107212ournal ofNeurology, Neurosurgery, and Psychiatry 1993;56:1072-1077 in adults: relation of J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.10.1072 on 1 October 1993. Downloaded from morphological aspects to clinical features and operative outcome

J M Stevens, W A D Serva, B E Kendall, A R Valentine, J R Ponsford

Abstract headache and neck pain, and those most To determine whether clinical features refractory were often and nystagmus. attributed to cerebellar ectopia could be Clearly not all patients with the adult related to the severity of the malforma- Chiari malformation benefit from operation. tion, and if morphological features could Attempts at risk stratification have been be related to operative outcome, a retro- made, and recently a careful study by spective study of 141 patients with the Menezes' group14 established three preopera- adult Chiari malformation was carried tive clinical features predictive of a poor oper- out, 81 receiving operative treatment. ative outcome-namely, ataxia, scoliosis, and Morphological parameters derived from muscle atrophy. Attempts at relating the preoperative clinical imaging were com- severity of the Chiari lesion to individual clin- pared with presenting clinical features ical features or syndromes have been few and and postoperative outcomes. Patients some of the conclusions conflicting.15-'7 There with the most severe cerebellar malfor- has been no formal attempt at operative risk mation, defined as descent of the cere- stratification based on morphological fea- bellar tonsils to or below the , had tures. disabling ataxia and nystagmus more The present study was begun several years frequently. Those with com- ago to investigate if the severity of the hind pression had limb weakness and muscle abnormality could be related to preop- wasting more frequently. Operative out- erative clinical features and postoperative out- come was significantly less favourable in come. Because the design of the investigation patients with severe cerebellar ectopia was retrospective, and large numbers of (12% improved, 690% deteriorated) than patients with detailed clinical data were in those with minor ectopia (50% required, it was necessary to use data mainly improved, 17% deteriorated). Patients from myelography and CT rather than MRI with a distended cervical syrinx had a which only became available at our institution more favourable outcome than those some time after the study began. Nevertheless, without. Morphological features help most of the criteria used in the study are predict operative risk. equally applicable to MRI.

(jNeurol Neurosurg Psychiatry 1993;56:1072-1077) http://jnnp.bmj.com/ Materials and methods The study consisted of a retrospective review The term adult Chiari malformation is used of 141 symptomatic adult patients diagnosed to describe descent of the hind brain into the as having a Chiari malformation by myelogra- cervical canal, in which meningomyelocoele is phy alone in 91 or in combination with CT in absent and hydrocephalus rare, and patients 50 patients. At least five of these patients also usually present in adult life.' The clinical fea- had MRI performed on a Vista MR (Picker tures which may be associated are protean26 International) operating at 05 or 0-26 T. on September 28, 2021 by guest. Protected copyright. and a causal relationship in a given case may One patient only had hydrocephalus, and not be established with certainty, especially as none had a meningomyelocoele. Eighty one modern computed imaging suggests that min- patients were operated on in one of three sur- imal protrusion of parts of the cerebellar gical units; all were subjected to a posterior The National the Hospitals for hemispheres through is suboccipital and upper cervical decompres- Neurology and relatively common, and that even marked sion, 12 had partial excision of herniated Neurosurgery, Queen protusion is often asymptomatic.7 The issue is cerebellar tonsils, and eight had an occlusion Square and Maida important because it is often maintained that of the . Thirty one patients also had a Vale, London, UK to J M Stevens disability due the adult Chiari malforma- syringostomy. Sixty patients did not receive W A D Serva tion usually is curable by operation.' 5 operations at our institution: 32 were lost to B E Kendall In the nine operative series which have follow up and may have been on A R Valentine operated included more than in J R Ponsford 20 patients and which elsewhere, 18 had significant co-existing dis- the mean was at Correspondence to: follow up least 6 months, the ease such as multiple sclerosis (10 patients) Dr J M Stevens, mean proportion and patients with sustained and severe spondylotic compres- Department of Radiology, Maida Vale Hospital, improvement has been 56% (range 29-82%), sion (eight patients), eight refused operation, London W9 ITL. the proportion who deteriorated 24% (range and it was unclear why no operation was per- Received 22 October 1992 6-31%) and the perioperative mortality 3% formed in two. and in revised form 8 January 1993. (range 0_5-6%).1 4 The clinical features Clinical data were gleaned from the case Accepted 19 January 1993 most likely to be alleviated have been notes and evaluated by one of two neurolo- Chiari malformation in adults: relation ofmorphological aspects to clinicalfeatures and operative outcome 1073 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.10.1072 on 1 October 1993. Downloaded from Table 1 Clinical groupings. arches of atlas and axis; and C, tonsils below the axis. The upper border of each neural (a) General 1 Headache: considered only if it was exacerbated by head arch was taken as the reference point for each movement, exercise, or coughing. group. 2 Drop attacks: unexplained falls with or without brief loss of consciousness. 3 Neck pain: with or without occipital radiation. FORAMEN MAGNUM (b) Functional anatomical The degree of obstruction at the foramen 1 Upper and brainstem (, ): dysfunction of cranial m, IV, V, VI, VII, and VIII. magnum was also assessed. Contrast medium 2 Medulla oblongata: dysfunction of IX, X, flowed freely into the head from the spinal XI, xII. 3 Brainstem-cerebelar connections: ataxia, vertigo, subarachnoid space via the anterior cere- nystagmnus. brospinal fluid pathway on myelography in all 4 Peripheral sensory disturbances: impaired pain and temperature sensation; loss of position and vibratory cases, although in some it was obstructed ; subjective complaints of numbness of posteriorly. Two types of observation were parasthesiae in limb or trunk. 5 Peripheral motor disturbances: all forms of weakness in made: limbs or trunk, with or without muscle wasting. 1 The size of the was assessed 6 Syringomyelia: this category included only those patients in which there was a firm diagnosis of both above and at the level of the foramen syringomyelia in the case notes, based on clinical and magnum and its mid-sagittal size determined radiological findings. by measuring the minimum distance between (c) Postoperative clinical categories 1 Original clinicalfeatures: The preoperative symptoms/ the internal occipital crest and posterior mar- signs were classified postoperatively as either improved, gin of the foramen magnum and the cerebel- unchanged or worse. 2 New clinicalfeatures: any complaints or signs appearing lum. CT or midline sagittal tomograms and after operation which was not present before, excluding MRI were used wherever possible for this those directly related to the wound itself. 3 Headache: this feature was considered separately when assessment. Accurate measurement was not present in the preoperative clinical picture. possible, but the cisterna magna was scored as obliterated if it was not visible, small if its mid-sagittal width was less than 2-3 mm, and gists (WS and JP) not involved in the routine normal if it was greater than 2-3 mm. management of the patients. A large volume 2 Contrast medium visible in thefourth ventri- of data was obtained which, for the purposes cle either during the myelogram, or on CT of analysis, was simplified by arranging it into within 1 hour of the myelogram provided the categories and groups indicated in table 1. concentration was similar to that in the poste- Missing data in any individual case were rior fossa cisterns, was recorded. This was recorded as such, and the case excluded from taken as evidence that either one or each of analysis of that feature. Data sets were judged the foramena of Magendie or of Lushka were to be adequate for 127 patients. The postop- patent. Arachnoiditis was considered present erative assessments were divided into early if the cisterna magna was obliterated and the and late. All early assessments were made failed to opacify on computed within 1 month of operation and features myelography. referable to the surgical wound were excluded, such as neck pain and stiffness. SYRINGOMYELIA The late assessments were made, as far as The upper spinal cord was also assessed for possible, from the last outpatient attendance. the presence of syringomyelia. Syringomyelia This varied from 3 months to 9 years after was considered probable if there was a central http://jnnp.bmj.com/ operation; but in only 18 did this period accumulation of contrast medium within the exceed 2 years and the mean follow up period spinal cord on early or delayed post- was 1 2 years. The available non-operated myelography CT, or a circumscribed central group was considered too small and heteroge- signal change was visible on MRI, or if there neous to be used for a meaningful parallel was a diffuse abnormality of cord size not group follow up study. explained by compression. A mid-sagittal

The radiological features were assessed by diameter of the spinal cord of greater than 10 on September 28, 2021 by guest. Protected copyright. three experienced neuroradiologists (aS, BK, mm was regarded as definitely enlarged, and AV). of less than 6 mm as definitely small. Measurements between were regarded as CEREBELLAR TONSILS normal. The level of the cerebellar tonsils was assessed with head and spine either slightly MEDULLA OBLONGATA extended or slightly flexed but never hyper- Compression of the medulla oblongata was flexed. The distance the most inferior part of also considered. This was often difficult to one or both cerebellar hemispheres assess on myelography, and only assessments descended below the foramen magnum was made on adequate computed cisternograms measured in millimetres on all available and MRI were analysed. The anterior and images, and the level also was noted relative posterior surfaces were considered separately, to easily recognisable anatomical landmarks. and compression was considered present Magnification was not adequately standard- when the whole or part of the contour was ised between examination types, so only the flattened or concave on axial images. The relative levels, reasonably consistent between posterior surface of the medulla was assessed myelography, computed myelography, and below the obex because the floor of the fourth MRI were considered in the final analysis. ventricle usually is concave. The compressing Patients were grouped into three categories: agent also was recorded, and was either the A, tonsils between foramen magnum and apex of the dens, an abnormality of the clivus, arch of the atlas; B, tonsils between the or the cerebellar tonsils acting either alone or 1074 Stevens, Serva, KendaU, Valentine, Ponsford J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.10.1072 on 1 October 1993. Downloaded from Table 2 Clinicalfeatures in 141 patients* Results The five radiological variables were tested for Symptoms* Signs* an association with each of the 15 clinical Headache 41 variables which included both early and late Neck pain 46 postoperative assessments. There were 75 Drop attacks 19 - Pons-midbrain 27 44 sets of contingency tables of which 17 showed Medulla oblongata 49 61 significant associations (24%). In addition, Brainstem/cerebellar 48 73 Peripheral sensory 67 67 the six postoperative clinical variables in the Peripheral motor 33 92 early and late assessments were compared *Patients often had more than one clinical feature and with each of the preoperative clinical vari- adequate data sets were available in only 127 patients. ables, to create a further 90 contingency tables, of which only four indicated signifi- Table 3 Postoperative clinical assessments cant associations (4%). The duration of symptoms was less than 5 Early Late years in 46% and more than 10 years in 26%. n % n % Patients' ages ranged from 11 to 78 years, but Original Features: 54% were aged between 32 and 45 years. Improved* 20 25 23 34 The distribution of preoperative clinical fea- Unchanged 31 38 19 28 Worse 30 37 25 37 tures in the sample is shown in table 2, Total 81 100 67 100 New features: 23/81 28 21/67 31 although it is noted that adequate sets were Headache: available in only 127 patients. The outcomes Improved 21 68 17 63 in the 81 operated patients are indicated in Unchanged 1 13 3 11 Worse 9 29 7 26 table 3. However, adequate sets of clinical Total 31 100 27 100 and radiological data were available for the *This implies that one or more cardinal features showed defi- early postoperative assessment in only 73, nite improvement; usually it was only one. No case appeared to and in 60 for the late assessment. Headache have improved in one direction and deteriorated in another. was included only if specifically mentioned in the postoperative notes, so that only 31 cases in combination. Compression was graded had valid early assessments, and only 27 valid visually as mild, moderate, or severe-severe late assessments. Sixty five patients had radi- indicating that the mid-sagittal diameter of ological evidence of syringomyelia (55%). the neural axis at the point of maximal com- Basilar invagination or other significant bony pression was reduced by 50% or more. abnormalities were recognised in 17 cases. All the clinical and radiological features Associations between several clinical fea- were assigned a numerical code and the data tures and radiological parameters which were analysed using the Statistical Package yielded probability values less than 0-05 are for the Social Sciences (SPSS) version 7 pro- indicated in tables 4 and 5. In addition, an gram, on the University of London Computer enlarged spinal cord was associated with the Centre Amdahl Computer. The incidence of presence of peripheral sensory disturbances each clinical feature was compared with the which were present in 73% of 63 valid cases presence or absence of each radiological fea- in whom the cord was enlarged, and in only ture, considered independently. In some 17-5% of 40 valid cases in whom the cord instances a relationship was sought between was normal or small (p = 0 0001). It was not http://jnnp.bmj.com/ individual radiological features, but the analy- possible to investigate the effect of a small sis was concerned mainly with relationship of spinal cord as opposed to a normal cord, radiological to clinical features, especially since the numbers recorded as having a small operative outcome. X2 tests of association cord were too small. Similarly there was a were used, except when an expected strong positive association between abnormal frequency was less than five, in which case a contrast accumulation in the cord on post- dis- Fisher's exact test was used. myelography CT and peripheral sensory on September 28, 2021 by guest. Protected copyright. turbance (21 of 28 valid cases with contrast in the cord had sensory disturbance com- pared with seven of 22 without, p = 0 002). Interestingly, syringomyelia showed no association with obliteration of the cisterna magna or opacification of the fourth ventricle Table 4 Clinicalfeatures related to tonsillar descent (p = 0-130). A significant association was found Level oftonsils* between operative outcome, and the level of (0-Cl) (Cl-C2) (>C2) p Value cerebellar tonsils (table 6). A further associa- tion was found between clinical improvement Medulla (symptoms) 16/53 14/49 14/25 0-04 (30%) (29%) (56%) and the presence of an enlarged spinal cord Peripheral motor (sign) 28/53 32/49 21/25 0-02 (53%) (65%) (84%) Brainstem/Cerebellar 19/53 26/49 18/25 0-01 (signs) (36%) (53%) (72%) Table S Clinicalfeatures related to size ofcisterna Peripheral sensory 29/53 27/49 6/24 0-03 magna (symptoms) (55%) (53%) (25%) Neck pain 24/53 11/48 5/25 0-02 Normal Small Obliterated p Value (47%) (23%) (20%) Syringomyelia 28/49 31/44 6/25 0-00 Medulla oblongata 5/18 31/74 10/12 0-02 (57%) (70%) (20%) (signs) (28%) (42%) (83%) Brainstem-cerebellar 3/18 39/74 10/12 0 00 *0-C1, tonsils between occiput and the atlas ring; C 1-2, tonsils between the atlas and axis; (signs) (16%) (53%) (83%) >C2, tonsils below the axis. Chiari malformation in adults: relation ofmorphological aspects to clinicalfeatures and operative outcome 1075 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.10.1072 on 1 October 1993. Downloaded from Table 6 Operative outcome with respect to level oftonsils 17, clivus 1) and posterior in three (cerebellar Level of tonsils tonsils). Analysis of other aspects of medullary mor- Preoperative (0-Cl) (Cl-C2) (>C2) clinicalfeatures n % n % n % phology, such as length, level of obex and , presence of spur or Early: Worse 5 18 14 47 1 1 69 kink, were possible in only 16 cases with Unchanged 13 48 1 1 37 3 19 computed cisternograms and the five cases Improved 9 33 5 17 2 12 Total 27 100 30 100 16 100 who also had MRI. In seven there was a cer- Late: vicomedullary spur just below the ectopic Worse 3 17 7 26 9 60 Unchanged 6 33 7 26 4 27 cerebellar tonsils over which the distal part of Improved 9 50 13 48 2 13 the fourth ventricle was prolonged, the obex Total 18 100 27 100 15 100 presumably lying under the spur. In all seven By combining the categories of unchanged and improved, p values were calculated as 0-02 for the cerebellar tonsils were at or below C2 early and 0 04 for late assessment. (one was in group B, and six were in group C). There were no independent clinical fea- on early (p = 0-01) but not on late (p = tures distinguishing this small group of 0 082) assessment (table 7). patients, though none had syringomyelia. The question of arachnoiditis could not be As would be expected from previous find- addressed directly in this study. The combi- ings, poor operative outcome showed a rela- nation of obliteration of the cistema magna tionship with the presence of and non-filling of the fourth ventricle may difficulties, truncal ataxia, and nystagmus (p have been due to arachnoiditis in many cases. = 0-025 in each instance), all of which had Both were associated with the cerebellar ton- shown an association with marked tonsillar sils lying below C2, but only non-filling of the descent. Patient age and duration of clinical fourth ventricle showed a relationship with symptoms were considered, and no signifi- poor operative outcome. When the fourth cant association was found with operative ventricle was opacified, two of 15 patients outcome on two sample t-tests. Mean ages in were worse after operation compared with 28 the worse, unchanged, and improved groups of 50 when the fourth ventricle did not were 41-6, 41-1, and 32-0 years; and mean opacify (p = 0-001). The role of arachnoiditis duration of symptoms in the same groups as opposed to apparent level of the cerebellar were 5 6, 4-1, and 4 5 years. tonsils could not be established by this analy- sis. Post-myelography CT studies all too fre- Discussion quently were of limited extent, and often The distribution of clinical features indicated delayed until the contrast density in the spinal that this sample of patients was similar to that and intracranial subarachnoid spaces was in other comparable studies, and the results much reduced, preventing detailed analysis. of subocciptal decompression, when due At the time the potential value of detailed allowance was made for varying methods of assessment was not adequately appreciated, a assessment, were generally compatible with situation which this study originally was the consensus of published reports already designed to redress. Therefore, only 28 cases cited. The mean follow up was relatively permitted what we now regard as an adequate short, many other series having mean periods http://jnnp.bmj.com/ estimation of medullary compression. The of more than 2 years. The group of operated majority of attempted clinical associations patients with mild cerebellar tonsillar descent with compression failed to suggest any associ- (group A), who generally fared best, were ation but of the 21 who had some compres- most frequently lost to follow up. Although sion 18 had peripheral motor signs (mainly the proportion of positive associations was weakness and wasting), and only three of similar to that which might have been

seven without compression had such signs (p expected by chance, the positive findings on September 28, 2021 by guest. Protected copyright. = 0 043, Fisher's exact test). Only 17 cases were broadly in accord with those of Elster et were suitable for analysis of operative out- al"7 and Dyste et al14 and, where relevant, come, and although five patients with mild showed the interrelationships expected from compression were unchanged, whereas nine true as opposed to spurious associations. For of 12 with moderate or severe compression example, marked tonsillar descent was associ- suffered a deterioration in original signs and ated with the preoperative ataxia and poor three also developed new features, a Fisher's operative outcome, and poor operative out- exact test indicated a p value of 0-06 which come was associated with preoperative ataxia. was not significant. The major compressing The most frequent relationships involved agent was judged to be anterior in 18 (dens the level of the cerebellar tonsils. It has been observed at operation that sometimes the ton- sils are found at a lower level than suggested Table 7 Operative outcome* with respect to spinal cord by myelography.l9 Possibly the altered size hydrodynamics at the craniovertebral junc- Nornal Enlarged tion caused by opening the dura explains at Preoperative clinical least some such discrepancies, and hyperflex- features (early) n % n % ion of the head may explain others. MRI is Worse 10 59 8 23 now the imaging method of choice and how Unchanged/better 7 41 28 82 estimates of tonsillar descent compared with Total 17 100 34 100 those from myelography in the same group of *Based on early assessment (p = 0-01). patients is unknown. The measurements 1076 Stevens, Serva, Kendall, Valentine, Ponsford J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.10.1072 on 1 October 1993. Downloaded from made by Aboulezz et al7 on MRI compared not severe malformations, the latter being with those made from myelography by associated with more disabling clinical mani- O'Connor et a120 and Baker,21 suggest that festations. they appear slightly lower on MRI. In this study, important associations also Observations from high definition computed were found between tonsillar descent and cisternography have indicated that the medial operative outcome. Unfavourable outcomes parts of the biventral lobules are nearly always were more frequent with increasing tonsillar inferior to the tonsils in normal cerebellar descent. With severe descent (group C) only hemispheres and lie in the plane of the fora- 12% were improved and 69% had deterio- men magnum posterolateral to the brainstem, rated on late assessment, whereas with mild very occasionally protruding slightly below.22 descent (group A) only 18-5% had deterio- Only when ectopic do the cerebellar tonsils rated. Clinical features least likely to improve enter the foramen magnum or protrude were those related to brainstem/cerebellar below it, usually lying directly posterior to the connections and medulla oblongata, espe- brainstem. Most MRI assessments are made cially ataxia, as was also the case in the recent from mid-sagittal images usually 3 to 5 mm report by Menezes' group.'4 The difference thick in which this distinction often cannot between outcomes on late assessment in reliably be made. It is even more difficult on group A and B was not significant, so once myelography unless thin section tomography again it was group C which was different. is used. In our cases, all that could reliably be Outcome of both preoperative cough and said was that some part of the cerebellar posture related headache showed no relation hemispheres was visible in the cervical part of to tonsillar descent or any other imaging the cisterna magna. In the cases with CT, parameter including size of cisterna magna, protrusion of the actual cerebellar tonsils yet the latter feature was significantly could be confirmed and the relative degree of improved in 62-6% cases. Posture and cough protrusion correlated reasonably on MRI and related headache, like drop attacks, are myelographic assessment. thought to result from intermittent tonsillar Computed cisternography and MRI indi- impaction in the foramen magnum27 which cate that patients with herniated cerebellar suboccipital craniectomy and duroplasty usu- tonsils often have an elongated and some- ally relieve very well.26 Therefore the lack of times even kinked medulla oblongata in association of such features with small or which the obex lies below the foramen mag- obliterated cisterna magna or low lying tonsils num,2325 as was the case in at least seven of both pre and postoperatively suggests that the our patients and probably also in many more. origin of these symptoms is more complex. In many publications such a configuration Small size or obliteration of the cisterna would be classified as Chiari type 1.f 13-17 24 magna, like increasing tonsillar descent, was The association between increasing tonsillar related to clinical features localisable to the descent and obliteration or smallness of the medulla oblongata and brainstem cerebellar cisterna magna probably accounted for the connections. The lack of a definite relation low incidence of filling of the fourth ventricle between the size of the cisterna magna and in group C. However, it was less clear why other clinical features which were related to more severe tonsillar descent should have level of the tonsils, such as operative out-

been associated with a significantly low inci- come, suggests that the level of tonsils is the http://jnnp.bmj.com/ dence of syringomelia. The explanation may more important of these interrelated parame- reside in the significance of abnormalities of ters. It was surprising that no association was the medulla oblongata such as the kink and found between the size of the cisterna magna spur, the incidence of which is increased in and syringomyelia, but once again the expla- patients with more severe malformations.25 nation in this study lies in group C where the Such morphology is associated with a lower cisterna magna usually was small or obliter- incidence of syringomyelia, possibly because ated and syringomyelia uncommon. However the obex lies below the obstructing tonsils.219 in a separate study using MRI, a lack of a on September 28, 2021 by guest. Protected copyright. The level of the cerebellar tonsils showed a relation between syringomyelia and the size of similar association with presenting clinical the CSF pathways across the foramen mag- features to those found in a recent MRI num also was observed.25 study. 17 Features referable to the medulla Opacification of the fourth ventricle with oblongata were mainly difficulties with swal- water soluble contrast medium nearly always lowing and phonation, and those referable to occurs during examination of the cranioverte- brainstem cerebellar connections were mainly bral junction in normal patients. It is a fea- ataxia and nystagmus. These, together with ture which cannot be determined using MRI peripheral signs, appeared and was of potential interest in view of theo- significant because of their relatively high ries still current about the causation of Chiari frequency in patients with severe tonsillar associated syringomyelia. It was recorded in descent (group C). The relationship with only about 30% in this study which probably peripheral sensory symptoms was mainly due was a sig-nificant departure from normal. The to the high frequency in patients with moder- maximum frequency of filling occurred in ate tonsillar descent (group B), and there was group B, which was the group with the high- a close relationship between peripheral sen- est prevalence of syringomyelia. This was sur- sory disturbances and radiological signs of prising perhaps, since it is widely accepted syringomyelia. It was no surprise perhaps to that obstruction of the outlets of the fourth have found that neck pain appeared a signifi- ventricle and cisterna magna by ectopic cere- cant feature in the presentation of mild but bellar tonsils and arachnoiditis are central to Chiari malformation in adults: relation ofmorphological aspects to clinicalfeatures and operative outcome 1077 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.10.1072 on 1 October 1993. Downloaded from the development of syringomyelia,2>32 and 3 Dobkin BH. in the adult Chiari anomaly. Neurology 1978;28:718-20. strategies to correct this underpin operative 4 Gordon D. Neurological syndromes associated with cran- treatment. Once again we believe the expla- iovertebral anomalies. Proc R Soc Med 1969;62:725-6. 5 Bronstein AM, Miller DH, Rudge P, Kendall BE. Down nation in this study was simply the low inci- beating nystagmus: magnetic resonance imaging and dence of fourth ventricle filling and of neurological findings. JNeurol Sci 1987;81:173-84. 6 Caetano de Barros M, Farias W, Ataide L, Lins S. Basilar syringomyelia in patients of group C, because impression and Arnold Chiari malformation. A study the differences between groups A and B were of 66 cases. J Neurol Neurosurg Psychiatry 1968;31: 596-605. not significant. This probably also is the main 7 Aboulezz AD, Sartor K, Geyer CA, Gado MH. Position reason why patients with filling of the fourth of cerebellar tonsils in the normal population and in patients with Chiari malformation: a quantitative ventricle and an enlarged spinal cord had a approach with MR imaging. Jf Comput Assis Tomogr better operative outcome than those without, 1985;9: 1033-6. 8 Banerafi NK, Millar JHD. Chiari malformation presenting because most of the former were in groups A in adult life. The relationship to syringomyelia. Brain and B. 1974;97: 157-68. 9 Saez R, Onofrio B, Yanagihara T. Experience with Arnold Medullary compression could be evaluated Chiari malformation, 1960-1970. J Neurosurg 1976;45: adequately in only a small cohort of patients 416-22. 10 Garcia-Uria J, Leunda G, Carrillo R, Bravo G. Syringo- and we did not attempt to relate this to other myelia: long-term results after posterior fossa decom- radiological variables such as level of the ton- pression. J Neurosurg 1981;54:380-3. 11 Logue V, Edwards MR. Syringomyelia and its surgical sils. Although multivariate analysis would be treatment; an analysis of 75 patients. J Neurol Neurosurg required to distinguish the roles of this, and Psychiaty 1981;44:273-85. 12 Di Lorenzo N, Aldo F, Guidetti B. Craniovertebral junc- other features in this study in relation to out- tion malformations. Clinico-radiological findings, long- come, it did seem to be related to one clinical term results and surgical indications in 63 cases. J Neurosurg 1982;57:603-8. feature, namely motor peripheral signs, but 13 Eisenstat DDR, Bemstein M, Fleming JFR, et al. Chiari only when compression was severe. There malformation in adults. A review of 40 cases. Can J Neurol Sci 1986;13:221-8. seemed to be no relationship with syringo- 14 Dyste SN, Menezes AH, Van Gilder JC. Symptomatic myelia which differed from the conclusions of Chiari malformations. An analysis of presentation, man- agement and longterm outcome. Jf Neurosurg 1989;71: Spinos et al24 but was similar to those of 159-68. Clifton et al,25 both based on MRI. It was 15 Curnes JT, Oakes WJ, Boyko OB. MR imaging of hind- brain deformity in Chiari II patients with and without expected that severe medullary compression symptoms of brainstem compression. AJNR 1989;10: also may have been relevant to some poor 293-302. 16 Wolpert SM, Scott RM, Platenberg C, Runge VM. The operative outcomes, which would be compat- clinical significance of herniation and defor- ible with the known significance of severe mity as shown on MR images of patients with Chiari II malformations. AJNR 1988;9:1075-8. spinal cord compression in cervical spondylo- 17 Elster AD, Chen MY, Chiari I malformations: clinical and sis33 and atlanto-axial subluxation,34 but num- radiological re-appraisal. Radiology 1992;183:347-53. 18 Batzdorf V. Chiari I malformation with syringomyelia. bers of patients in this study were too small to J Neurosurg 1988;68:726-30. permit any firm conclusion. The major com- 19 Rhoton AL. Microsurgery of the Arnold Chiari malforma- tion in adults with and without hydrocephalus. pression usually was anterior, where it was J Neurosurg 1976;45:473-83. likely to be made worse by flexing the head 20 O'Connor S, du Boulay G, Logue V. The normal position of the cerebellar tonsils as demonstrated by myelogra- to facilitate a satisfactory posterior operative phy. J Neurosurg 1973;39:387-9. approach.35 21 Baker H L. Myelographic examination of the posterior fossa with positive contrast medium. Radiology 1963; 81:791-801. 22 Stevens JM, Kendall E. Aspects of the anatomy of the on computed tomography. Neuroradiology Conclusion 1985;27:390-5.

The main outcome of this analysis has been 23 Naidich TP, McLone DS, Fulling KH. Chiari malforma- http://jnnp.bmj.com/ tion: Part IV. The hindbrain deformity. Neuroradiology to identify patients with tonsillar descent 1983;25: 179-97. below the upper border of the neural arch of 24 Spinos E, Laster DW, Moody P, et al. MR evaluation of the Chiari malformation at 0-15T. AJR 1985;144: C2 as a special group. They had the highest 1143-8. frequency of disabling ataxia and nystagmus, 25 Clifton AG, Stevens JM, Kendall BE. Idiopathic and Chiari associated syringomyelia in adults: observations the lowest frequency of syringomyelia, and on cerebrospinal fluid pathways at the foramen magnum suboccipital decompression was most likely to using static MRI. Neuroradiology 199 1;33 (suppl): 167-9. 26 Clifton A, Stevens JM, Kendall BE. Morphological fea- result in deterioration rather than improve- tures and their interrelationships in the adult Chiari ment. On the other hand patients with the malformation: an MRI study. Neuroradiology (in print). on September 28, 2021 by guest. Protected copyright. 27 Larson SJ, Sances A, Baker JB, Reigel DH. Herniated most favourable operative outcomes were cerebellar tonsils and cough syncope. J Neurosurg those with mild tonsillar descent and a dis- 1974;40:524-8. 28 Williams B. Cerebrospinal fluid pressure changes in tended cervical syrinx. Clinical malfunction response to coughing. Brain 1976;99:331-46. in the cerebellar/brainstem connections prob- 29 Barnett HJM, Foster JB, Hudgson P, eds. Syringomyelia. London: Saunders, 1973. ably is related to the severity of the hind brain 30 Newman PK, Terenty TR, Foster JB. Some observations anomaly, and when the anomaly is severe, on the pathogenesis of syringomyelia. JfNeurol Neurosurg Psychiatry 1981;44:964-9. clinical features are most likely to be end- 31 Newton EJ. Syringomyelia as a manifestation of defective points. This conclusion is supported by con- fourth ventricular drainage. Ann Roy Coll Surg Engl 1969;44: 199-214. tinuing observations at our institution where 32 du Boulay G, Shah S H, Currie J C, Logue V. The mech- MRI virtually is now the only imaging modal- anisms of hydromyelia in Chiari type I malformations. BrJRadiol 1973;47:579-87. ity used in preoperative assessment of this 33 Fujiwara K, Yenenobu K, Ebara S, Yamashita K, Ono K. condition. The prognosis of surgery for cervical compression myelopathy. An analysis of the factors involved. JfBone Joint Surg 1989;71-B:393-8. 34 Hunter J, Stevens JM, Kendall BE, Moskovich R, 1 Paul KS, Lye RH, Strang FA, Dutton J. Arnold Chiari Crockard HA. Radiological assessment for transoral malformation. Review of 71 cases. J Neurosurg 1983;58: surgery in rheumatoid arthritis, using dynamic com- 183-7. puted myelography. Neuroradiology 1991;33 (suppl): 2 Mohr PD, Strang FA, Sambrook MA, Boddie HG. 413-5. Clinical and surgical features in 40 patients with pri- 35 Brieg A. Biomechanics of the central . Some mary cerebellar ectopia (Adult Chiari malformation). basic normal and pathological phenomena. Uppsala: QJ7Med 1977;181:85-96. Almquist and Wiksells, 1960.