imaging ofthe spinal cord

J7ournal ofNeurology, Neurosurgery, and Psychiatry 1995;58:403-416

NEUROLOGICAL INVESTIGATIONS

Imaging of the spinal cord

John M Stevens

Methods of investigationHISTORICAL PERSPECTIVEAir myelographyThe first contrast medium used to show thespinal cord was air. The first reports of its useto localise intraspinal tumours came fromJacobeus in 1921 and Dandy in 1925.1 Thetechnique was later refined to show the entirespinal canal; this involved complete replace-ment of the CSF by air and distension of thespinal subarachnoid space.2 Adequate visuali-sation of the spinal cord usually requiredtomography. Spinal roots were not shown,and many types of pathology such as vascularmalformations or arachnoiditis were eithernot shown or easily misinterpreted. Despitethe fact that the technique was difficult toachieve and very hard on the patient, itremained the preferred method for visualisingthe spinal cord in many centres until theadvent of non-toxic water soluble contrastonly just over 10 years ago.

MRI Unit, Basementof Clarence Wing, StMary's Hospital,Praed Street, LondonWI INY, UKJ M Stevens

Oil myelographyAt the time air myelography was being devel-oped, it was found accidentally that iodisedoils could be moved through the spinal sub-arachnoid space under the influence of grav-ity. They proved easier to use than air, andthis quickly established oil myelography as

the technique of choice especially in the lum-bar spinal canal. It received even greaterimpetus from the appearances in 1940, ofiophendylate (Myodil; Pantopaque), a prepa-ration that was less viscous than the earlierLipiodol and better for demonstrating thespinal cord. Myodil was very opaque to x raysand special techniques such as tomographywere not required. It was very slowlyabsorbed and could be left in the canal andrerun postoperatively to check the adequacyof decompression. Its main disadvantage was

immiscibility with CSF; it tended to break upinto globules, forming a layer in the spinalcanal, which made it difficult to demonstrateboth anterior and posterior surfaces of thespinal cord unless large amounts were used.For nearly 40 years, Myodil was generally theagent of choice, with air being reserved forspecial situations, such as spinal dysraphismand syringomyelia. As late as 1989, eminentnames in spinal surgery were still declaringtheir preference for Myodil, mainly becauseof experiences with non-diagnostic water sol-

uble myelograms in which the contrastmedium had become too dilute. Myodil isnow no longer manufactured, and existingstocks have been withdrawn because of thefrequency with which it caused chronic adhe-sive arachnoiditis.34

WATER SOLUBLE MYELOGRAPHYThe advantages of water soluble over oilycontrast media were established by use of asubstance called Abrodil in Scandinaviancountries. This provided superior images ofthe cauda equina and root sheaths. Otherionic water soluble media such as Conray andDimer-X enjoyed limited use, but all weretoo toxic to use other than to show the lowerlumbar thecal sac. They are, therefore, irrele-vant to the present review. A revolution inmyelography occurred when the new non-ionic water soluble medium metrizamide(Amipaque) appeared. This was far less neu-rotoxic than previous ones. It could be usedsafely around the spinal cord. Inadvertentdeposition in the head when running the con-trast medium into the cervical region oftencaused generalised seizures, however; a riskminimised by introducing it by lateral C1-2puncture.5By 1983, similar and even less toxic media

had been developed and metrizamide wasquickly withdrawn. These agents, such asiohexol, do not cause arachnoiditis in theconcentrations used in clinical practice, andby 1989, Skalpe and Sortland were able tostate that "epileptic seizures have not beenreported following myelography withOmnipaque [iohexol], so it seems that fear ofthis complication can be virtually disre-garded."6

These media are still not perfect: mostpatients experience postmyelography head-ache and about 5% become confused ordevelop symptoms such as radicular pain ormeningism.7 Further agents have been devel-oped, but clinical trials are now difficult tomount because myelography is so little used.Iohexol should be injected by lumbar punc-ture wherever possible to reduce the risk ofinjury to the spinal cord or vertebral artery inlateral cervical puncture.8 Myelography maycause neurological deterioration due to thespinal puncture, causing injury to the spinalcord or intraspinal bleeding; the disease,resulting in increased cord compression during

403

group.bmj.com on April 8, 2018 - Published by http://jnnp.bmj.com/Downloaded from

http://jnnp.bmj.com/

http://group.bmj.com

Stevens

the positioning required for radiography; theprocedure, causing raised intraspinalpressure below a subarachnoid block.7 8

X RAY COMPUTED TOMOGRAPHY (CT) OF THESPINESoon after its introduction for cranial imagingCT was used in the spine. It provided thevaluable cross sectional perspective of thespinal canal, formerly very difficult to achievewith conventional radiography. Artefactscaused by bone degrade intraspinal contrastin spinal CT. Intravenous contrast mediaincrease contrast between extradural tissuesand CSF, and intrathecal contrast media pro-vide excellent visualisation of the spinal cordand other intradural structures. Patientsrequiring myelography usually are booked forCT as well, the findings on the myelogram orthe clinical features serving to direct the CTexaminations to specific levels. Computedtomography is almost exclusively a cross sec-tional technique; sagittal and coronal projec-tions require reformatting from stacks of axialslices, and resolution in reformatted planes isnot as good as in the plane of data acquisi-tion. Radiation dose can be very considerable,and new guidelines for the use of CT inBritain have been published recently.9

MAGNETIC RESONANCE IMAGING (MRI)During the period when myelography wasbeing greatly improved, MRI appeared, andwithin just a few years myelography waspushed almost into obsolescence. With mod-em high resolution MRI, almost all intraduralfeatures demonstrable by myelography can beshown, usually better, by MRI. Myelographyis now indicated only when satisfactory MRIcannot be obtained because it is contraindi-cated (pacemakers, mechanical heart valves,aneurysm clips); it cannot be done (claustro-phobia and anaesthesia refused or unsafe,patient cannot fit into the magnet due to obe-sity, scoliosis, or limb contractures); or it isnot available quickly enough, due to lack ofon call service or other logistical problems.

Several recent developments have had aspecial impact on MRI of the spinal cord.

Volumetric (3 D) acquisitionsThe deployment of fast image techniques haspermitted three dimensional spatial encodingwithin a few minutes. This results in multiplecontiguous images, no interslice gaps, andsection thicknesses down to 1 or 2 mm. Thebest images of the spinal cord structure are,however, still usually obtained from thickerslices. The best results from volumetric MRIacquisitions are usually obtained on high-fieldmachines.

Fast spin echo (FSE)Several phase encoding steps are made ateach excitation instead of just one, whichgreatly reduces acquisition times and permitsthe use of much larger matrices. This resultsin twice the resolution in even shorter dataacquisition periods. The penalty is slight lossof contrast and increased sensitivity to physi-

ological motion, for which it is more difficultto compensate than when single phase encod-ing steps are used.

Phased array coilsSpinal imaging requires surface coils, and thephased array configuration enables data to beacquired simultaneously from two or evenmore surface coils. This permits imaging ofthe entire spinal cord over one acquisitionperiod and greatly reduces imaging time.Vertebral level counting, sometimes difficultor impossible from single coils, especially inthe thoracic region, also becomes easy.A wide variety of postprocessing options

are available on most imagers, or can be pur-chased separately. These permit multiplanarreformatting in real time, three dimensionalsurface rendering, colour coding, and manyother modifications. The ability to recon-struct in a curved plane is potentially useful inscoliosis, but we have found this of only lim-ited value because most major curvatures arein more than one plane.

ArtefactsArtefacts are important to consider becausethey can closely simulate intraspinal disease.Phase dispersion across the image, due to mag-netic susceptibility variation and chemicalshift effects, reduces sensitivity to biologicalsignal differences and reduces boundary defi-nition. The truncation artefact, generated atboundaries by image processing, is particu-larly relevant at the CSF-spinal cord inter-face, and is one possible cause of the band ofhigh or low signal seen in the centre of thecord in midsagittal images. It also causes dif-ficulty in defining the position of the cord-CSF boundary where the problem iscompounded by susceptibility effects.Phantom studies have shown that both elec-tronic and caliper measurements of the spinalcord, especially in the phase encoding direc-tion, can be artificially reduced by over 2 mmand may create a spurious impression ofspinal cord flattening.'°

Motion artefactsMotion artefacts are generated by cardio-synchronous and oscillatory motion of CSF.This motion has been documented androughly quantified by MRI. At C2/3, move-ment is estimated at about 065 ml percardiac cycle, downwards on systole andupwards on diastole." New data confirmolder work indicating that the primary drivingforce behind intracranial and spinal canalCSF flow is expansion of the brain duringvascular systole. The spinal cord and brainstem also descend very slightly on systole andoscillate anteroposteriorly with CSF flow.'2 Inclassic studies, Rubin et all' showed how theoscillatory motions generate linear artefactsparallel to the cord-CSF interface at roughlyharmonic intervals across the images in thephase encoding direction, producing signalvariation in the spinal cord image that couldbe easily misinterpreted as intramedullarypathology.'3 Areas of turbulent CSF flow in

404





regions where subarachnoid septa exist, par-ticularly in the thoracic spine, can result insignal variations simulating intradural massesor enlarged vessels. Many strategies havebeen developed to minimise these problems,but results can be less consistent than isdesirable.The thoracic cord in particular is difficult

to image well, especially in cross section,where cardiac motion and the proximity ofthe aorta add to the motion generated arte-facts.

Metallic artefactsMetallic artefacts can be particularly destruc-tive of image quality. Many spinal operationsutilise metal stabilisation devices, such asplates, rods, screws, and loops, and patientsoften require postoperative imaging at somestage. Ferromagnetic substances such asstainless steel generate major local artefactsand usually render spinal imaging useless.Tiny fragments from drills and punches,invisible on plain radiographs, may also resultin devastating artefacts."4 1' Design ofimplants also can be important such as theavoidance of conductive loops.'6 The use oftitanium for manufacture may have advan-tages.'7 A wide range of titanium devices hasbecome available only recently, however, andthe long term stability and biological effectsof these new materials remain under evalua-tion. 18

Special techniquesNew methods of brain imaging are usuallyeventually applied to the spine. Many arebeing evaluated and only some will be givenpassing reference here.

Phase contrast imaging utilises bipolar phaseencoding gradients; the first brings all spinsinto phase, the second records reduced signalfrom moving spins as they dephase relative tostationary spins, due to their motion. Thiscan be used to demonstrate molecular diffu-sion. One method displays apparent diffusioncoefficients across the image and fluid filledcavities appear much brighter than solidareas.'9 Phase contrast imaging can also beused to demonstrate coherent CSF flow andmovement of the neural axis. Phase contrastcine MR produces images that representvelocity as a function of time throughout thecardiac cycle, and by the direction of thephase shift indicate flow direction.20 Spatialmodulation of magnetism (SPAMM), alsoreferred to as presaturation bolus tracking, is amethod whereby regions are tagged by apply-ing a narrow band of saturation before thepulse sequence. This produces a dark stripeacross the image that bends in the directionof movement. Multiple presaturation bandscan be used to produce a "zebra stripe" pat-tern, and coupling with a cine loop increasessensitivity in detecting minimal movement.2'This has been used extensively to studysyringomyelia. Susceptibility contrast weighteddynamic MRI has been used to study bloodflow in spinal tumours and arteriovenousmalformations.22 A susceptibility weighted

gradient recalled echo sequence is used todetect the large, transient signal reductionthat occurs on the first pass of a bolus dose ofintravenous gadolinium through the lesion.Fluid attenuated inversion recovery (FLAIR)sequences are said to offer improved lesiondetectability by permitting heavily T2weighted acquisition to be accomplished withsuppression of all signal from CSF. Thisremoves motion artefact from CSF, whichappears paradoxically as signal void on heav-ily T2 weighted images. Cysts and cysticlesions within the cord also appear, however,as signal void. Most other pathology appearsas hyperintensity. At present, imaging withFLAIR is slow and of low resolution,24 andlike most of the functional imaging methodslisted, it has found little general clinical appli-cation at present.

SPINAL SONOGRAPHYIntraoperative spinal sonography has been usedwidely in some centres. A minimum of a twoor three level laminectomy wound is required,filled with water or saline to act as an acousticwindow.2526 Although considerable utility hasbeen achieved,25 experience has shown that itmay not distinguish tumour from cord tissue,may incorrectly identify cysts as solid massesdue to unusual echogenicity of some cyst flu-ids, and cannot reliably distinguish tumoralfrom non-tumoral cysts.26 Surprisingly largetransverse excursion of the spinal cord androots is observed normally, lagging slightlybehind the cardiac cycle, and breathing andthe Valsalva manoeuvre produce additionalabrupt movement. Percutaneous spinal sonog-raphy is possible in infants.27 28 The spinalcanal can also be imaged in adults by anglingthe probe parallel to an intervertebral discspace. Transoesophageal transducers have beenused with some success to image the thoraciccord, but clinical utility is virtually non-exis-tent with the availability of MRI. Transuterinesonography has been used to identify dys-raphic states in the fetus,29 but MRI may alsobe employed for this purpose.

PLAIN RADIOGRAPHS OF THE SPINEPlain films have little part to play in the inves-tigation of spinal cord disease. Inferences canbe made about possible sites of spinal cordcompression in conditions such as spondy-lotic myelopathy and trauma, and may sug-gest the site and type of cord involvement indysraphic states. The role of plain films in thepreliminary investigation of patients withintraspinal lesions was reviewed by Naidich etal as late as 1986.30 The spinal canal is usuallyenlarged in children with intramedullarytumours. Greatest sensitivity is achieved bygraphing the interpediculate distances andcomparing them with normal values, but eventhen the false positive rate is at least 11%.Therefore, plain radiographs have only a verylimited screening value even in children.

SPINAL ANGIOGRAPHYThe technique of spinal angiography wasdeveloped over 30 years ago by Djindjian and

405




Stevens

Doppman. Today it is indicated only forlocalisation of the major radiculomedullaryarteries before operations on the spine, tostudy vascular malformations associated witharteriovenous shunts, and before endovascu-lar treatment of a variety of spinal lesions. Todemonstrate all the spinal cord arteries, selec-tive injections into the intercostal, lumbar,lumbosacral, vertebral, deep, and ascendingcervical arteries is required. Searches fordural fistulae may also require selective injec-tions into branches of the external carotidartery. General anaesthesia is desirablealthough not essential. Moderate, usuallytransient, neurological deterioration occursnot uncommonly after extensive spinalangiography,3' but paraplegia is rare.32 Spinalangiography can be a very laborious under-taking; short cuts have been devised usingother imaging modalities and these will bediscussed with vascular malformations.

Shape, internal structure, andbiomechanics of the spinal cordSIZE AND SHAPESpondylotic flattening of the formalin fixedspinal cord is commonly found at necropsy,and was often considered to be a fixationartefact. Modern imaging has made it abun-dantly clear that this is not so. Good qualita-tive agreement is found between theappearance of the spinal cord on cross sec-tional imaging in living subjects and formalinfixed cords at necropsy,3334 but clinical mea-surement is problematic. The most robustmeasure of size is cross sectional area. Areameasurements could theoretically be accuratewithin about ± 5%,35 but on clinical images itis clear that nothing like this is actuallyachieved. Electronic window settings pro-foundly influence all measurements of thecord on computed myelography36 and onMRI many other factors are also involved.This is shown by the mean values derived byvarious workers for their control populations.To give three examples: mean normal cordcross sectional areas at C2 (where spondylosisis not expected) of 62 mm2,37 86-6 mm2,35 and1 10 mm2 38 have been reported. Close correla-tions between area measurements made bycomputed myelography and MRI werereported by Fukushima et al (r = 0.901)34; butthe mean values were very different-namely,0-38 (SD 0-14) cm2 on computed myelogra-phy and 0 50 (0-16) cm2 on MRI. Yu and hiscolleagues concluded that each departmentneeded to define its own normal range.39 It ishardly surprising that no workers have showna stable relation between cord size, patientage, or body size. Few measurements havebeen published for the thoracic cord, but wecan console ourselves that medical science isnot too much the poorer for that.Two simple measurements of cord shape

have been used: circularity (417 area/circum-ference2),35 and the compression ratio (ratioof the anteroposterior to transverse diame-ters).35 4041 Abnormal shapes have also beenclassified qualitatively in specific types of cord

compression, especially cervical spondylosisand its variants, the most comprehensivebeing by Yu et al.39

INTERNAL STRUCTUREThe internal structure of the spinal cord onMRI seems remarkably similar to an anatomi-cal preparation stained for myelin. Moremyelinated regions generally yield a lower sig-nal than less myelinated ones. Magnetic reso-nance imaging has consistently shownvariations in texture in histologically uniformregions, such as the anterior horns in thesacral area, Clarke's column, and the dorsalhorn complex; the gracile fascicles yield aslightly higher signal than the cuneate.42 43 Insome types of image, the subpial zone hasyielded a high signal44 and it is uncertain ifthis represents an MR artefact, or the narrowband of subpial degeneration commonly seenin the cords of aging subjects. It is notablethat some authors have forgotten the Tishortening that occurs with formalin fixation,making it difficult to obtain images with Tiweighted contrast.

BIOMECHANICAL PROPERTIESDeformation of the cord by transverse com-pression of up to about 20% requires minimalforce, whereas deformation in excess of 50%requires forces exceeding capillary perfusionpressure and begins to disrupt both transverseand longitudinal axons.4546 Moreover, spinalcord substance has only a limited capacity forelastic recoil. It has been shown by measure-ments of cord deformation between flexionand extension on computed myelograms thatonly about 20% of a deformation recoverselastically.47 This has an important implica-tion for interpreting CT and MRI images thatare generally performed in positions wherethe available space for the spinal cord is max-imised. Because of its lack of elasticity, it issafe to conclude that if the spinal cord is nor-mal in cross sectional shape, it is not beingappreciably compressed in any situationoccurring in the patient during normal dailyactivities; and when deformity is present, it issafe to conclude that the deformity shownreasonably represents the magnitude of inter-mittent compression. We developed the sim-ple concept of congruous cord deformity, tohelp distinguish deformation due to intermit-tent compression from that due to atrophy.When the available subarachnoid spaceappears capacious on cross sectional images,compression is suggested, nevertheless, if thedeformity of the spinal cord is reciprocallycongruous with the visible (disc/osteophyte)or expected (reducible subluxation) deformityof the spinal canal; in cord atrophy the defor-mation is incongruous.48The compressed spinal cord usually

increases in size after operative decompres-sion. Fukushima et al34 measured a meanpostoperative increase in size of the com-pressed spinal cord of 13%, varying between5% and 20% in different subgroups ofpatients, which is of the order one wouldexpect from cord elasticity studies. The cord

406group.bmj.com on April 8, 2018 - Published by http://jnnp.bmj.com/Downloaded from




also often changes shape after decompression,becoming less flattened or altered in someother way, because it is very easily deformedby varying conditions.

CORD MOTIONThe cord lies about 2 mm more posteriorly inthe supine than in the prone position, and thecord and vertebral midlines differ by up to 2mm in over 40% of non-scoliotic subjects.49On phase contrast MRI the range of nor-

mal cardiosynchronous oscillatory longitudi-nal movement of the upper cord has beenmeasured to be only 0 4-05 mm.'2 Initialstudies caused excitement because they sug-gested that assessment of tethering could bemade without objective influence of morpho-logical appearances. Reduced oscillatorymotion has been seen in the presence of cordtethering, and increased oscillatory motion inChiari associated syringomyelia.21 The actualrelevance of any of these findings, however, inthe absence of morphological changes, suchas a low lying conus medullaris, has, in myopinion, yet to be demonstrated.

Pathological states in cord substanceNECROSISWater soluble contrast medium in the sub-arachnoid space diffuses freely into neural tis-sue, and x ray attenuation of spinal cordsubstance measured on CT equilibratesdynamically at about 20% of that in the sub-arachnoid space, being removed by the capil-lary-venous system.50 When necrosis occurs,hydrophobic lipid is broken down and thecapillary bed destroyed. Contrast mediumcontinues to accumulate in the necroticregion until the CSF concentration falls, andpassive diffusion out of the area takes severalhours. Necrotic areas thus appear as circum-scribed areas of contrast enhancement oncomputed myelograms, often most conspicu-ous after six to 12 hours.5'On MRI, colliquative necrosis appears as a

circumscribed area of signal change.Necrosis, from whatever cause, most ofteninvolves mainly the central parts of the spinalcord, in particular the grey matter and ventralparts of the posterior columns. Involvement isusually bilateral, producing either a localisedconfluent lesion or, more often, bilaterallesions resulting in an appearance likened tosnakes' eyes on cross sectional images.52

WALLERIAN DEGENERATIONAntegrade degeneration in the long tracts ofthe spinal cord is seen with pathological statesthat cause axonal damage. Typically itappears as descending degeneration in theanterior part of the lateral column andascending degeneration in the posteriorcolumns, which, in diseases of the cervicalspine, is often most severe in the fasciculuscuneatus."3 4144 53

Wallerian degeneration is shown only byMRI. Its appearances have been studiedmostly in the brain stem, and only recently inthe spinal cord. Four stages are distinguish-

able on imaging, which evolve over about 14weeks.54 For up to four weeks, MRI is normaland the chronic stage, characterised by vol-ume loss and increased MR signal, persistsindefinitely. Enhancement with intravenouscontrast agents does not occur at any stage.In the spinal cord only the chronic stage hasbeen described,4455 but recent experimentalstudies with magnetisation transfer imaginghave demonstrated abnormalities in the earli-est stage.56

CYSTIC DEGENERATION AND SYRINGOMYELIASyringomyelia may result from any pathologi-cal process that is liable to cause spinal cordnecrosis. It represents an end stage that itselfmay be progressive and promote further corddamage. The cavities are usually located dor-sal to the central canal, with which they mayor may not communicate. They may be singleor multiple. Only about 15% extend beyondC2, and those that do so usually bifurcatearound the decussations in the medullaoblongata and come to lie ventral to the floorof the fourth ventricle. Hydromyelia is a cavityconsisting mainly of a dilated central canaland communicating with the fourth ventricle.It is closely associated with hydrocephalus,and is collapsed by ventricular shunting.57

Chiari I associated syringomyeliaIn the currently most widely accepted hydro-mechanical theory of causation, this begins asa hydromyelia that subsequently losescommunication with the fourth ventricle earlyin life, resulting in the condition oftenbeing referred to as syringohydromyelia.Obstruction of the foramen magnum or theoutlets of the fourth ventricle by thedescended cerebellar tonsils is central to mosthydromechanical theories. In support, phasecontrast cine MRI has demonstrated absenceof CSF in the cisterna magna due to theabnormal cerebellar tonsils, and restoration ofCSF flow after foramen magnum decompres-sion, accompanied by collapse of the syrinx.Also, two groups have recently separatelyreported a new finding on dynamic MRI:accentuated caudal displacement of the cere-bellar tonsils and spinal cord with cardiac sys-tole, which was restored to the normal range,or obliterated altogether, by decompression ofthe foramen magnum.2' 58 Oldfield et al pro-posed intermittent piston like obstruction ofthe foramen magnum by the tonsils.58

Although these mechanisms may operatein cases where the cerebellar tonsils doobstruct the foramen magnum, other workershave shown that the foramen magnum is notobstructed by the abnormal cerebellar tonsilsin at least 20% of cases, and the cerebellumwould need to move more than 10 times asfar as has been recorded to cause intermittentobstruction.59 60 Furthermore, no associationexists between the distension or cranial extentof the syrinx and the presence or absence offoramen magnum obstruction, or degree oftonsillar descent59; indeed syringomyeliaoccurs significantly more often with mildrather than with severe tonsillar descent.606'

407




408

Additional complications have arisen withrespect to the nature of the Chiari I malfor-mation itself. Firstly, assessment of the levelof the cerebellar tonsils on midline sagittalMR images has greatly overestimated theprevalence of tonsillar descent below the fora-men magnum. Indeed, a prevalence ofbetween 15% and 20%61 is still generallyaccepted despite the previous assessments onmyelography and computed cisternographyindicating a prevalence of less than 1%. Therecent volumetric MRI study of Savy et al,however, has confirmed the second figure tobe correct.62 Although the apparent preva-lence of tonsillar ectopia on sagittal MRI inthe study was 20%, it was explained by par-tial volume averaging and was therefore spu-rious. Secondly, in over 50% of cases withtrue cerebellar ectopia, the medulla oblongatais also elongated. Indeed, a linear relation hasbeen shown between the presence ofmedullary elongation and severity of thedescent of the cerebellar tonsils,63 whichresults in the obex often lying in the cervicalcanal, not the cranial cavity, and actuallylying below the tonsils in about half of suchcases.59 60 63 These anatomical facts are usuallyignored in hydromechanical explanations ofsyringomyelia and pose appreciable difficul-ties. Finally, it now seems certain that theChiari I lesion is an acquired deformation ofthe rhombencephalon, and not a congenitalmalformation at all. Serial MRI examinationshave clearly shown the development of typicalChiari I deformities postnatally,6465 the causeapparently being a lower rate of growth of thebasicranium relative to the cerebellum in thefirst two years of life.

Dynamics and clinical aspectsOn air myelography, an important diagnosticobservation was whether an enlarged cervicalcord collapsed in the head up position.Syringomyelia was collapsing, cord tumoursor non-fluctuant cysts were not. Similar find-ings could be made on water soluble myelog-raphy, but the change in cord size was smallerand in the opposite direction, opacified CSFbeing denser than cyst fluid.5' Serial MRI has,however, sometimes revealed large fluctua-tions in cord size with no intervention what-soever, and not associated with any change inclinical status.59 60 66

There are numerous surgical strategies forcollapsing a distended syrinx,67 the common-est being foramen magnum decompressionand syringoperitoneal shunting, either ofwhich will collapse 70%-80%. The hypothe-sis of Williams holds that CSF enters the cordfrom above, due to intermittent pathologicalraising of intracranial over intraspinal pres-sure.68 This has been challenged by new intra-operative measurements indicating higherintraspinal pressures,69 lending support toalternative hypotheses, which propose thatraised intraspinal CSF pressure forces CSFinto the cord via the dorsal root entry zone,or Virchow-Robin spaces. Park et a169 treateda small series of patients with Chiari associ-ated syringomyelia by lumboperitoneal shunt-

ing, and found that this also collapsed thesyrinx in about 80% of cases. This approachhas been used in different types ofsyringomyelia, with similar success70- 72; theobvious advantage over syringoperitonealshunting is that it does not interfere with thespinal cord, and it is much less painful andhazardous than foramen magnum decom-pression.No correlation can be shown between the

degree of distension, or indeed the extent, ofa syrinx and severity of clinical features.51 73 75

Vaquero et al studied 30 patients clinicallyand with MRI both preoperatively and post-operatively and showed that collapse of thesyrinx was achieved in 29, but only 46%improved clinically, and 27% continued todeteriorate despite MRI showing persistentcollapse of the syrinx.75 Furthermore,Sherman et al used serial MRI to show thatChiari associated and post-traumaticsyringomyelia successfully collapsed by surgi-cal intervention may still continue to propa-gate through the cord.73 Assessment of CSFdynamics by MRI has been just as disap-pointing in predicting clinical progression oroutcome,73 despite several claims to the con-trary based on isolated cases and no follow updata.

Spinal cord compression and injuryIn 25% of their control population Yu et alfound the spinal cord to be at least moder-ately compressed by osteophytes at C5/6 or6l7.35Some workers have found an apparently

linear correlation between numerical evalua-tion of cord compression (cord cross sectionalarea or compression ratio) and clinical dis-ability34404'53 whereas others have not.377677All studies indicate, however, that when cordcompression is sufficient to reduce cord crosssectional area by more than 60%, clinical dys-function is usually present. Canine modelshave suggested that lower limb paralysisappears only when gradually applied cordcompression exceeds 50%, and that the rela-tion between compression and clinical dys-function is non-linear, following a catastrophemodel rather than a linear one.47 More isinvolved, however, than mere static compres-sion.

In a canine model of chronic spinal cordcompression at C5, Al-Sefty et al showed thatprogressive paraplegia developed in most ani-mals at a mean of seven months after only30% compression of the cord.78 In a series ofpatients with chronic malunited fracture ofthe dens, Crockard et al showed a log linearrelation between reduction in cord cross sec-tional area, as measured on computed myelo-grams, and time in years since injury.79 Inanother canine model, Anderson showed thata rapidly applied deformation of the cord ofonly 20%-30% produced much greater dam-age than slow compression of up to 50%-60%, and that this consisted mainly ofhaemorrhagic damage to the grey matter inthe first, and mainly white matter change in

Stevens





the second.80 In chronic compression,although blood flow reduction is maximal inthe anterior columns in contact with the com-pressive agent, paradoxically it is the lateraland the anterior parts of the posteriorcolumns that show pathological change.47 Inthe brain, diffuse axonal injury is the result ofshearing forces generated by rotatory acceler-ation. Similar forces are generated in thespinal cord, where the pia mater is restrainedmore than the rest of the cord structure bythe dentate ligaments and spinal roots, gener-ating shearing forces maximal in the lateraland dorsal columns.33 In the brain, verysevere rotatory acceleration causes diffusevascular injury, which is characterised byhaemorrhagic damage in the basal ganglia,similar to the grey matter damage seen in thespinal cord. Because the white matterchanges generally appear more suggestive ofvascular insufficiency than diffuse axonalinjury, however, most workers currentlybelieve that progressive cord damage is due torepeated episodes of momentary arrest of themicrocirculation. The resulting changes aremaximal in the vascular watershed area andtend to lead to cavitation, especially in theventral parts of the posterior columns; this isreadily shown by MRI.78

ImagingOnly MRI consistently shows the changes inthe spinal cord that result from compression.They are best shown on T2 weighted images.The distribution is usually characteristic, con-sisting of diffuse signal change at the site ofmaximal compression, with variable extensionto the central part of the cord, often bilateral,and resulting in an appearance reminiscent ofsnakes' eyes.4452788' These changes are shownon Ti weighted images only when the dam-age is particularly severe, and consist of lowsignal; when present they are a sign of poorprognosis. This is not so for signal changes onT2 weighted images, which often disappearcompletely after operative decompression, butpersist when operative outcome is poor.'282The pathological substrate for reversible MRIchanges is not known, but it is often assumedto be oedema. The size of the spinal cord atthe site of compression is also of prognosticimportance. Several studies have shown thatwhen the cord is reduced in size by more thanabout 50%-60%, operative outcome ispoor.'43776 This applies only to compressionin cervical spondylosis and subluxation. Thecord tolerates far greater compression frombenign tumours such as meningiomas andschwannomas and functional recoveryremains likely after decompression even whenthe cord is severely compressed.

In a recent computed myelographic studyof 56 patients with spondylotic myelopathywho had a poor operative outcome,83 an alter-native cause for the myelopathy (usually mul-tiple sclerosis) was established in only 14-3%.The spinal cord was reduced in size by 60%or more at the site of previous compression inonly 26-8%, and only 15-6% had evidence ofcord necrosis. In 57-1%, operation failed to

decompress the spinal canal. Another studyby the same workers has put paid to the ideathat osteophytes usually regress or disappearafter interbody fusion: some osteophytes wereas large as they had been up to eight yearsafter the preoperative computed myelogram,and in no case did any measurable regressionoccur.84

Clinicoradiological approach to cord compressionin spondylosis and subluxations in the cervicalspineIn patients with suspected compressivemyelopathy, osteophytes, disc protrusions, orsubluxation are irrelevant when the spinalcord is normal or only mildly flattened. Corddeformation of up to about 40% is also mostlikely to be irrelevant, unless appropriate sig-nal change is present in the cord on MRI atthe site of compression. Usually, however,signal change is present only in severely com-pressed cords.

Cord deformation of over 50% is likely tobe relevant, but this is also the point at whichclinical recovery from decompression isbecoming less likely. Therefore, it seemsappropriate to consider surgery in asympto-matic or mildly affected children and youngadults as a prophylactic measure when cordcompression is approaching 50%. In mostpatients, however, with cervical spondylosis,such operations will be less appropriatebecause of age and low expectation of deteri-oration within the relevant time frame.47

Acute spinal cord injuryIt is now established in animal models thatthe extent of signal change shown in thespinal cord on MRI is related to the severityof injury,8586 and clinical studies have alsoshown a general association between theextent of signal change on MRI and func-tional outcome.8789 Mild or transient loss offunction after spinal injury is not usuallyaccompanied by signal change in the cord onMRI.8789 In more severe injury, evidence ofhaematomyelia is present on MRI in onlyabout 50%. Cord swelling is mild and notalways present even within seven days ofinjury, and ongoing cord compression is usu-ally absent.87

Progression from an acute injury tolocalised cystic myelopathy has been followedby serial imaging, most cysts being asympto-matic.82 There is good evidence that cystsresult from colliquative necrosis, and exten-sion of spindle shaped cores of ectopicnecrotic tissue along the bases of the dorsalhorns into adjacent uninjured parts of thespinal cord is visible on MRI.84-86 89 Thesenecrotic cores are likely to be the basis of theelongated cavities that may occur within a fewweeks of injury,82 some of which distendand propagate and become associated witha progressive ascending myelopathy.5' 82Progressive post-traumatic myelopathy canalso occur in the absence of cavitation,5' 90and is associated with ascending centralnecrosis in the spinal cord, manifesting assignal changes on T2 and T1 weighted MR

409




Stevens

images" and abnormal accumulation ofcontrast medium on computed myelo-graphy.5' Other abnormalities such as adhe-sions, and occasionally cord compression, arefound in some cases; some workers considerthese to be the cause of progressive cord damage.

Spinal cord injuries in children differ insome ways from those in adults.9' Childrenmay have extensive cord contusion or infarc-tion with minor, remote, or no spinal frac-ture. Any signal change found on MRI isusually followed by considerable persistentfunctional loss.

Vertebral artery injury has recently beenreported as occurring in nearly 46% of casesof midcervical fracture dislocation.92 It isnotable, however, that spinal cord infarctiondue to vascular injury from subluxation,trauma, or cervical spondylosis and relatedconditions is exceptionally rare, and its docu-mentation is confined to only two or threecase reports over the past 30 years. Extensiveand severe adhesive arachnoiditis and superfi-cial siderosis are also described as rare, latecomplications.93

Vascular lesions of the spinal cordANATOMY AND PHYSIOLOGYThe blood supply of the spinal cord hasrecently been reviewed in detail by Lasjauniasand Berenstein.94 The anterior spinal arterysupplies a centrifugal arterial system andradial arteries from the vascular network onthe surface of the cord form a centripetal sys-tem. A watershed zone between these systemshas been defined consisting of the inner 25%of the white matter and the outer edge of thegrey matter, excluding the posterior 50% ormore of the posterior horns. Regional bloodflow in cord white matter of primates hasbeen estimated to be as low as 10 ml/100g/min and in grey matter as 58 ml/100 g/min,which is only about half cerebral blood flowmeasured by similar techniques. The bases ofthe dorsal horns seem the most vulnerableregions within the cord to ischaemia orhypoxaemia.Two main groups of veins drain the spinal

cord. The central veins, collecting from bothhalves and central parts of the cord, and theradial veins from capillary plexuses at theperiphery of both grey and white matter. Acoronal plexus of veins on the surface of thespinal cord forms a longitudinal networkwhich drains out of the spinal canal along themedullary veins that accompany the spinalroots at varying intervals. These veins are nar-rowed as they traverse the dura mater, thenarrowings perhaps functioning as weakantireflux valves. Although gravity favoursinferior venous drainage, in the cervicalregion cranial venous anastomoses seem ofparticular importance. High cervical obstruc-tion has been shown to cause venous conges-tion and stagnant hypoxia in the central partsof the spinal cord in the cervical enlargement.

SPINAL CORD INFARCTION

The MRI appearances have been described in

many cases.95 9 The commonest change hasbeen diffuse signal increase on T2 weightedimages, most often involving the lower tho-racic region. Cord swelling has been mild orabsent even in the acute phase. Centralhaemorrhage has been noted.94 In some casesonly the ventral part of the cord has beeninvolved, either limited in extent, confined togrey matter, or more diffusely in both greyand white matter.95 Diffuse contrast enhance-ment may be seen after intravenous gadolin-ium in patients examined 10-21 days afteronset, but not earlier or later.96 An associationwith infarction in adjacent vertebral bodieshas been noted.96

Venous infarction of the cord has beenreported less often, and in two recent casesthe MRI abnormality consisted of unilateralsignal change.4497 One was confirmed bynecropsy to be thrombosis of the posterolat-eral pial vein complex,44 and the other wasspeculated to be due to thrombosis in theseveins induced by a YAG laser during removalof an intradural neurinoma at C2. Serial MRIin the second case showed diffuse cordswelling and signal change from Cl to C3,which reduced within one month to a circum-scribed area of signal change involving nearlyall the lateral half of the spinal cord, confinedto the site of surgery.

SPINAL VASCULAR MALFORMATIONSDural arteriovenous fistulaeSpinal vascular malformations used to beclassified according to the extent of theabnormal intradural vessels, which was adescriptive rather than a functional approach.In the early years of spinal angiography, sur-geons believed that most of the vessels werearteries, as did many reputable neuropatholo-gists until very recently. The careful observa-tions of Kendall and Logue revolutionisedthinking about these lesions, however, andnow form the basis of the modern functionalclassification. Most spinal arteriovenous mal-formations are dural arteriovenous fistulae,the enlarged intradural vessels being veins notarteries. The fistula is located in the duramater close to the nerve roots, usually in thethoracic region in older patients,98 99 but canoccur in the lumbosacral theca or in the duramater around the foramen magnum or theposterior cranial fossa.'°" Slow, aberrantvenous drainage is an important feature, andis presumed to be due to thrombosis of radic-ular veins. Their precise anatomy requiresspinal angiography for elucidation. Treatmentis often straightforward, by operative orendovascular occlusion of the fistula.Haemodynamic improvement does notalways occur, however, because the throm-botic aspect of the disease may remain, withimpaired venous drainage of the spinal cord.99Complications of dural fistulae includeintramedullary haemorrhage, cord atrophy,and cavitation in the cord, usually above thefistula."9

High resolution MRI should detect mostclinically relevant arterialised veins, butoverdiagnosis is possible. Conventional

410





myelography probably remains the most sen-sitive and specific technique for their detec-tion; however, enlarged or conspicuousintradural veins that drain normally, even iffilled by a fistula, are not usually associatedwith clinical myelopathy.99 Virtually allpatients with clinical myelopathy have signalchanges in the lower part of the spinal cordon MRI, usually surrounded by a small rim ofapparently unaltered cord tissue. The signalchange often disappears partially or com-pletely when the fistula is successfully closedand symptoms remit, and may reappear if thefistula reopens. There may be patchyenhancement of spinal cord substance afterintravenous gadolinium.98100The site of the fistula can be detected reli-

ably by susceptibility contrast weighteddynamic MRI when the intradural veins arelarge enough to be shown easily. A series ofheavily susceptibility weighted fast images isacquired in the midsagittal plane after anintravenous bolus of gadolinium, and thepoint at which signal nulling first appears inthe intradural veins indicates the level of thefistula.23 Spinal angiography can then bedirected at this level, greatly speeding up thediagnostic and therapeutic process.

Intramedullary arteriovenous malformationsIntramedullary arteriovenous malformationsmay be either a nidus, or a direct arteriove-nous fistula, located within cord substance oron the pia mater. Fistulae are more commonin children. The nidus is often visible on MRIas a focal expansion of the cord closely asso-ciated with serpiginous signal voids indicatingthe draining veins and sometimes enlargedarteries. Successful endovascular treatmentmay be feasible, even for intramedullarylesions, but multiple sessions may berequired, obliteration is often incomplete, andthe recurrence rate is high.'0'

Cavernomas and capillary angiomasCavernomas and capillary angiomas are lesscommon in the cord than in the brain. OnMRI, they appear usually as localisedexpansions of the spinal cord, with sharplycircumscribed signal change98 102 that cannotbe distinguished from small intramedullaryhaemorrhages. Intravenous gadoliniumenhancement may demonstrate otherwiseinvisible lesions in rare cases of multiple cap-illary haemangiomas.'0'

Neoplastic and inflammatoryintramedullary processesNEOPLASTIC PROCESSESAstrocytomas and ependymomas occur aboutequally in the spinal cord itself, butependymomas are much more common inthe filum terminale. Extramedullary ependy-momas occur occasionally in the extraduralpart of the filum, involving the sacrum.50Glioblastomas are rare in all ages, as are oligo-dendrogliomas. Isolated reports are to be foundof subependymomas of the cervical cord andgangliogliomas of cervicothoracic cord and

filum terminale. Metastases are not uncom-mon. Primary lymphoma affecting only thespinal cord has now been reported severaltimes. Reports of exceptionally rare neo-plasms arising within or directly involving thecord have included melanoma; intramedullaryneurofibromas arising from Schwann cells innerves encasing blood vessels of the cord;intramedullary teratoma, associated with pre-cocious puberty; primitive neuroectodermaltumour; mesenchymal chondrosarcoma withoutdural attachment, and paraganglioma of thefilum terminale.

Finally, there is the relatively commonspinal capillary haemangioblastoma, and itswell known association with Von Hippel-Lindau disease. Screening of all family mem-bers with abdominal CT and spinal MRI withgadolinium enhancement has been recom-mended by several workers because 40% ofaffected patients may be asymptomatic at thetime of screening.On clinical imaging, the hallmark of

intramedullary neoplasms is expansion of thespinal cord, usually greater than in inflamma-tory conditions. Lobulation, or eccentricenlargement is extremely suggestive. Bothastrocytomas and especially ependymomascan appear as very well circumscribed signalchange on MRI. Sometimes circumscribedlesions appear etched out by a salient low sig-nal pseudocapsule, around the entire circum-ference or capping the cranial and caudalextremities, consisting of dense gliosis orhaemosiderin staining; this is more frequentwith ependymomas.50 Enhancement afterintravenous gadolinium is usual, unlikebenign intracranial gliomas, which usually donot enhance. Enhancement is patchy, anddoes not reliably indicate all neoplastic areas.

Haemangioblastomas and metastases usu-ally have a different appearance. They arewell defined, and enhance strongly after intra-venous gadolinium. About 50% of haeman-gioblastomas are associated with enlargedintrathecal veins, visible on all types of imag-ing including MRI. Spinal cord oedema iscommon with metastases and shows up wellin white matter with MRI44; it may be diffi-cult to distinguish from cavitation.

Three types of cyst occur in associationwith intramedullary neoplasms, and about70% will have at least one type: intratumoralcysts, the walls containing or consisting ofneoplastic tissue; capping cysts, cone shapedcavities extending for one or two spinal seg-ments into uninfiltrated cord cranial and caudalto the tumour; and syringomyelia, undistin-guishable from other causes remote from thetumour.50 Even on MRI and intraoperativesonography, it can be difficult to distinguishsome cystic from solid or necrotic tissue.

INFLAhMMTORY PROCESSESUntil the era of MRI, imaging was usuallynegative in these conditions, but virtuallyall inflammatory processes produce changesin cord substance detectable by MRI.Unfortunately, they all look alike and mosthave been confused with neoplasia. Diagnosis

41




Stevens

usually depends on clinical evolution, labora-tory tests, or even cord biopsy, and in manycases the diagnosis remains uncertain.

Multiple sclerosisThe primary demyelinations, which includeacute disseminated encephalomyelitis, pre-sent a spectrum of stage dependent changes,50and the stages follow a roughly predictabletime course that can be helpful in establishinga diagnosis.Stage 1: perivenous inflammation and oedema-On MRI the cord may show mild fusiformenlargement if the lesion is large enough, withpoorly defined signal change throughout theinvolved area, either diffuse or sparing thecord periphery. Clinical dysfunction is at itspeak during this phase. Patchy or diffuseenhancement occurs after intravenousgadolinium within the area of signal change,but not coextensive with it. Similarity withcord glioma is particularly close at this stage,which lasts two to eight weeks.Stage 2: demyelination and glial proliferationOnce MRI cord swelling has subsided, asmaller, more circumscribed area of signalchange is evident, and enhancement nolonger occurs after intravenous gadolinium.This is how most multiple sclerosis usuallypresents on MR images. Visualisation is con-siderably improved by the heavy T2 weight-ing provided by spinal FLAIR MRI.'04Lesions tend to involve sectors of the cordwhite matter extending to the periphery of thecord and are best shown on cross sectionalT2 weighted images. The posterior columnsand the posterior parts of the lateral columnsare most commonly involved. The lesions areusually unilateral, or extend across the mid-line; they do not have the appearance ofsnakes' eyes. These features should distin-guish plaques of demyelination arising in thecervical cord near sites of spondylotic com-pression, from cord damage due to the com-pression alone.Stage 3: atrophy-The spinal cord is small orfocally or diffusely flattened. In rare cases, itbecomes cavitated.

Nearly all multiple sclerosis lesions eventu-ally progress to stage 3. Acute disseminatedencephalomyelitis lesions usually arrest beforethis stage, often not progressing beyond stage1, and most regress completely. The prognos-tic significance of brain lesions, which arefound at presentation in about 60% ofpatients with clinically isolated cord syn-dromes, has been reviewed recently.'05

SarcoidosisInvolvement of the spinal cord is much lessfrequent than that of the brain or peripheralnerves in established cases. The appearanceson myelography and MRI can be dramatic,although only a few cases have been fullydescribed.'06 107 The cord may show pro-nounced and extensive fusiform or irregularexpansion, with variable signal changes on TIand poorly circumscribed high signal on T2weighted images. Patchy, non-uniformenhancement usually occurs after intravenous

gadolinium and may persist for months. Thesolid enhancing areas have been shown toconsist of astrocytic gliosis in which areembedded typical sarcoid granulomas. Oneoperated case was also found to have exten-sive cystic change involving almost the entirecord, the cyst containing xanthochromicfluid. Milder forms, indistinguishable fromfocal multiple sclerosis lesions, have also beendocumented.50 The cord is usually involvedalong with the brain. Lexa and Grossmandescribed cord involvement in three of 24established cases of neurosarcoidosis.106 Theassociation of changes in periventricular orperipheral white matter in the brain, and lep-tomeningeal enhancement after intravenousgadolinium, is particularly suggestive of sar-coidosis. Intravenous gadolinium is definitelyhelpful in identifying meningeal disease andlocating additional lesions, which may clincha difficult diagnosis. Rapid reduction in con-trast enhancement, accompanied by clinicalimprovement, was seen in 90% of patients inresponse to steroid treatment.'06

Spinal tuberculosisSpinal tuberculosis has a range of involve-ment similar to that of sarcoidosis. Meningealfibrosis with chronic cavitatory myelopathy ismore common, especially in countries wheretuberculosis has a high prevalence, such as inIndia, and an MR appearance consisting ofmultiple superficial enhancing lesions afterintravenous gadolinium is probably seenmore often.'08 The diagnosis should be madefrom the CSF.'08 Response to antituberculoustreatment is variable, as in the brain, and maybe preceded by a period of apparent worsen-ing of the appearances.

Intramedullary abscess (pyomyelia)Pyomyelia may occur from haematogenousdissemination, but is exceptionally rare. Moreoften there is an underlying abnormality, suchas a dermal sinus.'09'"° A peripherally enhanc-ing liquefying mass in a swollen oedematouscord is shown by MRI.

Acute varicella myelopathyHerpes zoster can present with neurologicaldisability before the onset of the cutaneousrash, usually consisting of unilateral limbweakness with or without long tract signs."'Magnetic resonance imaging has shown mildenlargement of the spinal cord, with diffusesignal change in the ipsilateral posterolateralportion and coextensive enhancement afterintravenous gadolinium. Three or four seg-ments are involved, a little more extensivethan the dermatome distribution of the cuta-neous rash when it appears. Only partial reso-lution may follow, with residual signal changein the cord and persistence of some dysfunc-tion. The condition can occur in fit patientsas well as those who are immunosuppressed.It is considered to be due to direct involve-ment of the spinal cord by the virus.

Tropical spastic paraparesisTropical spastic paraparesis is a progressivevacuolar leukomyelopathy showing a strong

412





association with human T cell lymphotrophicvirus type 1 HTLV I. The clinical course isrelentlessly progressive. The thoracic region isusually involved. Extensive patchy signalchange has been shown in the dorsolateralpart of the spinal cord, with patchy, some-times superficial, enhancement after intra-venous gadolinium.

Listeria meningoencephalomyelitisListeria monocytogenes produces an encephalo-myelitis characterised by multiple micro-abscesses. Mass lesions can form, simulatingmalignant tumours. Extensive brain stem andspinal cord involvement has been reportedseveral times. A case presenting as an isolatedabscess in the cervical spinal cord wasdescribed recently, showing the MRI featuresof an abscess.

Lyme diseaseLyme disease may cause an acute transversemyelitis with extensive cord involvement,often associated with involvement of theperipheral nerves. Demaerel et al recentlydescribed a case involving just the spinalmeninges, and we have encountered a similarcase."12 In both, only postgadolinium MRIwas abnormal, showing pronounced, diffuseenhancement of the pia mater of the brainstem and entire spinal cord. In our case, MRIof the head a few hours after the spinal exam-ination showed that the gadolinium had dif-fused into the CSF producing a positivecontrast cisternogram.

Granulomatous angiitis of the spinal cordGranulomatous angiitis is a condition charac-terised by granulomata involving vascularwalls, disseminated through the meningesand neural tissue, which only rarely involvesthe spinal cord. A case with extensive signalchange throughout the spinal cord, showingno enhancement after intravenous gadolin-ium, has been reported, and another other-wise similar case, which showed extensivemainly superficial enhancement after gadolin-ium suggestive of metastatic diseaseensheathing the cord.

Congenital abnormalities ofthe spinalcordCongenital abnormalities of the spinal cordhave been extensively reviewed by Naidich etal, to whom the interested reader is referredfor details."3"4 Some represent disorders ofneurulation of the neural plate and disjunc-tion of the neuroectoderm from the ecto-derm; and these include meningomyeloceleand lipomas of the spinal cord. In the first,the un-neurulated neural plate (placode)remains part of the integument. In the sec-ond, a localised region of the neural tube hasfailed to neurulate before disjunctionoccurred and mesenchyme contacting theexposed dorsal surface of the neural plate hasdifferentiated into adipose tissue; disjunctionusually was complete and the overlying duramater is intact, to create an apparently

intramedullary intradural lipoma. Spinalroots emerge from the ventral surface ofthe un-neurulated neural placode in bothconditions.The commonest malformations involve the

more caudal part of the neural tube, most ofwhich forms by canalisation of the caudal cellmass that develops in the tail fold of theembryo. When the tail fold disappears, thispart of the cord normally undergoes retro-gressive differentiation to form the filumterminale. A useful descriptive term for thisgroup of conditions is lipomyelomeningo-dysplasia to emphasise the elements usuallypresent to some degree in all. The spinal cordor thickened filum extends down, usually tothe sacral segments, and blends with a lipomathat extends through a dural defect andneural arch defect of variable size and lengthto blend with subcutaneous fat. The site ofblending with the dura mater and its extent,usually referred to as "tethering", is variable,as is the size and distribution of the lipomaand degree of meningeal ectasia. The spinalcord does not expand into a normal lumbarenlargement and the conus medullaris usuallylies at or below L3, its position sometimesbeing difficult to define.

Another group of conditions seems to bedue to much more focal, even punctate, fail-ures of disjunction. A dorsal dermal sinusextends from skin dimple through or betweenneural arches to the dura, and very occasion-ally intradurally; about 20% of spinal dermoidsand epidermoids are connected to a dorsal der-mal sinus. The neuroectoderm, and ectodermare normally briefly connected in embryoniclife, via the neurenteric canal or adhesion.Persistence of the normal adhesion, or aber-rant adhesions at other levels, can result in aconnection from foregut to spinal canal,along which neurenteric cysts may form.Intradural spinal neurenteric cysts were wellreviewed recently by Brooks et al,"' and pre-sent a reasonably characteristic appearanceon MRI. Persistence of a communicationwith the skin of the back is a dorsal entericfistula.

Diastematomyelia is a relatively commonanomaly often also considered to be due toaberrant neuroentodermal adhesions. Over avariable number of segments the spinal corddevelops as two, usually unequal hemicords,not duplications, although often there are twocentral canals, median sulci, and anteriorspinal arteries. Sometimes this division takesthe form only of a deep cleft, but usually thehemicords are entirely separate. In over 50%of cases, both are enclosed in a commondural tube; the rest are associated with split-ting of the dura mater also, and a bony spurarising from malformed thickened laminaethen often penetrates between the duraltubes. The spinal cord can be affected at anylevel, or rarely the filum terminale or medullaoblongata.

Finally, excessive retrogressive differentia-tion of the tail fold and caudal cell mass canlead to varying degrees of sacral andsacrolumbar agenesis, often referred to as the

413




414

caudal regression syndromes. The spinal cord isabnormally short, the conus lying in the tho-racic region at a variable level.Many of these often dramatic anomalies

are asymptomatic, and remain so throughoutlife. Much recent medical literature is stillconcerned with their diagnosis, which is noweasy, and often seems to exaggerate theimportance of timely surgical intervention.This is especially true in the concept of teth-ering. It has been documented recently thatthe filum terminale is thicker than 2 mm andfilled with fat in 4% of normal patients, andthat the conus medullaris lies at the level ofthe lower part of L2 in about 2% of the nor-mal population. Some workers have indicatedthat the conus may be tethered, but normal inposition, and others that functional MRI orsonography may demonstrate tethering in theabsence of any morphological abnormalities. Iremain aloof from such opinions at the pre-sent time. In cases with progressive disabilitydue to the lesion, structural imaging can beimportant in demonstrating normal cord tis-sue giving rise to spinal nerves and in suggest-ing an operative goal such as debulking of alipoma or the drainage of a cyst. High resolu-tion MRI has replaced the need for myelo-graphy and CT in preoperative assessment ofsuch cases.

1 Shapiro R. Myelography. 3rd ed. Chicago: Year BookMedical Publishers Inc, 1975.

2 Jirout J. Pneumographic examination of the cervicalspine. Acta Radiol 1958;50:221-45.

3 Johnson AJ, Burrows E. Thecal deformity after lumbarmyelography with iophendylate (Myodil) and meglu-mine iothalamate (Conray 280). Br I Radiol1978;51: 196-202.

4 Kendall BE, Stevens JM, Thomas D. Arachnoiditis.Current imaging 199 1;2:113-9.

5 Sortland 0, Skalpe IO. Cervical myelography by lateralcervical and lumbar injection of metrizamide: a com-parison. Acta Radiol 1977;355(suppl):154-63.

6 Skalpe IO, Sortland 0. Myelography. 2nd ed. Oslo:Tano, 1989.

7 Shaw DD, Back-Gansmo T, Dahlstrom K. Iohexol:summary of North American and European clinicaltrials in adult lumbar, thoracic and cervical myelo-graphy, with a new non-issue contrast medium. InvestRadiol 1985;20(suppl) :44-50.

8 Robertson HJ, Smith PD. Cervical myelography. Surveyof modes of practice and major complications.Radiology 1990;174:79-83.

9 National Radiation Protection Board. Protection of thepatient in x-ray computed tomography. Chiltern:NRPB, 1992.

10 Youser DM, Janick PA, Atlas SW, et al. Pseudo atrophyof the cervical portion of the spinal cord on MRimages: a manifestation of the truncation artefact?AJ7NR Am ] Neuroradiol 1990;11:373-7.

11 Enzmann DR, Pelc AJ. Cerebrospinal fluid-flow mea-sured by phase contrast cine MR. A.7NR Am TNeuroradiol 1993;14: 1301-7.

12 Mikulis DJ, Wood ML, Zerdoner OAM, Poncelet BP.Oscillatory motion of the normal cervical spinal cord.Radiology 1994;192:117-21.

13 Rubin JB, Enzmann DR. Dyke Award. Harmonic modu-lation of proton MR precessional phase by pulsatilemotion: origin of spinal CSF flow phenomenon. AJRAm _7Roentgenol 1987;148:983-94.

14 Heindel W, Friedmann G, Bunke J, Thomas B,Firsching R, Ernestus RI. Artefacts in MR imagingafter surgical intervention. 7 Comput Assist Tomogr1986;10:596-9.

15 Yoshino MT, Temeltas OM, Carter LP, et al. Metallicpostoperative artefacts on cervical MR. AJ7NR Am _7Neuroradiol 1993;14:747-9.

16 Clagman DA, Murakami ME, Vines FS. Compatibilityof the cervical braces with MR imaging: a study of ninenon-ferrous devices. A_7NR Am I Neuroradiol 1990;11:231-390.

17 Miruis SE, Gecsler F, Joslyn JN, Zrebeet H. Use of tita-nium wire in cervical spine fixation as a means toreduce artefacts. AJNR Am 37 Neuroradiol 1988;9:1229-3 1.

18 Williams DF. Editorial; titanium: epitome of biocompat-ibility or cause for concern. Br]I Bone J3oint Surg 1994;76B:348-9.

19 Berry I, Sigal R, Lebas J, Mark AS, Le Bihan D.Magnetic resonance imaging: principles, techniquesand imaging protocols. In: Manelfe C, ed. Imaging ofthe spine and spinal cord. New York: Raven Press, 1992:157-94.

20 Enzmann DR, Pelc NJ. Normal flow patterns of intracra-nial and spinal cerebrospinal fluid defined with phase-contrast cine MR imaging. Radiology 199 1;178:467-74.

21 Terae S, Miyasaka K, Abe S, et al. Increased pulsatilemovement of the hindbrain in syringomyelia associatedwith Chiari malformation: cine MRI with presentationbolus tracking. Neuroradiology 1994;36: 125-32.

22 Maeda M, Itoh S, Kimura H, et al. Vascularity of menin-giomas and neuromas: assessment with dynamic sus-ceptibility-contrast MR imaging. AIR Am ]f Roentgenol1994;163: 18 1-6.

23 Thorpe JW, Kendall BE, MacManus D, Miller DH.Dynamic gadolinium enhanced MRI with detectionand localisation of spinal arterio-venous malforma-tions. Neuroradiology 1994;36:522-9.

24 White SJ, Haginal JV, Young IR, Bydder GM. Use offluid attenuated inversion recovery (FLAIR) pulsesequences for imaging the spinal cord. Magn ResonMed 1992;28:153-62.

25 Montalvo BM, Quencer RM. Intraoperative sonographyin spinal surgery: state of the art. Neuroradiology1986;28:551-90.

26 Plainfosse B, Brunon J, Nelson MD, David P, Hurth M.Intraoperative ultrasound. In: Manelfe C, ed. Imagingof the spine and spinal cord. New York: Raven Press,1992:599-620.

27 Rowland Hill CA, Sibson PI, Britton JA, Hall DMB.Ultrasound of the neonatal conus medullaris: normalposition and use in identification of occult spinal dys-raphism [abstract]. Neuroradiology 1994;36: 165.

28 Di Pietro MA, Venes JL. Real time sonography of thepediatric spinal cord: horizons and limits. Concepts inPediatric Neurosurgery 1988;8: 120-32.

29 Winter RK, McKnight L, Byrne RA, et al.Diastematomyelia: prenatal ultrasonic appearances.Clin Radiol 1989;40:291-4.

30 Naidich TP, Doundoulakis SH, Poznanski AK.Intraspinal masses: effect of plain spine radiography.Paediatric Neuroscience 1986; 12:10-17.

31 Kendall BE. Spinal angiography. In: Du Boulay GH, ed.A textbook of radiological diagnosis. Vol 1. 5th ed. Thehead and CNS. London: Lewis, 1982:563-80.

32 Moseley IF, Tress BM. Extravasation of contrastmedium during spinal angiography, a cause of paraple-gia. Neuroradiology 1977;13:55-7.

33 Hughes JT. Disorders of the spine and spinal cord. In:Hume Adams J, Duchen IW, eds. Greenfield's neuro-pathology. London: Edward Arnold, 1992:1083-116.

34 Fukushima T, Takaaki I, Taoka Y, Takata S. Magneticresonance imaging study of spinal cord plasticity inpatients with cervical compression myelopathy. Spine1991;16:534-8.

35 Yu YL, Jones SJ. Somatosensory evoked potentials incervical spondylosis: correlation of median ulnar andposterior tibial nerve responses with radiological find-ings. Brain 1985;108:273-300.

36 Seibert CE, Barnes J, Dreisback JN, et al. Accurate CTmeasurement of the spinal cord using metrizamide:physical factors. AJNR Am]Nleurobiol 198 1;2:75-8.

37 Fijiwara K, Yonenobu K, Ebara S, Yamashita K, Ono K.The prognosis of surgery for cervical compression inmyelopathy. _7 Bone Joint Surg 1989;71B:393-8.

38 Sherman JL, Nassaux AB, Citrin CM. Measurements ofthe normal cervical spinal cord on MR imaging. AJNRAm .7 Neurobiol 1990;1 1:369-72.

39 Yu YL, Du Boulay GH, Stevens JM, Kendall BE.Computer assisted myelography in cervical spondyloticmyelopathy and radiculopathy. Brain 1986;109:259-78.

40 Fujiwara K, Yonenobu K, Hiroshima K, Ebara S,Yamashita K, Ono K. Morphometry in cases withcompression myelopathy. Spine 1988;13:1212-16.

41 Ogino H, Tada K, Okada K, et al. Canal diameter,antero-posterior compression ratio, and spondyloticmyelopathy of the cervical spine. Spine 1983;8:1-15.

42 Beuls E, Gelan J, Vandersteen M, et al. Microanatomy ofthe excised human spinal cord and the cervico-medullary junction examined with high resolution MRimaging at 9 4 Tesla. A7NR Am .7 Neurobiol 1993;14:699-707.

43 Solsberg MD, Lemaire C, Resch L, Potts DS. High reso-lution MR imaging of the cadaveric human spinal cord:normal anatomy. AJNRAm ] Neurobiol 1990;11:3-7.

44 Ohshio I, Hatayama A, Kaneda K, et al. Correlationbetween histopathological features and magnetic reso-nance imaging of spinal cord lesions. Spine 1993;18:1140-9.

45 Brieg A, Turnbull IM, Hasseter 0. Effects of mechanicalstresses on the cervical cord in cervical spondylosis: astudy on fresh cadaver material. _7 Neurosurg 1966;25:45-66.

46 Stevens JM, O'Driscoll DM, Yu YL, et al. Somedynamic factors in compressive deformity of the cervi-cal spinal cord. Neuroradiology 1987;29:136-42.

47 Stevens JM. The compressed spinal cord. Current med-ical literature. Medical Imaging 1993;5:3-8.

Stevens





48 Stevens JM, Kendall BE, Crockard HA. The spinal cordin rheumatoid arthritis with clinical myelopathy: acomputed myelographic study. J Neurol NeurosurgPsychiatry 1986;49: 140-51.

49 Holsheimer J, Den Boer JA, Struijk JJ, Rozeboom AR.MR assessment of the normal position of the spinalcord in the spinal canal. AJ7NR Am J Neuroradiol 1994;15:951-9.

50 Baleriaux D, Parizel P, Bank WD. Intraspinal andintramedullary pathology. In: Menelfe C, ed. Imagingof the spine and spinal cord. New York: Raven Press,1992:832-90.

51 Stevens JM, Olney JS, Kendall BE. Post-traumatic cysticand non-cystic myelopathy. Neuroradiology 1985;27:48-56.

52 Mehali TF, Pezzuti RT, Applebaum BI. Magnetic reso-nance imaging and cervical spondylotic myelopathy.Neurosurgery 1990;26:217-27.

53 Ono K, Ota H, Tada K, Yamomoto T. Cervical mye-lopathy secondary to multiple spondylotic protrusions:a clinicopathological study. Spine 1977;2:109-25.

54 Kuhn MJ, Mikulis JJ, Ayoub DM, et al. Wallerian degen-eration after cerebral infarction: evaluation withsequential imaging. Radiology 1989;172:179-82.

55 Terae S, Taneichi H, Aburni K. MRI of Walleriandegeneration of the injured spinal cord. J Comput AssistTomogr 1993;17:700-3.

56 Lexa FJ, Grossman RI, Rosenquist AC. MR of Walleriandegeneration in the feline visual system: characterisa-tion by magnetisation transfer rate with histopathologi-cal correlation. AJNR Am J Neuroradiol 1994;15:201-12.

57 Naidich TP, Zimmerman RA, McLone DG, et al.Congenital malformations of the spine and spinal cord.In: Manelfe C, ed. Imaging of the spine and spinal cord.New York: Raven Press, 1992:621-704.

58 Oldfield EH, Muraszko K, Shawker TH, Patronas NJ.Pathophysiology of syringomyelia associated withChiari I malformation of the cerebellar tonsils: implica-tions for diagnosis and treatment. J Neurosurg 1994;80:3-15.

59 Clifton A, Stevens JM, Kendall BE. Idiopathic andChiari associated syringomyelia in adults: observationon the cerebrospinal fluid pathways at the foramenmagnum. Neuroradiology 1991;33(suppl): 167-9.

60 Stevens JM, Serva W, Kendall BE, et al. Chiari malfor-mation in adults: relation of morphological aspects toclinical features and operative outcome. J NeurolNeurosurg Psychiatry 1993;56: 1072-7.

61 Barkovich AJ, Wippold FJ, Sherman JJL, Citrin CM.Significance of cerebellar tonsillar position on MRI.AJ7NR Am JfNeuroradiol 1986;7:795-9.

62 Savy L, Stevens JM, Taylor DJ. Apparent cerebellarectopia: a reappraisal using volumetric MRI.Neuroradiology 1994;6:360-3.

63 Stevens JM, Clifton A, Kendall BE. Relationshipbetween cerebellar tonsillar descent, medullary elonga-tion and the basi cranium in hindbrain deformities ofChiari type [abstract]. Neuroradiology 1994;36: 163.

64 Payner TD, Prenger E, Berger TS, Crone KR. AcquiredChiari malformations: incidence, diagnosis and man-agement. Neurosurgery 1994;34:429-34.

65 Huang PP, Constantine S. "Acquired" Chiari I malfor-mations. J Neurosurg 1994;80:1099-102.

66 Birbamer G, Buchberger W, Felber S, et al. Spontaneouscollapse of post-traumatic syringomyelia: serial mag-netic resonance imaging. Eur Neurol 1993;33:378-8 1.

67 Milhorat TH, Johnson WD, Miller JI, et al. Surgicaltreatment of syringomyelia based on magnetic reso-nance imaging criteria. Neurosurgery 1992;31:231-42.

68 Williams B. Pathogenesis of syringomyelia. Lancet 1972;i: 142-3.

69 Park TS, Cail WS, Broneldus WC, et al. Lumbo-peri-toneal shunt combined with myelotomy for treatmentof syringo-hydromyelia. _7 Neurosurg 1989;70:721-7.

70 Vissilouthis J, Panandreon A, Anagnostasas S.Thecoperitoneal shunt for post-traumatic syringo-myelia. _7 Neurol Neurosurg Psychiatry 1994;57:755-6.

71 Vissilouthis J, Panadreon A, Anagnostasas S.Thecoperitoneal shunt for syringomyelia. Report ofthree cases. Neurosurgery 1993;33:324-8.

72 Vengsarkar VS, Panchal VS, Tripathis PB, et al.Percutaneous theco-peritoneal shunt for syringomyelia.Report of three cases. .7 Neurosurg 199 1;74:827-3 1.

73 Sherman JL, Barkovich AJ, Citrin CM. The MR appear-ances syringomyelia: new observations. A7NR Am .7Neuroradiol 1986;7:985-95.

74 Grant R, Hadley DM, MacPherson P, et al.Syringomyelia-cyst measurement by magnetic-resonance imaging and comparison with symptoms,signs, and disability. _7 Neurol Neurosurg Psychiatry1987;50: 1008-14.

75 Vaquero J, Martinez R, Arias A. Syringomyelia-Chiaricomplex. Magnetic resonance imaging and clinicalevaluation of surgical treatment. . Neurosurg 1990;73:14-68.

76 Hunter JV, Stevens JM, Kendall BE, et al. Radiologicalassessment of transoral surgery in rheumatoid arthritisusing dynamic CT myelography. Neuroradiology 1991;33(suppl):41 3-5.

77 Yu YL, Stevens JM, Kendall BE, de Boulay GH. Cordshape and measurement in cervical spondyloticmyelopathy and radiculopathy. AJNR Am .7 Neuroradiol1 983;4:839-42.

78 Al-Mefty 0, Harkey HL, Marawi I, et al. Experimentalcompressive cervical myelopathy. _7 Neurosurg 1993;79:550-61.

79 Crockard HA, Heileman AE, Stevens JM. Progressivemyelopathy secondary to odontoid fractures: clinicalradiological and surgical features. _7 Neurosurg 1993;78:579-86.

80 Anderson TE. Spinal cord contusion injury.Experimental dislocation of haemorrhagic necrosis andsubacute long axonal conduction loss. _7 Neurosurg1985;62:115-9.

81 Schonman-Claeys E, Frija S, Caenol CA, et al. MRimaging of acute spinal cord injury: results of an exper-imental study in dogs. A.7NR Am _7 Neuroradiol 1990;11:459-65.

82 Yamashita Y, Takahaiki M, Matsumoto Y, et al. Chronicinjuries of the spinal cord: assessment with MR imag-ing. Radiology 1990;175:849-54.

83 Clifton AG, Stevens JM, Whitear PW, Kendall BE.Identifiable causes for poor outcome in surgery forcervical spondylosis. Post-operative computed myelo-graphy and MR imaging. Neuroradiology 1990;32:450-5.

84 Stevens JM, Clifton AG, Whitear P. Appearances of pos-terior osteophytes after sound anterior interbody fusionin the cervical spine: a high definition computed myel-ographic study. Neuroradiology 1993;35:227-8.

85 Hackney DB, Ford JC, Markowitz RS, et al.Experimental spinal cord injury: MR correlations tointensity of injury. .7 Comput Assist Tomogr 1994;18:357-62.

86 Fujii H, Yore K, Sakou I. Magnetic resonance imagingstudy of experimental acute spinal cord injury. Spine1993;18:2030-4.

87 Kulkami MR, McArdle CB, Kapanick D, et al. Acutespinal cord imaging: MR imaging at 1-5T. Radiology1987;164:837-43.

88 Silberstein M, Hennessy 0. Implications of focal spinalcord lesions following trauma-evaluation with mag-netic resonance imaging. Paraplegia 1993;31:160-7.

89 Beers GJ, Rague GH, Wagner SG, et al. Magnetic reso-nance imaging of spinal trauma. _7 Comput Axial Tomogr1988;12:755-61.

90 Falcone S, Quencer RM, Green BA, et al. Progressivepost-traumatic myelomalacic myelopathy: imaging andclinical features. A_7NR Am _7 Neuroradiol 1994;15:747-54.

91 Davies PC, Reisner A, Hudgins PA, et al. Spinal injuriesin children: role of MR. A.7NR Am 7 Neuroradiol 1993;14:607-17.

92 Willis BK, Greiner F, Orison WW, Benzel EC. Theincidence of vertebral artery injury after mid-cervicalfracture or subluxation. Neurosurgery 1994;34:435-42.

93 Bonito V, Agostinis C, Ferraresi S, Defanti CA.Superficial siderosis of the central nervous system afterbrachial plexus injury with pseudo meningoceles._7 Neurosurg 1994;80:931-4.

94 Lasjaunias P, Berenstein A. Surgical neuroangiography.Vol 3. Functional vascular anatomy of brain, spinal cordand spine. New York: Springer Verlag, 1990:15-87.

95 Mawad ME, Rivera V, Crawford S, et al. Spinal cordischaemia after resection of thoraco-abdominal aorticaneurysms: MR findings in 24 patients. A.7NR Am.7Neuroradiol 1990;11:987-91.

96 Yuh WT, Marsh CY, Wang AK, et al. MR Imaging ofspinal cord and vertebral body infarction. A3tNR Am JNeuroradiol 1992;13:145-54.

97 Henderson FC, Crockard HA, Stevens JM. Spinal cordoedema due to venous stasis. Neuroradiology 1993;35:312-5.

98 Rodesch G, Berenstein A, Lasjaunias P. Vasculature andvascular lesions of the spinal cord. In: Manelfe C, ed.Imaging of the spine and spinal cord. New York: RavenPress, 1992:565-98.

99 Willinsky R, Lasjaunias P, Terbrugge K, Hurth M.Spinal angiography in the investigation of spinal arteni-ovenous fistula. A protocol with application to thevenous phase. Neuroradiology 1990;32: 114-6.

100 Gaensler EHL, Jackson DE, Halbach VV. Arterio-venous fistulas of the cervico-medullary junctions as acause of myelopathy: radiological findings in two cases.

A3tNR Am _7 Neuroradiol 1990;11:518-22.101 Biondi A, Merland Ji, Reizine D, et al. Embolization

with particles in thoracic intramedullary arterio-venousmalformations: Long term angiographic and clinicalresults. Radiology 1990;177:651-8.

102 Barmwell SL, Dowd CF, Davis RL, et al. Cryptic vascu-lar malformations of the spinal cord: diagnosis by mag-netic resonance imaging and outcome of surgery._7 Neurosurg 1990;72:403-7.

103 Hida K, Tada M, Chandler WT, et al. Intramedullarydisseminated capillary haemangioma with spinal cordswelling-case report. Neurosurgery 1993;33: 1099-106.

104 Thomas DJ, Pennock JM, Hajnel, et al. Magnetic reso-nance imaging of the spinal cord in multiple sclerosisby fluid attenuated invasion recovery (FLAIR). Lancet1993;314:593-4.

105 Morrissey SP, Miller DH, Kendall BE, et al. The signifi-cance of brain magnetic resonance imaging abnormali-ties at presentation with clinically isolated syndromessuggestive of MS: a 5 year follow up study. Brain1993;116:135-46.

106 Lexa FJ, Grossman RI. MR of sarcoidosis in the headand spine: spectrum of manifestations and radiographic

415




416

response to steroid therapy. AJ7NR Am Jf Neuroradiol1994;15:973-82.

107 Stevens JM. Infections of the central nervous system. In:Butler P, ed. Imaging of the nervous system. London:Springer-Verlag, 1990:107-30.

108 Junger SS, Stem BJ, Levine SR, et al. Intramedullary sar-coidosis-clinical and magnetic imaging characteristics.Neurology 1993;43:333-7.

109 Rogg JM, Benzil DL, Haas RL, Knucky NW. Intra-medullary abscess, an unusual manifestation of a der-mal sinus. AJNR Am Jf Neuroradiol 1993;14: 1393-5.

110 Hardwidge C, Palsingh J, Williams B. Pyomyelia: anintramedullary abscess complicating lumbar lipomawith spina bifida. BrJNeurosurg 1993;7:419-22.

111 Esposito MB, Arrington JA, Murtaugh FR, et al. MR ofthe spinal cord on a patient with Herpes zoster. AJNRAm J Neuroradiol 1993;14:203-4.

112 Demaeral R, Wilms G, Van Lierde S, et al. Lyme diseasein childhood presenting as primary leptomeningealenhancement without parenchymal findings on MR.AINRAmJ Neuroradiol 1994;15:302-4.

113 Naidich TP, Zimmerman RA, McLone DG, et al.Congenital anomalies of the spine and spinal cord. In:Scott Atias, ed. Magnetic resonance imaging of the brainand spine. New York: Raven Press, 1991:865-920.

114 Naidich TP, McLone DG, Harwood-Nash D. Spinaldysraphism. In: Newton PH, Potts DG, eds. Modernneuroradiology. Vol 1. Computed tomography of the spineand spinal cord. San Anselmo, CA: Clavadel Press,1983:299-354.

115 Brooks BS, Duval ER, El Gammal T, et al. Neuro-imaging features of neurenteric cysts: analysis of ninecases and review of the literature. AJNR Am JNeuroradiol 1993; 14:735-46.

Disorders ofhigher corticalfunctionI have written before of Darius Clayhanger's dressingapraxia.' The account was based on Bennett's father.The underlying pathological process is not clear fromBennett's journals although I have suggested that it ispossible that the condition was a rare form of Pick'sdisease. Bennett returns briefly to the problem inThese twain. Most of the other extracts are concernedwith memory failure either as part of senescence, or as

part of a specific dementing illness. Mr Candy, in Themoonstone, develops a memory disorder after a flu-likeillness, conceivably, therefore, the sequela of an

encephalitic illness. Proust's musings, expressed in histypically convoluted sentences, belong more in therealm of philosophy than neurology. It has been sug-gested elsewhere that Mrs Gradgrind's curious inabil-ity to relate her pain to her own body is part of a

parietal disorder, although there is nothing in thenovel to support that interpretation.

Jonathan Swift, 1726, Gulliver's travelsAt ninety they lose their teeth and hair; they have atthat age no distinction of taste, but eat and drinkwhatever they can get, without relish or appetite. Thediseases they were subject to, still continue withoutencreasing or diminishing. In talking, they forget thecommon appellation of things, and the names of per-sons, even of those who are their nearest friends andrelations. For the same reason, they can never amuse

themselves with reading, because their memory willnot serve to carry them from the beginning of a sen-

tence to the end; and by this defect, they are deprivedof the only entertainment whereof they might other-wise be capable.

Charles Dickens, 1854, Hard times"I think there's a pain somewhere in the room," saidMrs Gradgrind, "but I couldn't positively say that Ihave got it."

George Eliot, 1863, RomolaHe was not mad; for he carried within him the piteousstamp of sanity, the clear consciousness of shatteredfaculties; he measured his own feebleness....Would any believe that he had ever had a mind

filled with rare knowledge, busy with close thoughts,ready with various speech? It had all slipped awayfrom him-that laboriously gathered store . . . buthe found, to his acute distress, that of the new

details he learned he could only retain a few, andthose only by continual repetition; and he beganto be afraid of listening to any new discourse, lest itshould obliterate what he was already striving toremember ....

Old men's eyes are like old men's memories; theyare strongest for things a long way off.

Wilkie Collins, 1868, The moonstoneHere, he got on glibly enough. Trumpery littlescandals and quarrels in the town, some of them asmuch as a month old, appeared to recur to hismemory readily. He chattered on, with something ofthe smooth gossiping fluency of former times. Butthere were moments, even in the full flow of histalkativeness, when he suddenly hesitated-lookedat me for a moment with the vacant inquiry once morein his eyes-controlled himself-and went on again.

Fyodor Dostoyevsky, 1869, The idiotThe General talked for ten minutes, heatedly andrapidly, as though too engrossed for time to expressthe thoughts that crowded in his head; towards theend, tears glistened in his eyes. And yet it was onlysentences without beginning or end, unexpectedwords and unexpected ideas, rapidly and unexpectedlybursting forth and stumbling over one another.

Arnold Bennett, 1910, ClayhangerFor many months now he had helped Darius to dress,when he came up from the shop for breakfast, and toundress in the evening. It was not that his fatherlacked the strength, but he would somehow lose him-self in the maze of his garments, and apparently hecould never remember the proper order of doffing ordonning them. Sometimes he would ask, "am I dress-ing or undressing?" And he would be capable of soinvolving himself in a shirt, if Edwin were not there todirect, that much patience was needed for his extrica-tion. His misapprehensions and mistakes frequentlyreached the grotesque. As habit threw them more andmore intimately together, the trusting dependence ofDarius on Edwin increased. At morning and eveningthe expression of that intensely mournful visageseemed to be saying as its gaze met Edwin's, "here isthe one clear-sighted, powerful being who can guideme through this complex and frightful problem of myclothes." A suit, for Darius, had become as intricate asa quadratic equation.

Arnold Bennett, 1916, These twain. . .but it had witnessed hundreds of monotonoustragic meals at which the progress of his father's men-tal malady and the approach of his death could bemeasured by the old man's increasing disability to dis-tinguish between his knife and his fork.

Marcel Proust, 1919, Remembrance of things past:within a budding groveFor our memory, relatively to the complexity of theimpressions which it has to face while we are listening,is infinitesimal, as brief as the memory of a man whoin his sleep thinks of a thousand things and at onceforgets them, or as that of a man in his second child-hood who cannot recall a minute afterwards what onehas first said to him.

Continued on page 421

NEUROLOGY IN LITERATURE

Stevens




Imaging of the spinal cord.

J M Stevens

doi: 10.1136/jnnp.58.4.4031995 58: 403-416 J Neurol Neurosurg Psychiatry

http://jnnp.bmj.com/content/58/4/403.citationUpdated information and services can be found at:

These include:

serviceEmail alerting

box at the top right corner of the online article. Receive free email alerts when new articles cite this article. Sign up in the

Notes

http://group.bmj.com/group/rights-licensing/permissionsTo request permissions go to:

http://journals.bmj.com/cgi/reprintformTo order reprints go to:

http://group.bmj.com/subscribe/To subscribe to BMJ go to:


http://jnnp.bmj.com/content/58/4/403.citation

http://group.bmj.com/group/rights-licensing/permissions

http://journals.bmj.com/cgi/reprintform

http://group.bmj.com/subscribe/



imaging ofthe spinal cord

Documents