Konzo: a Distinct Disease Entity with Selective J Neurol Neurosurg Psychiatry: First Published As 10.1136/Jnnp.56.6.638 on 1 June 1993

63868ournal ofNeurology, Neurosurgery, and Psychiatry 1993;56:638-643

Konzo: a distinct disease entity with selective J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.6.638 on 1 June 1993. Downloaded from upper motor neuron damage

T Tylleskar, W P Howlett, H T Rwiza, S-M Aquilonius, E Stalberg, B Linden, A Mandahl, H C Larsen, G R Brubaker, H Rosling

Abstract Histories were taken and neurological Two Tanzanian patients with konzo were examinations were carried out on both severely disabled by a non-progressive patients during visits in their homes in May spastic paraparesis, since the sudden 1985 and repeated in 1986, 1988 (WPH), onset during an epidemic six years ear- and again in Sweden in 1991 (S-M A). lier. At the time of onset they had a high Magnetic resonance imaging (spin echo) of dietary intake of cyanide from exclusive the brain and spinal cord was performed with consumption of insufficiently processed a Philips T5 MR-scanner (05 Tesla unit) bitter cassava roots. MRI of brain and without use of intravenous contrast. Ti spinal cord were normal but motor weighted, proton density, and T2 weighted evoked potentials on magnetic brain images were obtained in sagittal, coronal, and International Child Health Unit, stimulation were absent, even in the only axial planes of the brain and in the sagittal Department of slightly affected upper limbs. Other neu- plane of the whole spinal cord. Axial proton Pediatrics, University rophysiological investigations were density and T2 weighted images over the Hospital, Uppsala, Sweden largely normal but the more affected conus medullaris were also obtained. T Tylleskar patient had central visual field defects. Neurophysiological investigations were H Rosling Konzo is a distinct disease entity with performed with conventional techniques Kilimanjaro Christian selective type upper motor neuron dam- including motor and sensory nerve conduc- Medical Center, age. tion, small fibre tests such as respiratory Moshi, Tanzania W P Howlett dependent heart rate variation,4 thermal per- Department of (7 Neurol Neurosurg Psychiary 1993;56:638-643) ception thresholds for warmth and cold and Internal Medicine, pain threshold for heat and cold; concentric Muhimbili Medical electromyography (EMG) both at rest, at Center, Dar-es- maxi- Salaam, Tanzania Konzo is an upper motor neuron disease, slight voluntary contraction and during H T Rwiza characterised by abrupt onset of a varying mal contraction (automatic turn/amplitude Department of degree of symmetrical, isolated, and perma- analysis,5 single fibre EMG, fibre density Neurology, University nent but non-progressive spastic paraparesis. measurement, jitter analysis, blink reflexes, Hospital, Uppsala, It was first described in Zaire in 1938 and is electroenchephalography (EEG), somatosen- Sweden S-M Aquilonius named after the local designation in the first sory evoked potentials (SEP), full field visual

and brainstem http://jnnp.bmj.com/ Department of report.' In the last decade it has been evoked potentials (VEP), Clinical reported from remote rural areas of evoked response audiometry (BERA). Neurophysiology, Mozambique, Tanzania, and Zaire. The uni- Transcranial stimulation of the motor cortex University Hospital, and clinical was performed with a magnetic stimulator to Uppsala, Sweden form epidemiological findings E Stalberg have identified konzo as a distinct disease elicit motor evoked potentials (MEP). Department of entity induced by a combined effect of high Stimulation was also performed over the C7 Radiology, Falun cyanide and low sulphur intake from exclu- vertebra. Pure tone audiometry and caloric Hospital, Falun, sive consumption of insufficiently processed stimulation of the vestibularis were studied in

Sweden on September 29, 2021 by guest. Protected copyright. B linden bitter cassava roots.2 The study of konzo has conventional ways. Electronystagmography been limited to clinical examinations (ENG) in a dark room supervised by an Department of formerly Ophthalmology, during field surveys. infrared camera was used for the registration University Hospital, of spontaneous nystagmus, head shake nys- Uppsala, Sweden tagmus, positional nystagmus, and gaze nys- A Mandahl Subjects and methods tagmus at 300 eye deviation. Smooth pursuit Department of Audiology, University Two male konzo patients aged 25 and 19 movements of 600 with a fixed velocity of Hospital, Uppsala, years were invited to Sweden in October 20°/s 10 movement to the right and 10 to the Sweden 1991. They were diagnosed in 1985 during left was recorded with ENG and analysed by H C Larsen an epidemic in Tarime district,"3situated east computer for velocity, accuracy, and superim- Shirati Hospital, of Lake Victoria in the northern part of posed saccades according to standardised Musoma, Mara Region, Tanzania Tanzania. procedure.6 Voluntary saccades (that is, rapid G R Brubaker The study was approved by the ministry of eye movements) of 600 was performed 20 Correspondence to: health of Tanzania and the ethical committee times to the right and 20 times to the left Dr T Tylleskiir, International Child Health of Uppsala University. Local civil and health analysed by computer for start latency, accu- Unit (ICH), Departnent of authorities informed the patients and their racy, and velocity.6 Pediatrics, University Hospital, S-751 85 Uppsala, families about the aim and procedures to be Conventional ophthalmological investiga- Sweden. undertaken and obtained written consent tions included: visual acuity, eye pressure, Received 18 May 1992 from the patients to participate. A local direct/consensual pupillary reflex, corneal and in revised form 16 October 1992. health worker accompanied the patients to sensitivity, colour vision, dark adaptation, Accepted 28 October 1992 Sweden and acted as interpreter. and binocularity. Visual fields were examined Konzo: a distinct disease entity with selective upper motor neuron damage 639

with Goldmann perimetry. The ocular fundi months. He is able to move around with a J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.6.638 on 1 June 1993. Downloaded from were photographed and scrutinised for four point walking frame and he makes his defects in the nerve fibre layer. living as a shoe maker. Serum thiocyanate was used as biomarker In the month preceding the onset the fam- for dietary cyanide exposure.2 An extensive ily was almost exclusively eating cassava, the battery of haematological and biochemical only crop to survive the drought in 1985. His routine analyses were carried out including youngest brother, born in 1973, was similarly analyses of blood cells, platelets, haemoglo- but less severely affected in March 1985. bin, serum levels of electrolytes, different proteins, enzymes, vitamin B6, B12, thyroid PATENT 2 hormones, a plasma electrophoresis, and rou- This man was born in 1972 as the fifth born tine urine tests. Serological testing included: in a family of 12 children. There was no fam- IgM anti-hepatitis A virus, hepatitis B surface ily history of neurological disorder. He was antigen, anti-hepatitis C virus, antibodies to healthy until March 1985 when one night he syphilis, retroviruses HIV-1 + 2 (Enzygnost noticed weakness in both legs while walking AntiHIVi + 2 EIA, Wellcosyme Recom- to the latrine. He returned to bed with diffi- binant HIV-1 EIA, PCR HIV-1), HTLV-I culty but in the morning he was unable to (Abbot HTLV-I EIA, PCR HTLV-I). stand, and over the next three days the weak- Characterisation of lymphocyte subpopula- ness increased and he was unable to get out tions by immunofluorescence and flow of bed. At this stage he also experienced diffi- cytometry as well as an attempt to isolate culties in speaking, visual disturbances and HIV-1 and HTLV-I from blood specimens some tingling sensations in his legs, and a was performed. Direct microscopy of faeces cold feeling in his feet. He also experienced was performed for cysts and eggs of intestinal clumsiness in his hands. He improved over parasites and direct microscopy of blood films some weeks and since May 1985 he is able to for malaria parasites, trypanosomes, and stand and to walk with the help of two sticks. filaria. ELISA serology for amoeba, His hoarseness cleared and his vision echinococcus, filaria, schistosoma, and improved. immunofluorescence for leishmania, try- At the time of onset of the paralysis the panosomes, giardia, and malaria was per- diet of the family consisted almost exclusively formed. The cerebrospinal fluid was screened of cassava because of drought. Out of the for syphilis and examined for cells, protein, total household of two parents and eight liv- glucose, and lactate. A protein electrophoresis ing children four other siblings aged 6, 11, 12 and an isoelectric focusing was done. and 14 were similarly affected in March-April 1985. The only three to escape the disease were the oldest child who was away at a Results boarding school, the youngest who was still PATIENT 1 breast-feeding, and a four year old boy. This man was born in 1966, the eldest of three children. There was no family history of CLINICAL FINDINGS* neurological disorder. He was healthy until Figure 1 shows the two patients in October the 16 March 1985. The night before the 1991, six years after onset. The clinical find- http://jnnp.bmj.com/ onset of his illness he attended a local dance ings from the initial examination in May and experienced no difficulties with his legs. 1985, six weeks after the onset, and the Later that night at home on the way to the examination in October 1991 were almost latrine he noticed a weakness of the legs. The identical. Both patients were unable to walk following morning the weakness was more unaided because of severe spastic paraparesis. pronounced and the legs were trembling on The degree of disability on the expanded dis-

standing. He also experienced some sensa- ability status scale (EDSS) was rated at 7-0 on September 29, 2021 by guest. Protected copyright. tions in the legs, and felt "as if the skin was for patient 1 and 6-5 for patient 2. This dis- not his". However, he managed to go and ability was characterised by extensive motor work in the field and when returning for envolvement in the legs and a typical toe scis- lunch he sat down on a chair to rest. Since sors gait on walking. The main abnormal then he has never been able to walk again. neurological findings were in the legs in both There was a further deterioration during the patients and were as follows: hypertonia, first days with weakness in the arms and most obvious in the strongest muscle slurred speech, but no visual or hearing groups-that is, hip adductors, knee exten- impairments. He was confined to bed for sors and ankle plantar flexors, bilateral sus- approximately two months after which speech tained ankle clonus, loss of power (MRC difficulties cleared and there was a functional range 0/5 ankle dorsi flexors in patient 1 due improvement in his legs and especially in his to contractures, to 4/5 knee flexors in the arms allowing him to stand with the support same patient and 3/5 hip flexors to 4 5/5 of one person. He did not have any bladder plantar flexors in patient 2), hyperreflexia, problems. About one year later, he had an episode of visual impairment when he was unable to read a book but this cleared in *A 7 minute video tape showing the main neurological find- some weeks. In 1989 he married and the cou- ings and the magnetic stimulation is available from the author. ple have since then had two children. His dis- It is free of charge for colleagues from developing countries and for others we charge the equivalent of £20 sterling for ability has remained unchanged apart from production and postage. Unless otherwise requested we will the functional improvements in the first distribute a VHS/PAL tape. 640 Tylleskdr, Howlett, Rwiza, Aquilonius, Stalberg, Linden, Mandahl, Larsen, Brubaker, Rosling J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.6.638 on 1 June 1993. Downloaded from

Figure (A) Frontal view of the patients showing the typical toe scissors gait and their usual walking aids. (B) Side view ofpatient 2.

crossed adductor reflexes, and extensor plan- stimulation over the neck at C7 level. Both tar responses. Sensation (superficial and patients showed normal peripheral nerve con- deep) and superficial reflexes were intact. In duction, EMG, BERA, and blink reflexes.

the upper limbs there was an obvious impair- Respiratory dependent heart rate was also http://jnnp.bmj.com/ ment of fine repetitive and rapid alternating normal. In patient 1 the EEG amplitudes of hand movements affecting both patients. all frequencies were abnormally low. The Patient 2 had a slight (MRC 4-5/5) loss of recordings showed nearly no a activity during power affecting the extensors at the right wakefulness, drowsiness, hyperventilation, or elbow. Otherwise the upper limbs were nor- photic stimulation. In patient 2 EEG was mal. The cranial nerves and mental function normal but sensory nerve action potentials were normal apart from an eye motion abnor- were abnormally low and thermal thresholds on September 29, 2021 by guest. Protected copyright. mality and a visual field defect (both pre- were abnormally high. VEP showed asymme- sented below). The jaw jerk was absent. The try with longer latency from the right eye but rest of the neurological assessment including the absolute values for each eye was within cerebellar, basal ganglia and autonomic func- normal limits. tion was normal. The pure tone audiometry, electronystagmography, and caloric reactions were normal. MRI AND NEUROPHYSIOLOGY In the oculomotor test none of the patients On MRI no abnormalities were found in the had any eye muscle paralysis but both motor cortex, the pyramidal tracts or else- patients had pathological smooth pursuit where in the brain, brainstem, or the spinal movements with regard to velocity and by cord. superimposed saccades. The voluntary sac- The main neurophysiological finding was cades were normal in accuracy and velocity the complete absence of response at repeated but the start latency was too slow magnetic stimulation of the motor cortex (288-308 ms) in both patients. using maximal output of the stimulator at muscle rest and at various degrees of volun- OPHTHALMOLOGY tary muscle contraction for facilitation; both Patient 1 had bilateral central visual field in the affected legs and in the only slightly defects and a corresponding atrophy of the affected arms. There was twitching of facial papillomacular nerve fibre layer with tempo- muscles on magnetic stimulation and normal ral pallor of the optic discs. His colour vision responses were obtained from arms upon (blue-yellow discrimination) was borderline. Konzo: a distinct disease entity with selective upper motor neuron damage 641

Patient 2 had small white spots in the macula typical for konzo, that is, a sudden onset of a J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.6.638 on 1 June 1993. Downloaded from lutea, compatible with a chloroquine symmetric, isolated, and non-progressive retinopathy, and a vasculitis with arteriolar spastic paraparesis, unchanged over time. sheathing in the nasal part of the left retina. The characteristic toe scissors gait, the sym- Other ophthalmological examinations were metrically exaggerated and clonic reflexes in normal. the lower limbs and the extensor plantar responses in the absence of sensory or auto- LABORATORY INVESTIGATIONS nomic disturbances are all constant findings Serum thiocyanate was 532,umol/l for patient in patients with severe konzo.' 7 In the villages 1 and 230 pmol/l for patient 2 in May 1985 of the two patients, konzo constitute two and 62 and 10, umol/ respectively in October thirds of all locomotor disabilities3 and the 1991 (reference value 8-76 umol/1). patients studied belong to the most severely Other laboratory findings were normal affected subjects. except in patient 1 (reference values in paren- A retrovirus aetiology of konzo has been thesis): white blood cell count 3-2 (4-9 109/1) suspected7 but the two konzo patients studied with 26% polymorphonuclear cells; albumin were seronegative to all retroviruses (HTLV- 41 (42-55 g/l); a-amylase 8-0 (1 4-5 pkat/1); 1, HIV-1, and HIV-2), like all other konzo plasma Ig G 21L2 (7-0-18-0 g/l), plasma IgM patients tested. 27 Further tests such as PCR, 3-62 (0A4-2-75 g/l) but no monoclonal virus isolation, and lymphocyte subpopula- bands were found. Patient 2 had normal tions showed no sign of retrovirus affection in values except: albumin 36 (42-55 g/l); the two patients. The distribution of HTLV-I a-amylase 11 -2 (1-4-5 pukat/l); alkaline phos- associated myelopathy/tropical spastic para- phatase 9 4 (0 8-48,ukat/l); AST 0-92 (< 0-6 paresis (HAM/TSP) and konzo overlap geo- ,ukat/l); ALT 0-66 (< 06,ukat/1); LDH 8-5 graphically8 but apart from negative serology (3-8-6-7,ukat/l), and plasma IgM 3-67 konzo also differs clinically from HAM/TSP (0-4-2.75 g/l). by its abrupt onset and non-progressive Both patients were seronegative to HTLV- course. Plasma immunoglobulins were 1, HIV-1, HIV-2, syphilis, and hepatitis in increased in both patients, as expected in 1985 as well as in 1991. No retroviruses subjects from an environment with high rate could be isolated from the blood and of infectious diseases. Patient 1 had a some- CD4+/CD8+ lymphocyte ratios were nor- what increased IgG index of CSF but no mal. oligoclonal bands could be detected. Such Microscopy of faeces and blood films were elevated IgG index are occasionally found in negative in both patients. ELISA serologies non-inflammatory neurological diseases as a were negative for amoeba and echinococcus non-specific sign of neurodamage.9 An earlier and weakly positive to filaria and schistosoma study in Zaire also reported absence of oligo- in both patients. Immunofluorescence for clonal bands in the CSF of konzo patients.7 leishmania, giardia, and trypanosomes was In conclusion, neither this nor earlier epi- weakly positive in both patients but strongly demiological or clinical studies support the positive for malaria (1:5120 and 1:81920). involvement of an infectious agent in konzo. Patient 1 had a traumatic bleed on lumbar At the time of the onset in 1985 both puncture and the collected CSF contained patients subsisted on cassava that, because of http://jnnp.bmj.com/ 7:2 x 109/l erythrocytes, 24 x 106/1 polymor- food shortage, was insufficiently processed.' 3 phonuclear cells, and 18 x 106/l lympho- They had high dietary cyanide intakes as sup- cytes. The bleeding also increased the protein ported by very high serum thiocyanate. High content to 700 mg/l (150-500 mg/l). CSF cyanide exposure has been verified at onset in albumin was 198 mg/l (<290 mg/l), and IgG all areas where konzo have been reported. 122 mg/l (<33 mg/l). Plasma albumin was 37 This and other findings strongly indicate a

g/l (35-46 g/l) and IgG 21-2 g/l (7-15 g/l). toxiconutritional aetiology in konzo.' 2 on September 29, 2021 by guest. Protected copyright. The albumin ratio (CSF/plasma) was 5-3 Lathyrism is another toxiconutritional dis- (<6-5) and IgG index 1-0 (<0 7). Isoelectric ease that is similar to konzo but their geo- focusing showed no oligoclonal bands. graphical distributions do not overlap and Glucose in CSF/blood was 3-2/5 7 mmol/l they are attributed to different types of and CSF lactate was 1-4 mmol/l. Syphilis monotonous diets. The only clinical differ- test was negative. ences between the diseases are the occasional Patient 2 had normal cerebrospinal fluid visual involvement in severe konzo and the with no erythrocytes or polymorphonuclear occasional autonomic involvement in lathy- cells and 1 x 106/1 lymphocytes. The protein rism."0 A common final pathogenetic mecha- content was 240 mg/l, albumin 98 mg/l, and nism has been suggested' but no comparison IgG 30 5 mg/l. Plasma albumin was 36 g/l of the neurodamage in the two diseases can and IgG 17-7 g/l (7-15 g/l) and the albumin be made since similar neuroradiological and ratio (CSF/plasma) was 2-7 and IgG index neurophysiological examinations are lacking 0-65. Isoelectric focusing showed no oligo- for lathyrism. clonal bands. Glucose in GSF/blood was The first interesting result of the present 3.9/5.7 mmol/l and CSF lactate was 1 9 study is the absence of pathological findings mmol/l. Syphilis test was negative. on MRI in the two severely handicapped konzo patients. This is in sharp contrast to the extensive lesions found on MRI in Discussion patients with multiple sclerosis" (MS) with The clinical findings in both patients were the same degree of disability-that is, 642 Tylleskdr, Howlett, Rwiza, Aquilonius, Stdlberg, Lindin, Mandahl, Larsen, Brubaker, Rosling

EDSS > 6. The multifocal lesions frequently nerve fibre layer with temporal disc pallor. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.6.638 on 1 June 1993. Downloaded from seen on MRI in HAM/TSP12 differ in distrib- These findings are characteristic of a ution and are less impressive than those "toxic/deficiency optic neuropathy".Y Earlier found in patients with MS with a similar studies of konzo have noted reduced visual degree of disability. But in contrast to our acuity in severe cases of konzo, especially in normal findings in the two konzo patients, the first weeks following onset.'7 Both the pathological MRI changes were found in 13 optic neuropathy found in this study and ear- out of 14 HAMJTSP"3 patients with a disabil- lier reported findings of reduced visual acuity ity of EDSS > 6. Patients with motor neuron show that in severe cases of konzo an optic disease'4 have limited lesions along the pyra- nerve lesion may add to the isolated damage midal tract that are related to the severity of of corticospinal tracts. the disease. The normal MRI in konzo thus indicates a more selective pathogenetic mech- CYANIDE AND THE NERVOUS SYSTEM anism than in MS, HAM/TSP, and motor The types of neurodamage that has been neuron disease. attributed to cyanide seem to depend on the The second interesting result is the dose rate. Firstly it is well known that very absence of motor evoked potentials on trans- high exposure leads to rapid death through an cranial magnetic stimulation of the motor effect on central neurons. Secondly, the fact cortex in both konzo patients. Difficulties in that sublethal acute intoxications may result exciting the motor cortex electrically or by in. Parkinsonian symptoms following damage magnetic stimulation have been seen in to basal ganglia has been shown on MRI in patients with hyperreflexia, spasticity'5 and such patients.2' Thirdly, several years to extensor plantar responses'6 in relation to the decades of low dietary cyanide exposure from severity of the upper motor neuron involve- cassava has in Nigeria, been associated with a ment.'7 It is therefore not surprising that no syndrome of slowly progressive ataxia, response could be elicited in the severely peripheral neuropathy, and optic atrophy. affected legs but the inexcitability of the This syndrome is known as tropical ataxic upper limbs is surprising since both patients neuropathy22 and is clinically distinct from had a good voluntary control of their arms. konzo. Only one study reports some few patients The abrupt corticospinal damage in konzo with motor neuron disease having absent has been attributed to a fourth pattern of responses after electrical brain stimulation in cyanide exposure, that is, several weeks of muscles with preserved voluntary control.'8 uninterrupted high but sublethal blood Antiepileptic drug medications have also been cyanide levels.2 These result from a high reported to induce a dissociation between dietary cyanide intake in combination with magnetic excitability and voluntary control'9 impaired cyanide to thiocyanate conversion although the patients in this study were not due to low sulphur intake. This metabolic taking any medication. pattern results from exclusive consumption of Frequent involvement of the upper limbs insufficiently processed bitter cassava roots manifested by weakness and hyperreflexia has without protein rich supplementary foods. been reported at onset of konzo.' This usually The dietary situation has been identically clears but may persist in the more severely extreme in all five areas of Africa where konzo http://jnnp.bmj.com/ affected cases. We have observed a similar has been well documented.'2 From the pre- pattern in minimally affected lower limbs sent study we hypothesise that several weeks where residual disease is manifested by per- of high blood cyanide levels at a certain sistent hyperreflexia only. In the two patients threshold, directly or after further metabolic studied the inability to make fine repetitive events, cause an irreversible damage to a hith- and alternating finger and hand movements erto unidentified structure of some of the

in the absence of cerebellar disease is the only upper motor neurons. on September 29, 2021 by guest. Protected copyright. clinically detectable sign of minimal damage The effect on CNS of the pattern of of the corticospinal tracts controlling the cyanide exposure associated with konzo has upper limbs. The complete inexitability of the not been tested in an animal model. An arms to cortical stimulation, while supporting observation of hindlimb paralysis in rats this clinical finding, also suggests that konzo exposed to the combination of a high cyanide may constitute a rare pattern of corticospinal and low sulphur, while testing its diabeto- failure where a subpopulation of neurons may genic effect, is so far the only experimental be damaged at cortical or corticospinal level support for a cyanide aetiology of konzo.2' while relatively normal muscle power and However, continuous exposure of primates to function may be still preserved in the affected the related substance cyanate (OCN) resulted limbs. This supports the case for a further after six weeks in an abrupt onset of a perma- spectrum of minimally detectable or subclini- nent spastic paraparesis identical to konzo.24 cal disease among the konzo affected popula- As only minor histological changes were tions. The abnormal findings in eye motion found in the corticospinal tracts of the that is, saccadic eye movements and inability cyanate exposed primates the toxicological to rapidly alternate the fixation, are most eas- mechanism may be the same as in konzo. The ily interpreted as a bulbar analogue to the proposed toxiconutritional aetiology of konzo findings in the upper limbs. should be possible to confirm in an animal The more severely affected patient in this model. The inexcitability of motor cortex on study had central visual field defects and a transcranial stimulation may be used to corresponding atrophy of the papillomacular identify subclinical forms of konzo in such Konzo: a distinct disease entity with selective upper motor neuron damage 643

as well as in future fluid. ScandJ Clin Lab Invest 1977;37:391-6. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.6.638 on 1 June 1993. Downloaded from models, epidemiological 10 Ludolph AC, Hugon J, Dwivedi MP, et al. Studies on the studies. aetiology and pathogenesis of motor neuron diseases. 1. Lathyrism: Clinical findings in established cases. Brain 1987;110:149-65. The study would not have been possible without support from 11 Truyen L, Gheuens J, Parizel P, et al. Long term follow- the Ministry of Health in Tanzania, Dr Festo Kavishe and Mr up of multiple sclerosis by standardized, non-contrast- Nicolas Mlingi at the Tanzania Food and Nutrition Centre, enhanced magnetic resonance imaging. Jf Neurol Sci the regional and local health authorities, Mr Daniel Kumbi, 1991;106:35-40. Miss Helena Karlen, and involved staff in different depart- 12 Newton M, Cruickshank K, Miller D, et al. Antibody to ments at Uppsala University and Falun hospital. human T-lymphotropic virus type 1 in West-Indian- We also thank Professors Kjell Bergstrom and Anders born UK residents with spastic paraparesis. Lancet Hemmingsson, Department of Diagnostic Radiology, and 1987;i:415-6. Professor Lars Klareskog, Department of Clinical 13 Kira J, Fujihara K, Itoyama Y, et al. Leucoencephalopathy Immunology and Transfusion Medicine, University Hospital, in HTLV-I-associated myelopathy/tropical spastic para- Uppsala, Sweden for valuable advice and Dr Vasilios paresis: MRI analysis and a two year follow-up study Kostulas, Department of Neurology, Huddinge Hospital, after corticosteroid therapy. Jf Neurol Sci 1991 ;106:41-9. Stockholm for performing the isoelectric focusing on the cere- 14 Sales Luis M, Hormigo A, Mauricio C, et al. Magnetic brospinal fluids. resonance imaging in motor neuron disease. Jf Neurol The study was supported by SAREC, the Swedish Agency 1990;237:471-4. for Research Cooperation with Developing Countries. 15 Caramia M, Cicinelli P, Paradiso C, et al. Excitability changes of muscular responses to magnetic brain stimulation in patients with central motor disorders. Electroencephalogr Clin Neurophysiol 1991;81 :243-50. 1 Howlett WP, Brubaker GR, Mlingi N, Rosling H. Konzo, 16 Ugawa Y, Shimpo T, Mannen T. Central motor conduc- an epidemic upper motor neuron disease studied in tion in cerebrovascular disease and motor neuron dis- Tanzania. Brain 1990;113:223-35. ease. Acta Neurol Scand 1988;78:297-306. 2 Tylleskar T, Banea M, Bikangi N, et al. Cassava 17 Hugon J, Lubeau M, Tabaraud F, et al. Central motor cyanogens and konzo, an upper motoneuron disease conduction in motor neuron disease. Ann Neurol found in Africa. Lancet 1992;339:208-1 1. 1987;22:544-6. 3 Howlett W, Brubaker G, Mlingi N, Rosling H. A geo- 18 Thompson P, Day B, Rothwell J, et al. The interpretation graphical cluster of konzo in Tanzania. Jf Trop Geograph of electromyographic responses to electrical stimulation Neurol 1992;2:102-8. of the motor cortex in diseases of the upper motor neu- 4 Stalberg E, Nogues M. Automatic analysis of heart rate rone. JNeurol Sci 1987;80:91-110. variation, Part I: Method and reference values in healthy 19 Hufnagel A, Elger C, Marx W, Ising A. Magnetic motor- controls. Muscle Nerve 1989;12:993-1000. evoked potentials in epilepsy: effects of the disease and 5 Stalberg E, Chu J, Bril V, et al. Automatic analysis of the of anticonvulsant medication. Ann Neurol 1990;28: EMG interference pattem. Electroencephalogr Clin 680-6. Neurophysiol 1983;56:672-81. 20 Miller N, ed. Walsh and Hoyt's clinical neuro- 6 Bergenius J. Computerized analysis of voluntary eye ophthalmology. 4th ed. Baltimore/London: Williams & movements. A clinical method for evaluation of smooth Wilkins, 1982: vol I). pursuit and saccades in oto-neurological diagnosis. Acta 21 Carella F, Grassi M, Savoiardo M, et al. Dystonic- Otolaryngol (Stockh) 1984;98:490-500. Parkinsonian syndrome after cyanide poisoning: clinical 7 Carton H, Kazadi K, Kabeya, et al. Epidemic spastic para- and MRI findings. 7 Neurol Neurosurg Psychiatry 1988; paresis in Bandundu (Zaire). Jf Neurol Neurosurg 51:1345-8. Psychiatry 1986;49:620-7. 22 Osuntokun BO. Cassava diet, chronic cyanide intoxica- 8 Kayembe K, Goubau P, Desmyter J, et al. A cluster of tion and neuropathy in Nigerian Africans. World Rev HTLV-1 associated tropical spastic paraparesis in Nutr Diet 1981;36:141-73. Equateur (Zaire): ethnic and familial distribution. 23 McMillan DE, Geevarghese PJ. Dietary cyanide and trop- Neurol Neurosurg Psychiatry 1990;53:4-10. ical malnutrition diabetes. Diabetes Care 1979;2:202-8. 9 Link H, Tibbling G. Principles of albumin and IgG analy- 24 Shaw CM, Papayannopoulou T, Stamatoyannopuolos G. ses in neurological disorders. II. Relation of the con- Neuropathology of cyanate toxicity in Rhesus monkeys. centration of the proteins in serum and cerebrospinal Pharmacology 1974;12:166-76. http://jnnp.bmj.com/ on September 29, 2021 by guest. Protected copyright.