Letters to the editor
438
spinal cord; also the cord was not atrophic. The images of the brain were also judged to be normal. On a T2-weighted image there was an abnormal high intensity area in the centre of the spinal cord where the syrinx had been. There was no CSF flow-void sign or abnormality of cord size. On MRI, after a follow up period of two years and three months, the Chiari malformation and syrinx had disappeared on the T1weighted image. On the T2-weighted image, however, a high intensity area was clearly visible in the spinal cord. A high intensity area on T2-weighted image showed a nonspecific change and could be due to gliosis, oedema, microcystic changes and/or demyelination.67 We therefore assumed that the syrinx had collapsed and been replaced by
gliosis. The reduction of pain, hand muscle weakness, pathological reflexes and severity of anaesthesia are likely to reflect the disappearance of the syrinx, even though the hyporeflexia of the left arm remained unchanged and the area of sensory disturbance was somewhat enlarged. There are several theories about the pathogenesis of syringomyelia. 3 They are based on the concept of the block of the flow of the CSF around the posterior fossa due, for example, to Chiari malformations or arachnoiditis. Decompression of the posterior fossa may reduce the symptoms and signs as well as the size of the syrinx.34 We presume that in our case, the Chiari malformation was mild and that growth of the patient and enlargement of the posterior fossa may have led to restoration ofthe flow of CSF and the subsequent disappearance of the
syrinx. We conclude that the spontaneous reduction and the eventual disappearance of a syrinx can occur. Therefore, careful follow up and cautious therapeutic decisions are needed, especially in young patients. KAZUMASA SUDO SHIZUKI DOI YASUNORI MARUO KUNIO TASHIRO Department of Neurology SATOSHI TERAE KAZUO MIYASAKA Department of Radiology TOYOHIKO ISU Department of Neurosurgery, Hokkaido University School of Medicine, Sapporo, Japan
Correspondence to: Kazumasa Sudo, Department of Neurology, Hokkaido University School of Medicine, Kita-14 Nishi-5 Kita-Ku, Sapporo 060 Japan.
1 Barnett HJM, Foster JB, Hudgson P. Syringomyelia. In: Walton JN, ed. Major problems in neurology. London: WB Saunders, 1973:16-29. 2 Tashiro K, Fukazawa T, Moriwaka F, et al. Syringomyelic syndrome: clinical features in 31 cases confirmed by CT myelography or magnetic resonance imaging. J Neurol 1987; 235:26-30. 3 Tashiro K, Iwasaki Y, Fukazawa T, Isu T. Syringomyelia. Its classification, associated anomalies and treatment. Neurological Medicine (Tokyo) 1985;23:31-8. 4 Williams B. A critical appraisal of posterior fossa surgery for communicating syringomyelia. Brain 1978;101:223-50. 5 Levy WJ, Marson L, Hahn JF. Chiari malformation presenting in adults: A surgical experience in 127 cases. Neurosurgery 1983; 12:377-90. 6 Sherman JL, Barkovich AJ, Citrin CM. The MR appearance of syringomyelia: new observations. AJR 1987;148:381-91.
7 Gebarski SS, Maynard FW, Gabrielsen TO, Knake JE, Latack JF, Hoff PT. Posttraumatic progressive myelopathy. Radiology 1985; 157:379-85.
Syringomyelia associated with a spinal schwannoma: a case report Schwannomas are the most common intraspinal tumours. They constitute 16 to 30 per cent of all spinal tumours in reported series.' The presence of syringomyelia in association with a spinal schwannoma is rare being more often seen with intramedullary tumours. There are only two reports of the former in the literature.23 However, it is possible that with the increasing availability of MRI, the association will be recognised more frequently. We report a case of a schwannoma of the filum terminale associated with syringomyelia that was demonstrated on the preoperative MRI. A 38 year old female cashier had suffered mild back pain since her teens and for a year this had been associated with pain in the anterolateral and proximal aspects of her left leg. The pain was constant, burning in nature, worse at night and not related to exertion. In the same period walking became more difficult. She had no sphincteric disturbance. On examination, she had an obvious scoliosis convex to the left. There was no cutaneous stigmata of neurofibromatosis. Lumbar movements and straight leg raising were full. There was generalised wasting of the left leg and a spastic paraparesis with Grade 4 power, the weakness being a little more marked on the left. Knee and ankle jerks were brisk and plantars extensor. Joint position sense was slightly reduced in the legs, left more than right. Light touch, pinprick and temperature sensation were all reduced in the left leg below LI and sparing the perineal segments. Gait was spastic with a tendency to circumduct the left leg. Plain spinal radiographs showed an expanded lower dorsal and upper lumbar spinal canal. CT myelograph showed a mass adjacent to and displacing the conus at L1/2. MRI (fig a, b) showed a mass at the level of the conus, oval in shape and indistinguishable
Figure T, weighted MRI, coronal (a) sagittal (b), show a mass indistinguishable from the cord at the level of the conus. There is a syrinx cavity above the lesion.
from the cord in terms of signal intensity and plane of separation. In addition, there was slightly irregular cavitation extending proximally from the lesion into the upper
thoracic cord. The cervical cord was normal in shape and size with no evidence of syringomyelia. There were no abnormalities at the craniocervical junction, in particular, no Arnold Chiari malformation. The tonsils were in a normal position at the level of a line joining the inferior tip of the clivus and the posterio-inferior lip of the foramen magnum. The patient had a D 1 to L2 laminectomy. This revealed an intradural firm pinkish tumour sized 5 x 3 cm excavating the conus from the left. Above the lesion, the cord was expanded. A small mid-line myelotomy above the tumour revealed a syrinx containing clear fluid identical with cerebrospinal fluid (CSF). The tumour was totally excised with the help of a Cavitron Ultrasonic Surgical Aspirator and it appeared to arise from the filum terminale. After the tumour was excised, a small hole in the cord just above the conus was visable at the upper end of the tumour bed which appeared to communicate with the syrinx cavity. Histology confirmed a typical appearance of a schwannoma with no malignant changes. Post operatively, there was a temporary deterioration of the patient's paraparesis which has since improved. Spinal schwannomas almost invariably originate from the posterior roots of the spinal nerves4 and this may account for the high incidence of radicular pain as a presenting symptom. These tumours may be single or multiple and may or may not be associated with generalised neurofibromatosis. In the report of 1 15 cases of spinal schwannomas by Gautier-Smith,' only 12 cases had some cutaneous manifestations of neurofibromatosis. The presence of cord signs in association with lumbar schwannomas is uncommon since there must be spread upward of the tumour to produce this.' However, in our case the degree of spastic paraparesis was disproportionate to the extent of cord compression by the tumour which was effectively confined to the conus only. Presumably the ascending syrinx was responsible for the upper motor neuron signs. In 1969 Williams6 introduced the concept of communicating and non-communicating syringomyelia referring to any existing connection between the syrinx cavity and the fourth ventricle. The majority of patients with a communicating syrinx also have some form of hind brain abnormality such as Arnold Chiari malformation or Dandy Walker cyst with resulting obstruction of the normal outlets of the fourth ventricle.4 Non communicating syringomyelia is less common and relates to the extension of cystic cavities within the cord or to blockage of the central canal within the cord. The commonest causes are, intramedullary tumours, post traumatic paraplegia, and dense arachnoiditis. Syringomyelia occurring with intramedullary spinal tumours is a well known entity but extramedullary tumours causing syringomyelia are less common. In a review of the world's literature up to 1973, Barnett et a!' reported seven cases of pathologically proven syringomyelia associated with extramedullary tumours, five of which were thoracic and two were cervical in location. Few cases have been reported since and these have been in the main meningiomas.8 There were three cases of schwannomas. In attempting to explain the pathogenesis
Letters to the editor
439
of syringomyelia related to a spinal schwannoma, the possibility of a co-incidental relationship is remote. It is true that the MRI will continue to attract notice to more and more cases of double pathology, however, in this case a causal relationship is more likely. This is supported by the observation that the syrinx cavity communicated with the tumour bed and the lack of evidence of communication between the syrinx and the fourth ventricle, as well as the absence of any other responsible pathology such as another tumour. In our case there was no evidence that the syrinx cavity was an extension of the tumour or that the fluid was produced by the tumour. There was no cyst formation around the tumour, unlike the case reported by Gooding,' where the secondary cyst extended into the cord above the tumour and communicated with the syrinx. The presumed localised impairment of vascular supply ofthe cord related to the tumour could be responsible for the cavitation within the cord.9 However, we feel that the pathogenesis is best explained in a similar way to the proposition put by Quencer.' The longstanding presence of the schwannoma had caused a block in the spinal subarachnoid space obstructing the pulsatile CSF. In addition, the tumour itself must have caused local damage and atrophy to the spinal cord around the area it was excavating. After a period of time the altered CSF dyanmics caused a hole in the cord. Through this and possibly through the dilated perivascular spaces CSF entered the syrinx cavity. Fluid movement inside the cavity leads to expansion. We conclude that the syringomyelia in association with intradural extramedullary tumours is now a well recognised condition and must be considered on each occasion where the neurological deficits could not be explained by the position ofthe tumour. With the greater availability of MRI this pathology may be more frequently identified. We are grateful to Mr R D Weeks for his permission to report this case. A B JAMJOOM K G DAVIES Department of Neurosurgery, University Hospital of Wales, Heath Park, Cardiff CF4 4XW
Correspondence to: Mr A Jamjoom, 23 Windmill Lane, Henbury, Bristol BS1O 7XE
1 Yeomans JR. Neurological Surgery Vol 5. Philadelphia: Saunders, 1982:3204. 2 Gooding MR. Syringomyelia in association with a neurofibroma of the filum terminale. J
Neurol, Neurosurg Psychiatry 1972;35:560-4.
3 Quencer RM, El Gammal T, Cohen G. Syringomyelia associated with intradural extramedullary masses of the spinal canal. Am J Neuroradiol 1986;7(1):143-8. 4 Miller JD, ed. Northfield's Surgery of the Central Nervous System, 2nd ed. Blackwell: Oxford 1987, 586. 5 Gautier-Smith PC. Clinical Aspects of spinal neurofibromas. Brain 1967;90:359-94. 6 Williams B. The distending force in the production of communicating syringomyelia. Lancet 1969;ii:189-93. 7 Barnett HJM, Foster JB, Hudgson P. Syringomyelia, Vol 1. Major Problems in Neurology.
London: Saunders, 1973.
8 Wesserberg J, Marks P, Hardy D.
Syrin-
gomyelia of the thoracic spinal cord associated with spinal meningiomas. Br J Neurosurg
1987;1:485-8. 9 Hall PV, Muller J, Campbell RL. Experimnental
hydrosyringomyelia, ischaemia myelopathy and syringomyelia. J Neurosurg 1975;43: 464-70.
Cervical cord compression and femoral neuropathy in hereditary multiple exostoses Hereditary multiple exostoses is an inherited dysplasia of bone, characterised by the presence of multiple cartilage capped exostoses. The exostoses arise mainly from the metaphyses of long bones, but any bone formed by endochondral ossification may be affected. Patients with this condition are usually below average height, and have typical deformities such as ulnar deviation of the wrist, subluxation of the radio-humeral joint, limb length discrepancy, and valgus of the knees and ankles.' The exostoses grow slowly until skeletal maturity, and behave as space occupying lesions. The most frequent local complications are bursa formation, interference with tendon function and restriction of joint motion. Neurological complications are uncommon, but when they occur most commonly take the form of a peripheral compression neuropathy. Substantiated cervical cord compression in hereditary multiple exostoses is rare. Thirteen cases have been reported from 1843 to the present,2 3one of which had in addition confirmed involvement of the peripheral nervous system due to exostoses at separate sites.4 We report a case presenting with a peripheral compression neuropathy ofthe left femoral nerve caused by a large iliac exostosis, which masked the onset of a spastic paraparesis from an exostosis arising at C2. A 15 year old boy with a family history of hereditary multiple exostoses, presented with painful knees and difficulty in walking. Examination revealed multiple large exostoses, involving the shoulders, elbows, and knees and small exostoses at the wrist and ankles. The standing intermalleolar separation was 18 cm. Knee flexion and extension were restricted by large exostoses arising from the distal femoral and proximal tibial metaphyses. A large left sided abdominal mass was present, arising from the left innominate bone restricting flexion of the left hip to 80'. His left quadriceps were weak (MRC grade four), and the left knee jerk was absent. His gait was clumsy. Radiographs of the pelvis showed a large exostosis arising from the left iliac crest extending from the anterior superior iliac spine to the ala of the sacrum with an upper border at the level of the second lumbar vertebra. Excision was performed and histopathological examination confirmed a benign exostosis. Postoperatively the patient had a complete femoral nerve palsy, which recovered rapidly to MRC grade three. Electromyography confirmed a lower motor neuron lesion. Four months later he was readmitted for correction of his genu valgum. Examination on this occasion revealed a spastic paraparesis, with weakness in both legs. The tendon reflexes were increased, apart from the left knee jerk, which was still absent. The plantar responses were extensor. Unsustained clonus was present at both ankles. Myelography revealed a large extradural mass at the C2 level, compressing the dural sac. Computed tomography confirmed the presence of a large exostosis arising from the neural arch of C2 (fig). At decompressive laminectomy the exostosis was found to arise from the neural arch of C2 extending upwards and anteriorly to the arch of Cl. Histopathological examination confirmed a
Figure CT scan of the cervical spine with metrizimide showing the origin of the exostosis from the neural arch of C2 with marked cord compression. benign exostosis. Postoperatively the spastic paraparesis rapidly improved. Hereditary multiple exostoses is unusual amongst skeletal dysplasias, in that an average life span and normal function may be anticipated.' The incidence ofvertebral exostoses is between 1-7%.' They commonly arise from the neural arches but spinal cord compression is rare as they are usually small and stop growing at puberty. The majority of cases of confirmed cervical cord compression by an exostosis in association with hereditary multiple exostoses occur in males, in the second or third decade of life.' In the fatal case of cervical cord compression with confirmed associated peripheral femoral and sciatic neuropathy described by Chiurco the patient presented with tetraparesis.4 The diagnosis of cord compression was less apparent in our case because the lower limb deformities and peripheral femoral neuropathy, in the absence of other neurological signs provided an explanation for his gait problems. The need for vigilance when managing these patients is stressed as prompt cord decompression should lead to full recovery. D A NUNEZ, H K GRAHAM, J A FIXSEN, The Orthopaedic Department, The Hospitalfor Sick Children, Great Ormond Street, London
Correspondence to: Mr D A Nunez, North Riding Infirmary, Middlesbrough, Cleveland TS1 5JE.
1 Solomon L. Hereditary multiple exostoses. J Bone Joint Surg 1963;45-B:292-304. 2 Misra UK, Nag D, Dave VS, Shukla R, Kar AM. Cervical cord compression due to chondromatous change in a patient with metaphysial aclasis. J Neurol Neurosurg Psychiatry 1983;46:1 155-7. 3 Gottlieb A, Severi P, Ruelle A, Lasio G. Exostosis as a Cause of Spinal Cord Compression. Surg Neurol 1986;26:581-4. 4 Chiurco AA. Multiple exostoses of bone with fatal spinal cord compression: report of a case and brief review of the literature. Neurology
(Minneap) 1970;20:275-8.
5 Stocks P, Barrington A. Hereditary disorders of bone development. In: Pearson D, ed. The treasury of human inheritance. London: Cambridge University Press, 1925.
438
spinal cord; also the cord was not atrophic. The images of the brain were also judged to be normal. On a T2-weighted image there was an abnormal high intensity area in the centre of the spinal cord where the syrinx had been. There was no CSF flow-void sign or abnormality of cord size. On MRI, after a follow up period of two years and three months, the Chiari malformation and syrinx had disappeared on the T1weighted image. On the T2-weighted image, however, a high intensity area was clearly visible in the spinal cord. A high intensity area on T2-weighted image showed a nonspecific change and could be due to gliosis, oedema, microcystic changes and/or demyelination.67 We therefore assumed that the syrinx had collapsed and been replaced by
gliosis. The reduction of pain, hand muscle weakness, pathological reflexes and severity of anaesthesia are likely to reflect the disappearance of the syrinx, even though the hyporeflexia of the left arm remained unchanged and the area of sensory disturbance was somewhat enlarged. There are several theories about the pathogenesis of syringomyelia. 3 They are based on the concept of the block of the flow of the CSF around the posterior fossa due, for example, to Chiari malformations or arachnoiditis. Decompression of the posterior fossa may reduce the symptoms and signs as well as the size of the syrinx.34 We presume that in our case, the Chiari malformation was mild and that growth of the patient and enlargement of the posterior fossa may have led to restoration ofthe flow of CSF and the subsequent disappearance of the
syrinx. We conclude that the spontaneous reduction and the eventual disappearance of a syrinx can occur. Therefore, careful follow up and cautious therapeutic decisions are needed, especially in young patients. KAZUMASA SUDO SHIZUKI DOI YASUNORI MARUO KUNIO TASHIRO Department of Neurology SATOSHI TERAE KAZUO MIYASAKA Department of Radiology TOYOHIKO ISU Department of Neurosurgery, Hokkaido University School of Medicine, Sapporo, Japan
Correspondence to: Kazumasa Sudo, Department of Neurology, Hokkaido University School of Medicine, Kita-14 Nishi-5 Kita-Ku, Sapporo 060 Japan.
1 Barnett HJM, Foster JB, Hudgson P. Syringomyelia. In: Walton JN, ed. Major problems in neurology. London: WB Saunders, 1973:16-29. 2 Tashiro K, Fukazawa T, Moriwaka F, et al. Syringomyelic syndrome: clinical features in 31 cases confirmed by CT myelography or magnetic resonance imaging. J Neurol 1987; 235:26-30. 3 Tashiro K, Iwasaki Y, Fukazawa T, Isu T. Syringomyelia. Its classification, associated anomalies and treatment. Neurological Medicine (Tokyo) 1985;23:31-8. 4 Williams B. A critical appraisal of posterior fossa surgery for communicating syringomyelia. Brain 1978;101:223-50. 5 Levy WJ, Marson L, Hahn JF. Chiari malformation presenting in adults: A surgical experience in 127 cases. Neurosurgery 1983; 12:377-90. 6 Sherman JL, Barkovich AJ, Citrin CM. The MR appearance of syringomyelia: new observations. AJR 1987;148:381-91.
7 Gebarski SS, Maynard FW, Gabrielsen TO, Knake JE, Latack JF, Hoff PT. Posttraumatic progressive myelopathy. Radiology 1985; 157:379-85.
Syringomyelia associated with a spinal schwannoma: a case report Schwannomas are the most common intraspinal tumours. They constitute 16 to 30 per cent of all spinal tumours in reported series.' The presence of syringomyelia in association with a spinal schwannoma is rare being more often seen with intramedullary tumours. There are only two reports of the former in the literature.23 However, it is possible that with the increasing availability of MRI, the association will be recognised more frequently. We report a case of a schwannoma of the filum terminale associated with syringomyelia that was demonstrated on the preoperative MRI. A 38 year old female cashier had suffered mild back pain since her teens and for a year this had been associated with pain in the anterolateral and proximal aspects of her left leg. The pain was constant, burning in nature, worse at night and not related to exertion. In the same period walking became more difficult. She had no sphincteric disturbance. On examination, she had an obvious scoliosis convex to the left. There was no cutaneous stigmata of neurofibromatosis. Lumbar movements and straight leg raising were full. There was generalised wasting of the left leg and a spastic paraparesis with Grade 4 power, the weakness being a little more marked on the left. Knee and ankle jerks were brisk and plantars extensor. Joint position sense was slightly reduced in the legs, left more than right. Light touch, pinprick and temperature sensation were all reduced in the left leg below LI and sparing the perineal segments. Gait was spastic with a tendency to circumduct the left leg. Plain spinal radiographs showed an expanded lower dorsal and upper lumbar spinal canal. CT myelograph showed a mass adjacent to and displacing the conus at L1/2. MRI (fig a, b) showed a mass at the level of the conus, oval in shape and indistinguishable
Figure T, weighted MRI, coronal (a) sagittal (b), show a mass indistinguishable from the cord at the level of the conus. There is a syrinx cavity above the lesion.
from the cord in terms of signal intensity and plane of separation. In addition, there was slightly irregular cavitation extending proximally from the lesion into the upper
thoracic cord. The cervical cord was normal in shape and size with no evidence of syringomyelia. There were no abnormalities at the craniocervical junction, in particular, no Arnold Chiari malformation. The tonsils were in a normal position at the level of a line joining the inferior tip of the clivus and the posterio-inferior lip of the foramen magnum. The patient had a D 1 to L2 laminectomy. This revealed an intradural firm pinkish tumour sized 5 x 3 cm excavating the conus from the left. Above the lesion, the cord was expanded. A small mid-line myelotomy above the tumour revealed a syrinx containing clear fluid identical with cerebrospinal fluid (CSF). The tumour was totally excised with the help of a Cavitron Ultrasonic Surgical Aspirator and it appeared to arise from the filum terminale. After the tumour was excised, a small hole in the cord just above the conus was visable at the upper end of the tumour bed which appeared to communicate with the syrinx cavity. Histology confirmed a typical appearance of a schwannoma with no malignant changes. Post operatively, there was a temporary deterioration of the patient's paraparesis which has since improved. Spinal schwannomas almost invariably originate from the posterior roots of the spinal nerves4 and this may account for the high incidence of radicular pain as a presenting symptom. These tumours may be single or multiple and may or may not be associated with generalised neurofibromatosis. In the report of 1 15 cases of spinal schwannomas by Gautier-Smith,' only 12 cases had some cutaneous manifestations of neurofibromatosis. The presence of cord signs in association with lumbar schwannomas is uncommon since there must be spread upward of the tumour to produce this.' However, in our case the degree of spastic paraparesis was disproportionate to the extent of cord compression by the tumour which was effectively confined to the conus only. Presumably the ascending syrinx was responsible for the upper motor neuron signs. In 1969 Williams6 introduced the concept of communicating and non-communicating syringomyelia referring to any existing connection between the syrinx cavity and the fourth ventricle. The majority of patients with a communicating syrinx also have some form of hind brain abnormality such as Arnold Chiari malformation or Dandy Walker cyst with resulting obstruction of the normal outlets of the fourth ventricle.4 Non communicating syringomyelia is less common and relates to the extension of cystic cavities within the cord or to blockage of the central canal within the cord. The commonest causes are, intramedullary tumours, post traumatic paraplegia, and dense arachnoiditis. Syringomyelia occurring with intramedullary spinal tumours is a well known entity but extramedullary tumours causing syringomyelia are less common. In a review of the world's literature up to 1973, Barnett et a!' reported seven cases of pathologically proven syringomyelia associated with extramedullary tumours, five of which were thoracic and two were cervical in location. Few cases have been reported since and these have been in the main meningiomas.8 There were three cases of schwannomas. In attempting to explain the pathogenesis
Letters to the editor
439
of syringomyelia related to a spinal schwannoma, the possibility of a co-incidental relationship is remote. It is true that the MRI will continue to attract notice to more and more cases of double pathology, however, in this case a causal relationship is more likely. This is supported by the observation that the syrinx cavity communicated with the tumour bed and the lack of evidence of communication between the syrinx and the fourth ventricle, as well as the absence of any other responsible pathology such as another tumour. In our case there was no evidence that the syrinx cavity was an extension of the tumour or that the fluid was produced by the tumour. There was no cyst formation around the tumour, unlike the case reported by Gooding,' where the secondary cyst extended into the cord above the tumour and communicated with the syrinx. The presumed localised impairment of vascular supply ofthe cord related to the tumour could be responsible for the cavitation within the cord.9 However, we feel that the pathogenesis is best explained in a similar way to the proposition put by Quencer.' The longstanding presence of the schwannoma had caused a block in the spinal subarachnoid space obstructing the pulsatile CSF. In addition, the tumour itself must have caused local damage and atrophy to the spinal cord around the area it was excavating. After a period of time the altered CSF dyanmics caused a hole in the cord. Through this and possibly through the dilated perivascular spaces CSF entered the syrinx cavity. Fluid movement inside the cavity leads to expansion. We conclude that the syringomyelia in association with intradural extramedullary tumours is now a well recognised condition and must be considered on each occasion where the neurological deficits could not be explained by the position ofthe tumour. With the greater availability of MRI this pathology may be more frequently identified. We are grateful to Mr R D Weeks for his permission to report this case. A B JAMJOOM K G DAVIES Department of Neurosurgery, University Hospital of Wales, Heath Park, Cardiff CF4 4XW
Correspondence to: Mr A Jamjoom, 23 Windmill Lane, Henbury, Bristol BS1O 7XE
1 Yeomans JR. Neurological Surgery Vol 5. Philadelphia: Saunders, 1982:3204. 2 Gooding MR. Syringomyelia in association with a neurofibroma of the filum terminale. J
Neurol, Neurosurg Psychiatry 1972;35:560-4.
3 Quencer RM, El Gammal T, Cohen G. Syringomyelia associated with intradural extramedullary masses of the spinal canal. Am J Neuroradiol 1986;7(1):143-8. 4 Miller JD, ed. Northfield's Surgery of the Central Nervous System, 2nd ed. Blackwell: Oxford 1987, 586. 5 Gautier-Smith PC. Clinical Aspects of spinal neurofibromas. Brain 1967;90:359-94. 6 Williams B. The distending force in the production of communicating syringomyelia. Lancet 1969;ii:189-93. 7 Barnett HJM, Foster JB, Hudgson P. Syringomyelia, Vol 1. Major Problems in Neurology.
London: Saunders, 1973.
8 Wesserberg J, Marks P, Hardy D.
Syrin-
gomyelia of the thoracic spinal cord associated with spinal meningiomas. Br J Neurosurg
1987;1:485-8. 9 Hall PV, Muller J, Campbell RL. Experimnental
hydrosyringomyelia, ischaemia myelopathy and syringomyelia. J Neurosurg 1975;43: 464-70.
Cervical cord compression and femoral neuropathy in hereditary multiple exostoses Hereditary multiple exostoses is an inherited dysplasia of bone, characterised by the presence of multiple cartilage capped exostoses. The exostoses arise mainly from the metaphyses of long bones, but any bone formed by endochondral ossification may be affected. Patients with this condition are usually below average height, and have typical deformities such as ulnar deviation of the wrist, subluxation of the radio-humeral joint, limb length discrepancy, and valgus of the knees and ankles.' The exostoses grow slowly until skeletal maturity, and behave as space occupying lesions. The most frequent local complications are bursa formation, interference with tendon function and restriction of joint motion. Neurological complications are uncommon, but when they occur most commonly take the form of a peripheral compression neuropathy. Substantiated cervical cord compression in hereditary multiple exostoses is rare. Thirteen cases have been reported from 1843 to the present,2 3one of which had in addition confirmed involvement of the peripheral nervous system due to exostoses at separate sites.4 We report a case presenting with a peripheral compression neuropathy ofthe left femoral nerve caused by a large iliac exostosis, which masked the onset of a spastic paraparesis from an exostosis arising at C2. A 15 year old boy with a family history of hereditary multiple exostoses, presented with painful knees and difficulty in walking. Examination revealed multiple large exostoses, involving the shoulders, elbows, and knees and small exostoses at the wrist and ankles. The standing intermalleolar separation was 18 cm. Knee flexion and extension were restricted by large exostoses arising from the distal femoral and proximal tibial metaphyses. A large left sided abdominal mass was present, arising from the left innominate bone restricting flexion of the left hip to 80'. His left quadriceps were weak (MRC grade four), and the left knee jerk was absent. His gait was clumsy. Radiographs of the pelvis showed a large exostosis arising from the left iliac crest extending from the anterior superior iliac spine to the ala of the sacrum with an upper border at the level of the second lumbar vertebra. Excision was performed and histopathological examination confirmed a benign exostosis. Postoperatively the patient had a complete femoral nerve palsy, which recovered rapidly to MRC grade three. Electromyography confirmed a lower motor neuron lesion. Four months later he was readmitted for correction of his genu valgum. Examination on this occasion revealed a spastic paraparesis, with weakness in both legs. The tendon reflexes were increased, apart from the left knee jerk, which was still absent. The plantar responses were extensor. Unsustained clonus was present at both ankles. Myelography revealed a large extradural mass at the C2 level, compressing the dural sac. Computed tomography confirmed the presence of a large exostosis arising from the neural arch of C2 (fig). At decompressive laminectomy the exostosis was found to arise from the neural arch of C2 extending upwards and anteriorly to the arch of Cl. Histopathological examination confirmed a
Figure CT scan of the cervical spine with metrizimide showing the origin of the exostosis from the neural arch of C2 with marked cord compression. benign exostosis. Postoperatively the spastic paraparesis rapidly improved. Hereditary multiple exostoses is unusual amongst skeletal dysplasias, in that an average life span and normal function may be anticipated.' The incidence ofvertebral exostoses is between 1-7%.' They commonly arise from the neural arches but spinal cord compression is rare as they are usually small and stop growing at puberty. The majority of cases of confirmed cervical cord compression by an exostosis in association with hereditary multiple exostoses occur in males, in the second or third decade of life.' In the fatal case of cervical cord compression with confirmed associated peripheral femoral and sciatic neuropathy described by Chiurco the patient presented with tetraparesis.4 The diagnosis of cord compression was less apparent in our case because the lower limb deformities and peripheral femoral neuropathy, in the absence of other neurological signs provided an explanation for his gait problems. The need for vigilance when managing these patients is stressed as prompt cord decompression should lead to full recovery. D A NUNEZ, H K GRAHAM, J A FIXSEN, The Orthopaedic Department, The Hospitalfor Sick Children, Great Ormond Street, London
Correspondence to: Mr D A Nunez, North Riding Infirmary, Middlesbrough, Cleveland TS1 5JE.
1 Solomon L. Hereditary multiple exostoses. J Bone Joint Surg 1963;45-B:292-304. 2 Misra UK, Nag D, Dave VS, Shukla R, Kar AM. Cervical cord compression due to chondromatous change in a patient with metaphysial aclasis. J Neurol Neurosurg Psychiatry 1983;46:1 155-7. 3 Gottlieb A, Severi P, Ruelle A, Lasio G. Exostosis as a Cause of Spinal Cord Compression. Surg Neurol 1986;26:581-4. 4 Chiurco AA. Multiple exostoses of bone with fatal spinal cord compression: report of a case and brief review of the literature. Neurology
(Minneap) 1970;20:275-8.
5 Stocks P, Barrington A. Hereditary disorders of bone development. In: Pearson D, ed. The treasury of human inheritance. London: Cambridge University Press, 1925.
