148 Letters haustable ankle clonus and extensor plantar lesions or compression of the spinal cord MICHEL D FERRARI reflex, all on the right side. After a 15 most unlikely. Based on the results of CSF PIETER HE HILKENS minutes rest, upper-motor neuron signs and studies, multimodality evoked potentials BERRY KREMER symptoms had disappeared. Routine labor- tests and MRI, in both cases a diagnosis of THEO W POLDER atory tests including syphilis serology and "laboratory-supported-definite multiple Department of Neurology, complement studies were normal or nega- sclerosis" was made, according to the criteUniversity Hospital Leiden, tive. ANA and rheumatoid factor were ria of Poser et al.7 In the literature on neuPO Box 9600 negative. The CSF showed high IgG and rogenic intermittent claudication, multiple 2300 RC Leiden, one extra oligoclonal band at iso-electric sclerosis is not generally cited as a cause for The Netherlands focusing. Myelography, spinal angiography IPC."2On the other hand, textbooks on References and cranial CT scan were normal. MRI multiple sclerosis mention IPC as "a freshowed multiple hemispheral white matter quent early symptom of multiple sclerosis", I Verbiest H. Neurogenic intermittent claudication. In: Vinken PJ, Bruyn GW. Handbook lesions, suggesting multifocal demyelin- however, without referring to well documenof Clinical Neurology. Vol 20. Amsterdam: ation. Visual-evoked-responses (VER) and ted observations.45 Descriptions of signs North-Holland Publishing Company, 1976: somatosensory-evoked-responses (SSER) and symptoms are variable and sometimes 611-724. showed increased latencies. vague.45 9 The period during which upper- 2 Jellinger K, Neumayer E. Claudication of the Patient 2: Six years before admission, this motor neuron signs and symptoms really are spinal cord and cauda equina. In: Vinken PJ, 58 year old book-keeper repeatedly noticed intermittent (that is, they completely disapBruyn GW. Handbook of Clinical Neurology. brief jerks of his right leg after walking for pear after rest) appears to be short.49 Only Vol 12. Amsterdam: North-Holland 10 minutes, which disappeared after resting three short case histories could be found in Publishing Company. 1972:507-47. for 5 minutes. One and a half years later he the literature.6 '1 McAlpine6 described two 3 Dejerine J. Sur la claudication intermittente de la moelle epiniere. Rev, Neurol (Paris) was troubled by a heavy feeling in the right patients who, as early feature of multiple 1906;14:341-50. arm with numbness of the first three fingers, sclerosis, showed IPC for several weeks. In 4 Matthews WB, Acheson ED, Batchelor JR, on using it extensively. These symptoms dis- both cases this disappeared for about a Weller RO. McAlpine s Multiple Sclerosis. appeared after 30 minutes rest. Four years year and then recurred as a prelude to 1985. Edinburgh: Churchill after onset he noticed weakness of the legs a progressive course of the disease. 5 Kelly R. Clinical aspectsLivingstone, of multiple sclerosis. on walking for 5 minutes, which disap- Godlewski'째 mentioned a multiple sclerosis In: Vinken PJ, Bruyn GW. Handbook of Clinical Neurology. Vol 47. Amsterdam: peared during rest. These symptoms forced patient who had IPC for 4 months, before Elsevier Science Publishers 1985:49-78. him to work for half days only. Six months permanent spastic paraplegia developed. later, walking felt stiff and awkward with Multiple sclerosis presenting with slowly 6 McAlpine D, Compston ND, Lumsden CE. Multiple Sclerosis. Edinburgh: Livingstone, pain and tingling of the legs and an imper- progressive IPC as isolated symptom for 1955. ative micturation developed. When lying in nearly 6 years has, to the best of our 7 Poser CM, Paty DW, Scheinberg, el al. New bed he noticed jerking movements knowledge, not been described before. The criteria for Multiple Sclerosis: diagnostic of the legs. Six years after onset of the pathophysiological mechanism of IPC in Guidelines for research protocols. Ann symptoms he was admitted to our hospital. multiple sclerosis is believed to be related to Neurol 1983;13:227-31. Neurological examination revealed no diminished motor conduction in demy- 8 Claude H, Egger W. Quelques symptomes nouveaux de la sclerose en plaques. Rev abnormalities except for absent abdominal elinated axons, resulting from even a slight Neurol (Paris) 1906;14:275-7. reflexes. His gait was normal initially, but rise of body-temperature caused by exerbecame spastic after walking for 1 hour and cise.4 51 A temperature increase of only 9 Guillain G. La sclerole en plaques. Rev Neurol (Paris) 1 924;24:648-83. a brisk left biceps jerk, exaggerated ankle 0 5째C appears to be sufficient to produce 10 Godlewski S. Les claudications intermittentes jerks and bilateral extensor plantar reflexes reversible conduction-block in demyelinated medullaires et radiculaires. L'HWpital (Horsappeared. After a 10 minute rest upper- fibres."2 Indeed in our first patient, walking serie) 1969:529-48. motor neuron signs and symptoms for 30 minutes increased tympanic tem- 11 Davis FA. Axonal conduction studies based on some considerations of temperature effects disappeared. Routine laboratory tests were perature by 0 5째C. Obviously, comparison in multiple sclerosis. Electroencephalogr Clin normal. CSF examination revealed high IgG of pre- and post-exertion corticospinal tract Neurophysiol 1970;28:281-6. and multiple extra oligoclonal bands at conduction time"3 in patients with IPC, may iso-electric focusing. Myelography was give better insight in the pathophysiological 12 Rasminsky M. The effects of temperature on conduction in demyelinated single nerve normal. MRI revealed multiple para- mechanism. fibres. Arch Neurol 1973;28:287-92. ventricular white matter lesions consistent IPC is generally attributed to transient 13 Mills KR, Murray NMF. Corticospinal tract with multifocal demyelination. VER showed spinal ischaemia.' However, actual evidence conduction time in multiple sclerosis. Ann prolongation of P100 on stimulation of the for such causal relationship is sometimes Neurol 1985;18:601-5. right eye, SSER showed absent cortical poor.3 14 Conduction or angiographic evi- 14 Dupre ME. Claudication intermittente par arteriosclerose de la moelle. Rev Neurol responses on bilateral tibial nerve stimu- dence are rarely found. In retrospect, some (Paris) 1914;28:498-500. lation and electro-oculography showed of these cases might in fact have been caused square wave jerks on eye closure. by multiple sclerosis. 27 The main features of both case histories In conclusion, these two case histories Accepted July 1987 are intermittent, exertion-induced, upper- illustrate that IPC may be a solitary manimotor neuron symptoms and signs, disap- festation of multiple sclerosis for protracted Unexplained chronic subarachnoid bleeding pearing after rest. During rest or mild periods. A simple exertion-provocation-test and a slowly progressive neurological exercise both patients were subjectively and may be essential to reveal the neurological syndrome objectively asymptomatic. Normal "vascu- substrate of the complaints and to establish Sir: Superficial haemosiderosis of the central litis tests", angiography, myelography and the diagnosis. nervous system (CNS) (subpial cerebral MRI of the spinal cord made vascular 149 Letters siderosis) is the result of chronic and repeated leakage of blood into the subarachnoid space. The syndrome is unique in that the symptoms are mainly caused by the deposition of iron-containing pigments (haemosiderin) in the marginal zones of the CNS1 resulting in gliosis and neuronal loss. 6 years superficial haemosiderosis of the CNS was suspected. Neuropsychological examination showed a discrepancy between the verbal IQ (114) and the performance IQ (77) on the WISC-R scale. An audiological examination was normal. At the age of 8 years she was readmitted after two generThe cardinal feature is the finding of xan- alised convulsions and an increase of her thochromic cerebrospinal fluid (CSF) on cerebellar signs. A new finding at neurorepeated lumbar punctures. The clinical pic- logical examination were absent ankle jerks. ture is quite varying, but one or more of the The CSF was still xanthochromic. following signs are usually present: cere- Myelography and MRI of the brain and spibellar ataxia, nerve deafness, dementia and nal cord were normal. Neuropsychological spasticity.2 These signs do not indicate the examination showed no deterioration in site and cause of the bleeding. Two patients comparison to the examination of 2 years (adults) have been reported who were diag- before. She was discharged and 3 years later nosed during life;34 we report one young her clinical state is unaltered. In this 9 year old child CSF xanthopatient with a remarkable benign course chromia has been present for 5 years with over 4 years. A female child was admitted to hospital at elevated protein, methaemoglobin and the age of 5 years because of abdominal bilirubin, indicating continuous blood leakpain, vomiting and failure to thrive. After age in the subarachnoid space (table). Since thorough investigations no cause was found other causes of the neurological signs have for the abdominal complaints. An EEG been excluded, the syndrome is very likely to unexpectedly showed generalised bilateral be caused by superficial haemosiderosis of synchronous spike and waves. A lumbar the CNS. Apart from an initial period after puncture was then performed which showed the first lumbar puncture, there have been xanthochromic CSF with a raised protein no signs of meningism, so the bleeding has content of 2 g/l (see table). One day later she been chronic and slow. The clinical picture consisted of bilateral complained of headache, consciousness became impaired, and signs of meningism cerebellar ataxia, seizures, a low performance IQ and absent ankle jerks. There developed. A CT scan was normal. On transfer to our hospital, examination has been no evidence of spasticity or nerve revealed a slightly stuporous child with signs deafness. From the patients described by of meningism, without lateralising neuro- Tomlinson and Walton2 it can be concluded logical signs and with normal fundi. In the that none of the clinical features is a sine qua course of the next days she recovered spon- non for superficial haemosiderosis of the taneously. Repeated EEG showed bilateral CNS. Absent ankle or knee jerks have been synchronous epileptic discharges of short described1 67 as likely caused by the sideduration. Four-vessel angiography revealed rosis as by a tumour of the cauda equina no abnormaliti'es. Repeated lumbar punc- with repeated bleeding. Epilepsy was tures disclosed persistent xanthochromia, reported in one patient.2 The dementia, if elevated protein, methaemoglobin and present, occurs in the terminal stage of the bilirubin (see table). A month after admis- disease as a result of necrosis of the cortex sion slight cerebellar ataxia became evident. due to the haemosiderosis; in addition At discharge 2 months later the CSF abnor- hydrocephalus may be a causative factor. malities were still present. Six months later Only one author8 has reported on necropshe developed absences. At the age of sies in two children with this syndrome: a 12 Table CSFfindings in the reported patient Total protein (mg/dl) Bilirubin (mg/dl) October 1981 November 1981 December 1981 May 1 67 07 2 15 09 16-2 11 red - 1-96 09 14-3 2-2 0-6 15 8 8 white Monocytes macrophages some with haemosiderin Methaemoglobin (mg/dl) 11-8 6 white Cytology Monocytes Cells/mm3 macrophages - - 1984 Normal values: protein 0 15-045 mg/dl; bilirubin not present; methaemoglobine not present. June 1984 1-7 0-8 157 7 white - years old boy and a 7 month old baby. In our patient the disease has lasted 5 years with only a minimal progression of the symptoms. Despite an extensive search no cause or place of the bleeding was found. This is not too much of a surprise since even at necropsy in approximately half of the patients no cause of the bleeding has been found.8 In the remaining cases a tumour in contact with the CSF was usually found and sometimes a vascular malformation. The presence of methaemoglobin in the CSF is remarkable and not reported before in superficial haemosiderosis. It is usually formed in longstanding collections of encapsulated blood, due to the lack of oxygen.5 It could be argued that a relative lack of oxygen arises in the condition because of the chronic inflammation of the meninges resulting in the formation of methaemoglobin. In our patient haemosiderin deposition in macrophages in the CSF was seen supporting the diagnosis of superficial haemosiderosis. We suggest that this clinical syndrome is underreported as it is only mentioned in two major textbooks9 SO and in the literature of the last 13 years we found only one publication4 paying attention to this syndrome. M] ZWARTS JH BEGEER R LE COULTRE Department of Child Neurology, University Hospital, Oostersingel 59, 9713 EZ Groningen, The Netherlands. References I Dastur DK, Sinh G. Toxic iron and the nervous system. Acda Neuropathol (Berl) 1962;2: 161-20 1. 2 Tomlinson BE, Walton JN. Superficial haemosiderosis of the central nervous system. J Neurol Neurosurg Psychiatr) 1964;27:332-9. 3 Sherwin I, Toll K. Superficial hemosiderosis of the central nervous system. Dis Nerv Svst 1972;33:41 3-7. 4 Pinkston JW, Ballinger WE, Lotz PR, Friedman WA. Superficial Siderosis: A cause of leptomeningeal enhancement on computed tomography. J Comput Assist Tomogr 1983;7: 1073-6. 5 Fishman RA. Cerebrospinal Fluid in Disease of the Nervous SYstem. Philadelphia: Saunders, 1980:173-5. 6 McGee DA, Van Patter HJ, Morotta J, Olszewsky J. Subpial cerebral siderosis. Neurology 1962:12:108-13. 150 LetUers 7 Noetzel H. Diffusion von Blutfarb in der inneren Randzone und aussere Oberflache des Zentralnervensystems bei subarachnoidaler Blutung. Arch Psychiat Nervenkr 1940;111:129-38. 8 Trevor Hughes J, Oppenheimer DR. Superficial siderosis of the central nervous system. Acta Neuropathol (Berl) 1 969;13:56-74. 9 Bradley WG. Diseases of the spinal roots. In: Dyck PJ, Thomas PK, Lambert EH, Bunge R, eds. Peripheral Neuropathy. Philadelphia: Saunders, 1984:1377. 10 Adams RD, Victor M. Principles of Neurology. New York: McGraw-Hill 1985:472. Accepted 4 August 1987 Complete gaze palsy in pontine haemorrhage Sir: Henn et at1 have recently shown that in monkeys lesions confined to the reticular formation of the pons (paramedian pontine reticular formation, PPRF) cause abnormalities of vertical as well as horizontal gaze. In man it has long been recognised that pontine lesions cause horizontal gaze defects but vertical palsies have rarely been recorded. We report a patient with a pontine haemorrhage in whom there was both horizontal and vertical gaze palsy. A 55 year old hypertensive male was admitted to hospital with sudden onset of severe walking difficulty and left sided paraesthesia. On examination his blood pressure was 230/140 mmHg. He was co-operative and alert. Visual acuity and clinical visual field testing were normal. His fundi showed Grade II retinopathy. There was bilateral symmetrical miosis, but the pupils reacted to light and the eyelids were normally posi- tioned. The patient's eyes were in the primary position and he could not move them on command or in response to stationary or moving visual targets. Doll's head manoeuvre in the horizontal plane, (vestibulo-ocular reflex), as well as caloric irrigation with water at about 20째C failed to induce any eye movement. There was a right sided lower motor neuron facial weakness. The patient was unable to walk because of severe gait ataxia but sensorimotor examination of the limbs while lying down was normal. Tendon reflexes and plantar responses were also normal. Investigations included normal CSF examination. A CT scan (fig a) showed a pontine haemorrhage occupying the medial tegmental area, slightly more on the right, extending up to the ponto-mesencephalic junction. Mesencephalic and mesencephalon-diencephalic junction sections (fig b) were normal. The patient's hypertension was controlled with propranolol (160 mg a day) and nifedipine (40 mg a day). On the third day the patient noticed diplopia and oscillopsia and examination showed he had recovered a few degrees of conjugate vertical eye movements and abduction of the left eye, accompanied by left beating nystagmus on attempted gaze to that side. After a week the patient's gait improved and he was able to walk with help although he suffered from vertigo on standing. During the second week vertical gaze and convergence were restored. Abduction in the left eye was partially recovered. There was a conjugate, torsional nystagmus to the left (clockwise) and a superimposed horizontal pendular nystagmus in the left eye. By the eighth week the haematoma had resolved on CT scanning. When examined 2 years later the patient had mild cerebellar signs in the left limbs and an unsteady, broad based gait with left lateropulsion. He complained of positional vertigo and horizontal oscillopsia which was worse when viewing with the left eye alone. There was a horizontal pendular nystagmus with an elliptical trajectory more marked on the left eye, a constant clockwise torsional nystagmus in both eyes, a severe right VI nerve palsy, bilateral horizontal gaze palsy much worse towards the right, and a right internuclear ophthalmoplegia (one and a half syndrome); these horizontal gaze defects were less marked on pursuit and doll's head manoeuvre. The rest of the eye movements were normal. Henn and co-workers showed that restricted caudal lesions in the PPRF can produce permanent abolition of horizontal and vertical saccades and quick components of nystagmus in the monkey.' Clinical examples of documented central pontine lesions in cooperative patients are not very common and the effects on vertical gaze have been variable and contradictory.`6 For instance, no case reported had a permanent vertical gaze palsy; some of the patients showed normal vertical eye movements whereas others had only slowed vertical saccades. The case reported here had clinical and radiological evidence of a pontine lesion. However, the lesion was not restricted to the PPRF, since smooth pursuit, saccades and the vestibulo-ocular reflex were also affected. It is possible that involvement of other neighbouring pontine structures such as the VI nuclei, MLF and/or vestibular pathways may partly account for this. A;N. w Fig (a) Enhanced CT scan showing tegmental haemorrhage in the pons. (b) Normal appearance of the enhanced CT scan at the level of the superior colliculi.