgery isindicated, although halo vest immo-bilization for 3 months may be at-tempted, usually accompanied byrecumbent traction for up to 6 weeks.If radiographs in the halo vest revealan ADI $4 mm or further LMD,surgical stabilization is indicated.21

Surgery is reserved for patients withTAL disruption because of the poten-tial for further lateral mass widening,unacceptable C1-C2 or occipital-C1alignment, and pseudarthrosis. Con-troversy remains as to what sub-stantiates C1-C2 instability. In thepresence of atlas fractures, instabilitycan be seenwith unacceptable bilateraloverhang of the C1 over the C2 lateralmasses or with a unilateral sagittalsplit that allows asymmetric sub-luxation of the C1 lateral mass, thuspermitting the occipital condyle tosettle onto the C2 lateral mass andcreate a cock-robin–type deformity.22

Surgical options include C1-C2transarticular screw arthrodesis orsegmental fixationwithC1 lateralmassscrews and C2 pedicle or translaminarscrews connected by plates or rods, al-lowing for fracture reduction. To pre-serve C1-C2 motion, internal fixationof C1 alone has been proposed, eitheranterior or posterior (the most com-mon), but indications remain unclear.Considerable treatment challenges

exist when unilateral sagittal C1 frac-ture variants maintain TAL integritybut the fracture fragment is lateral tothe TAL insertion and at risk forcontinued lateral displacement. Ratesof unacceptable outcomes are highwith external immobilization alone;the resulting upper cervical deformitymay require a complex upper cervical




osteotomy for correction.22 Recom-mended management options includeextended recumbent cranial traction(ie, $3 weeks), followed by halo vestimmobilization, internal C1 ring fix-ation to avoid the need for fusion,compared with C1-2 posterior fusionwhen both C1 lateral masses areamenable to screw fixation or occipi-tocervical fusion when there is toomuch comminution to accommodateC1 lateral mass screw fixation.23 Iso-lated horizontal C1 fractures may beassociated with CCD, thus promptingclose evaluation of the CCJ.24

In atlas fractures, severe complica-tions are rare and neurologic sequelaeare uncommon. External immobiliza-tion of isolated atlas fractures yieldsgood radiographic results, with rarecases of nonunion or instability, butmost patients continue to have neckpain.25 Dvorak et al26 reported that anLMD$7 mm is associated with worselong-term outcomes and that 91% ofpatients did not believe they had re-turned to their preinjury state of health;this puts into question whether exter-nal immobilization is the best treat-ment. Severe displacement of a lateralmass fracture and healing in a dis-placed position may result in painfultorticollis, requiring realignment andfusion from the occiput to C2.

Atlantoaxial Instability

Three injury patterns, designated as A,B, and C, may occur either in isolationor in combination. Type A injuries arerotationally displaced in the transverseplane, typeB injuries are translationallyunstable as a result of TAL disruption,and type C injuries are a variant ofCCD and are vertically unstable.27

Type A instability is often non-traumatic, but mild rotational sub-luxation and complete dislocationhave been described. Because criticalCCJ ligaments are intact, treatmentincludes reduction and immobiliza-tion. Survivorsof typeB injuries,whichare commonly fatal, generally requireC1-C2 arthrodesis. In nondisplacedor minimally displaced fractures withbonyTAL insertionavulsions, a periodof recumbent traction followed by haloimmobilization is an option.20 If non-surgical management is attemptedand instability remains after 3 months(ie, flexion/extension ADI .3 mm),arthrodesis is indicated. Type Cdistraction injuries in which C1-C2distraction is .2 mm are analogousto CCD and are treated similarly,with C1-C2 posterior arthrodesis; iffindings show occiput-C1 jointinvolvement, occiput-C2 stabiliza-tion is warranted.

TraumaticSpondylolisthesis

Traumatic spondylolisthesisof theaxis(ie, hangman’s fracture) is the secondmost common fracture of the axis(38%), usually resulting from hyper-extension and axial loading (Figure 6).The classification system by Effendi,modified by Levine and Starr, de-scribes five injury patterns.28-30 Type Ipatterns are minimally displaced parsinterarticularis fractures with trans-lation,3mm; this must be verified onupright views. Subtype IA are atypi-cal, unstable fractures from side-bending forces, typically creating anoblique fracture through one pars in-terarticularis and anterior to the parswithin the body of the contralateralside. Type II injuries occur whenflexion follows hyperextension andaxial loading. Type II injuries mayappear similar to type I injuries onsupine imaging, but upright radio-graphs in a rigid collar show dis-placement .3 mm. The type IIAsubgroup, which results from flex-ion and distraction, has a morehorizontal fracture pattern, kypho-sis relatively greater than trans-lation, and associated C2-C3 diskand posterior longitudinal ligamentinjuries. Type III patterns are similar

Figure 6

Illustrations demonstrating five injury patterns in traumatic spondylolisthesis. A, Type I.B, Type IA. C, Type II.D, Type IIA. E,Type III.




to type I patterns with respect to thepars fracture pattern but are asso-ciated with dislocation of the C2/C3facet joints and are usually irre-ducible without posterior surgicalintervention.Surgical treatment of hangman’s

fractures is rare. Patients with type Iand IA injuries without neurologiccompromise typically do well withambulatory immobilization for 12weeks, although atypical fracturesmay be unpredictable and should befollowed closely. Type II injuries aretreated with halo immobilization for12 weeks, whereas type IIA injuriesgenerally require surgery consisting ofC2-C3 anterior cervical diskectomyand fusion or posterior fixation.Whereas anterior C2-3 fixation pre-serves atlantoaxial motion becausethe anterior longitudinal ligament andpossibly the anterior annulus are theonly intact C2-C3 stabilizing struc-tures, posterior fixation is also possi-ble and is more biomechanicallysound. However, additional motionmay be lost if posterior C2 screwpurchase across the fracture is inade-quate; the instrumentation must beextended cranially to include C1.Type III injuries usually require openposterior reduction and stabilization.Stabilization options include posteriorC1-C3 fusion, compared with possi-ble approaches that preserve C1-2motion, such as posterior C2-C3fusion with interfragmentary screwsacross the C2 fracture or conversionof a type III pattern to a type I or IIpattern (ie, fusing the C2-C3 facetsusing a C2 screw and stopping shortof the fracture), followed by externalimmobilization with a halo vest orcervical collar. Because the pars frac-ture is usually minimally displaced intype III hangman’s fractures, posteriorinterfragmentary fixation across C2with C2-3 posterior arthrodesis isusually feasible.Neurologic injury is uncommon in

hangman’s fractures unless the injurypattern causes narrowing of the spinal

canal, generally restricted to type IA(33%) and type III (60%) injuries.31

Other injury patterns result in wid-ening of the spinal canal and thusrarely result in neurologic compro-mise, although severely displaced typeII injuries may result in compressionbetween the posterior arch of C1 andthe posterosuperior aspect of C3.Nonsurgical management results in

successful healing rates approaching95%. Type IA, IIA, and III injuries aremorechallenging tomanagebecauseoftheir displacement patterns and inher-ent instability, but anterior or posteriorfixation yields good results. Yinget al32 reported a fusion rate of 100%at 6 months in 30 patients whounderwent C2-C3 anterior cervicaldiskectomy and fusion for type II, IIA,and III traumatic spondylolisthesis.Xu et al33 reported equivalent resultswith similar treatment in 28 patientswith unstable injury. Ma et al34 re-ported a 100% healing rate at 6months in 35 patients with unstabletraumatic spondylolisthesis who weretreated with posterior C2-C3 fixation.In our experience, the need for surgicalintervention in type II injuries is rare.

Summary

Injuries to the upper cervical spineinclude awide spectrumof pathology,from benign to life threatening.Occipital condyle fractures may rep-resent major ligament avulsions, buttreatment is typically nonsurgical.Craniocervical dissociation resultsfrom disruption of the primary os-seoligamentous stabilizers betweenthe occiput and C2. Dynamic fluo-roscopy can help guide treatment.Management of atlas fractures is dic-tated by transverse alar ligamentintegrity, whereas treatment of odon-toid fractures is controversial anddictated by fracture characteristics,patient comorbidities, and radio-graphic findings. Atlantoaxial dis-locations are rotated, translated,

or distracted and are treated witha rigid cervical orthosis or fusion.Hangman’s fractures of the axis arerarely treated surgically; however,atypical patterns and displaced frac-tures may cause neurologic injury andtherefore should be reduced andfused. An in-depth understanding ofCCJ anatomy and its radiographicprojections is essential for evaluatingconditions related to trauma of theupper cervical spine, making a properinitial diagnosis, and subsequentlyproviding definitive management ofthese injuries.

References

Evidence-based Medicine: Levels ofevidence are described in the table ofcontents. In this article, references 4,6, 10, and 11 are level III studies.References 5, 7, 9, 13-20, 23-26, 31,32, and 34 are level IV studies.

References printed in bold type arethose publishedwithin thepast 5 years.

1. Tubbs RS, Hallock JD, Radcliff V, et al:Ligaments of the craniocervical junction.J Neurosurg Spine 2011;14(6):697-709.

2. Panjabi M, Dvorak J, Crisco J III, Oda T,Hilibrand A, Grob D: Flexion, extension,and lateral bending of the upper cervicalspine in response to alar ligamenttransections. J Spinal Disord 1991;4(2):157-167.

3. Steinmetz MP, Mroz TE, Benzel EC:Craniovertebral junction: Biomechanicalconsiderations. Neurosurgery 2010;66(suppl 3):7-12.

4. Harris JH Jr, Carson GC, Wagner LK:Radiologic diagnosis of traumaticoccipitovertebral dissociation: 1. Normaloccipitovertebral relationships onlateral radiographs of supine subjects.AJR Am J Roentgenol 1994;162(4):881-886.

5. Spence KF Jr, Decker S, Sell KW: Burstingatlantal fracture associated with rupture ofthe transverse ligament. J Bone Joint SurgAm 1970;52(3):543-549.

6. Pollack CV Jr, Hendey GW, Martin DR,Hoffman JR, Mower WR; NEXUS Group:Use of flexion-extension radiographs of thecervical spine in blunt trauma. Ann EmergMed 2001;38(1):8-11.

7. Bellabarba C, Mirza SK, West GA, et al:Diagnosis and treatment of craniocervical




dislocation in a series of 17consecutive survivors during an 8-yearperiod. J Neurosurg Spine 2006;4(6):429-440.

8. Saternus KS: Forms of fractures of theoccipital condyles. Z Rechtsmed 1987;99(2):95-108.

9. Anderson PA, Montesano PX: Morphologyand treatment of occipital condylefractures. Spine (Phila Pa 1976) 1988;13(7):731-736.

10. Maddox JJ, Rodriguez-Feo JA III,Maddox GE, Gullung G, McGwin G,Theiss SM: Nonoperative treatment ofoccipital condyle fractures: Anoutcomes review of 32 fractures. Spine(Phila Pa 1976) 2012;37(16):E964-E968.

11. Mueller FJ, Fuechtmeier B, Kinner B, et al:Occipital condyle fractures: Prospectivefollow-up of 31 cases within 5 years ata level 1 trauma centre. Eur Spine J 2012;21(2):289-294.

12. Davis D, Bohlman H, Walker AE, Fisher R,Robinson R: The pathological findings infatal craniospinal injuries. J Neurosurg1971;34(5):603-613.

13. Bucholz RW, Burkhead WZ: Thepathological anatomy of fatal atlanto-occipital dislocations. J Bone Joint Surg Am1979;61(2):248-250.

14. Gabrielsen TO, Maxwell JA: Traumaticatlanto-occipital dislocation; with casereport of a patient who survived. Am JRoentgenol Radium Ther Nucl Med 1966;97(3):624-629.

15. Chaput CD, Torres E, Davis M, Song J,Rahm M: Survival of atlanto-occipitaldissociation correlates with atlanto-occipital distraction, injury severity score,and neurologic status. J Trauma 2011;71(2):393-395.

16. Bellabarba C, Bransford RJ, Chapman JR:Timing to diagnosis and neurologicaloutcomes in 48 consecutive craniocervicaldissociation patients. Spine J 2011;11(suppl 10):S57-S57.

17. Traynelis VC, Marano GD, Dunker RO,Kaufman HH: Traumatic atlanto-occipitaldislocation: Case report. J Neurosurg 1986;65(6):863-870.

18. Kazemi N, Bellabarba C, Bransford R,Vilela M: Incidence of bluntcerebrovascular injuries associated withcraniocervical distraction injuries. EvidBased Spine Care J 2012;3(4):63-64.

19. Cooper Z, Gross JA, Lacey JM, Traven N,Mirza SK, Arbabi S: Identifying survivorswith traumatic craniocervical dissociation:A retrospective study. J Surg Res 2010;160(1):3-8.

20. Dickman CA, Greene KA, Sonntag VK:Injuries involving the transverse atlantalligament: Classification and treatmentguidelines based upon experience with 39injuries. Neurosurgery 1996;38(1):44-50.

21. Kakarla UK, Chang SW, Theodore N,Sonntag VK: Atlas fractures. Neurosurgery2010;66(suppl 3):60-67.

22. Bransford R, Falicov A, Nguyen Q,Chapman J: Unilateral C-1 lateral masssagittal split fracture: An unstable Jeffersonfracture variant. J Neurosurg Spine 2009;10(5):466-473.

23. Bransford R, Chapman JR, Bellabarba C:Primary internal fixation of unilateral C1lateral mass sagittal split fractures: A seriesof 3 cases. J Spinal Disord Tech 2011;24(3):157-163.

24. Vilela MD, Bransford RJ, Bellabarba C,Ellenbogen RG: Horizontal C-1 fractures inassociation with unstable distractioninjuries of the craniocervical junction.J Neurosurg Spine 2011;15(2):182-186.

25. Levine AM, Edwards CC: Fractures of theatlas. J Bone Joint Surg Am 1991;73(5):680-691.

26. Dvorak MF, Johnson MG, Boyd M,Johnson G, Kwon BK, Fisher CG: Long-termhealth-related quality of life outcomesfollowing Jefferson-type burst fractures of theatlas. J Neurosurg Spine 2005;2(4):411-417.

27. Fielding JW, Hawkins RJ: Atlanto-axialrotatory fixation. (Fixed rotatorysubluxation of the atlanto-axial joint). JBone Joint Surg Am 1977;59(1):37-44.

28. Effendi B, Roy D, Cornish B, Dussault RG,Laurin CA: Fractures of the ring of the axis:A classification based on the analysis of 131cases. J Bone Joint Surg Br 1981;63(3):319-327.

29. Levine AM, Edwards CC: Themanagement of traumaticspondylolisthesis of the axis. J Bone JointSurg Am 1985;67(2):217-226.

30. Starr JK, Eismont FJ: Atypical hangman’sfractures. Spine (Phila Pa 1976) 1993;18(14):1954-1957.

31. Ding T, Maltenfort M, Yang H, et al:Correlation of C2 fractures and vertebralartery injury. Spine (Phila Pa 1976) 2010;35(12):E520-E524.

32. Ying Z, Wen Y, Xinwei W, et al: Anteriorcervical discectomy and fusion for unstabletraumatic spondylolisthesis of the axis.Spine (Phila Pa 1976) 2008;33(3):255-258.

33. Xu H, Zhao J, Yuan J, Wang C: Anteriordiscectomy and fusion with internalfixation for unstable hangman’s fracture.Int Orthop 2010;34(1):85-88.

34. Ma W, Xu R, Liu J, et al: Posterior short-segment fixation and fusion inunstable Hangman’s fractures.Spine (Phila Pa 1976) 2011;36(7):529-533.




upper cervical spine traumas9ab564d99fcaaf1d.jimcontent.com/download/version/1435337413/… ·...

Documents