Subacute Necrotizing Encephalomyelopathy (Leigh's Disease): a Consideration of Clinical Features and Etiology

  • Published on

  • View

  • Download


  • Develop. Med. Child Neurol. 1912, 14, 81-101

    Subacute Necrotizing Encephalomyelopathy (Leighs Disease): a Consideration of

    Clinical Features and Etiology Jonathan H . Pincus

    Introduction Subacute necrotizing encephalomyelop-

    athy (SNE), or Leighs disease, is a degener- ative nervous system affliction; in the large majority it starts before two years, and the great variability in the clinical picture often makes ante-mortem diagnosis very difficult. Feeding problems, weakness, external ophthalmoplegia, loss of vision, swallowing difficulties, ataxia, convulsions and peri- pheral neuropathy have all been described in different combinations. Essentially, the diagnosis has rested upon autopsy findings of necrotizing lesions, with a characteristic distribution within the ner- vous system. The changes which are considered typical include cell loss, demyel- ination and reactive microgliosis, with a relative sparing of neurons in affected regions. There is always a marked increase in the number of capillaries with plump endothelial cells. These lesions, which usually vary in age, are seen mainly in the midbrain and lower brain-stem regions, the spinal cord, thalamus, basal ganglia, dentate nucleus of the cerebellum and, often, the optic nerve and tracts. Involve- ment of peripheral nerves, cerebellar and cerebral cortex and white matter have also been recorded, but less commonly. The

    mammillary bodies are usually spared, otherwise the nature and distribution of the lesions most closely resemble those of Wernickes encephalopathy : for this reason, it has long been supposed that SNE might in some way reflect a disturbance in thiamine metabolism.

    Although the use of thiamine in doses which would be appropriate for dietary thiamine deficiency states has not in any way altered the clinical status of patients with SNE, several children have seemed to benefit from the use of very large amounts of thiamine and thiamine derivatives such as thiamine propyldisulfide (TPD) (Pincus et al. 1971a, Sheremata et al. 1971, Weil et al. 1971). Evaluating the results of therapy in any rare condition is difficult, particu- larly so in conditions such as this where spontaneous remissions have been known to occur. It is therefore of great importance to establish as accurately as possible the natural course of the disease, and the frequency, length and completeness of spontaneous remissions; these seem to vary not only from case to case but also within a single family in which more than one child is affected.

    It seemed also to be worthwhile to review the clinical material available in the

    Associate Professor of Neurology, Department of Neurology, Yale University School of Medicine,

    87 333 Cedar St., New Haven, Connecticut 06510.


    literature to discover clues which might be helpful in diagnosis. Most of the cases reported have been substantiated by autopsy, which immediately prejudices the results and makes the disease appear to be almost uniformly fatal. Even so, the disease is marked by extraordinary varia- bility in age of onset, rapidity of onset and course, as well as by the actual symptoms and signs of disturbance. I t is possible that more benign forms of the disease might be discovered if a reliable chemical test for its presence could be developed and a diagnosis sustained without autopsy. Such a test has been devised and is under evaluation as a clinical diagnostic tool (Cooper et al. 1970).

    In this review we have included 86 cases reported in the literature. We have not included 12 more with whom we have had contact through the urinary test we have developed, and who are as yet unreported.

    Age at Onset As can be seen in Table I, the onset of

    illness occurred before one year of age in more than half the reported cases, and before two years in the large majority of cases. Nonetheless considerable variability exists, and two children were in their second decade at the onset of symptoms (Garcin et al. 1956, Tuthill and Henn 1964). More recently, two cases with onset in adult life have been described (Feigin and Goebel 1969), but exceptional features in these cases have led us not to include them in our Tables.

    The development of the symptoms that lead parents to seek medical attention tends to be insidious in most cases, but it is occasionally subacute with symptoms developing over the course of a week to ten days. More than a dozen cases have occurred with an acute presentation, usually in association with seizures (Poser and Van Bogaert 1960, Tuthill 1960, Tom and Rewcastle 1962, Aronson and Okazaki

    TABLE I Clinical features of SNE

    Age at onset of symptoms < 12 months 52/86 -= 24 months 78/86 > 24 months 8/86

    Rhythm of onset Insidious Subacute Acute

    Course length < 12 months < 24 months > 24 months

    61/86 12/86 13/86

    43/83 55/83 27/83

    Course rhythm Chronic unremitting 46/83 Chronic remitting 23/83 Subacute unremi t ting 12/83 Acute unremitting 2/83

    Respiratory 61 Cranial nerves 3-12 57 Hypotonia 51 Movement disorder (cerebellar or extra-pyramidal) 34 Blindness 31 Babinski sign(s) 30 Seizures 21 Absent tendon reflexes 21 Spasticity 13 M icroencephal y 5

    Incidence of symptoms and signs

    1963, Worsley et al. 1965, Anderson 1966, Bignami et al, 1966, Guazzi et al. 1968, Kamoshita et al. 1968, Crome 1970, Montpetit et al. 1971). As can be seen in Table 11, seizures are characteristic of cases with acute presentations, but other- wise are unusual in this disease. It should be noted, however, that two of these cases were microcephalic. Microcephaly has been reported in a total of six cases (Feigin and Wolf 1954, Guazzi et al. 1968, Crome 1970, de Villard 1970), and its presence suggests that the nervous system has been affected in the fetal, perinatal or postnatal period, even in children whose presenting symptoms are acute and whose early de- velopment is said to be unremarkable. The significance of the mode of presentation for the prognosis is suggested by the short course of the illness in those with an acute presentation, as compared with the longer course in those with an insidious present-



    TABLE I1 Comparison of courses between those with acute and insidious onset

    Age at onset of symptoms: < 12 months < 24 months

    Course length: < 12 months < 24 months > 24 months

    Course rhythm: chronic unremitting chronic remitting subacute unremitting acute unremitting


    Acute onset

    Range: 3 to 36 months

    Range: I day to 36 months

    10113 12/13

    8/12 I I j12 1/12

    4/13 4/13 3/13 2/13


    Insidious onset

    Range: birth to 7 years 36/61 51/61

    21/61 31/61 30161

    45/60 13/60 2/60

    15/61 *

    Range: I month to 15 years


    *Five of these lived for longer than 24 months after onset and five for less than 12 months.

    ation. Seizures themselves did not seem to influence the course of the illness.

    Presenting Complaints The presenting complaints tended to be

    fairly uniform when patients became ill in the first year of life (Table 111). In 45 out of the 52 cases reported, weight loss, weak- ness and some degree of psychomotor retardation were prominent features, either alone or in combination. Anorexia and vomiting were very frequent. Eight cases had periodic unexplained fevers, often associated with the acute development of

    TABLE 111 Most common presenting complaints

    Psychomotor retardation

    Vomiting Weight loss Weakness Seizures

    ( 73 64 60 48 48 29

    ( %) 25 12

    12 20


    Fevers I 15 I - I Movement disorder 2 Coma Hemiplegia Blindness

    38 12 12 6

    I - l l 2 1 Nystagmus

    neurological symptoms such as seizures, which were present in 15 of the 52 cases.

    The presenting complaints were much more variable in those who became ill after 18 months of age. Six of the 14 in this age- group had some form of movement prob- lem which included clumsiness, falling, loss of ability to walk or ataxia. Four of the 15 had some form of mental retardation such as loss of speech, progressive apathy, lethargy or frank dementia. Seizures occurred in three, vomiting in only two and hemiplegia in two cases.

    Remissions In most cases the disease is chronic and

    progresses inexorably over more than six months, but occasionally its tempo is subacute, or even acute with death occurring within two weeks of onset. About 25 per cent of the cases have spontaneous remissions (Table IV), but these are seldom complete and often consist merely of a brighter mood or the resolution of an episode of hyperventila- tion. Remissions are usually brief, lasting only for several weeks, but prolonged periods of improvement have been re- ported which lasted for months or even years. In most cases, remissions have been single, though repeated alternation be-



    TABLE IV Remissions in SNE (23 cases reported)*


    Status during remission Normal

    5 Abnormal


    Clearing of acute exacerbation

    Nature of remission 9

    Reversal of chronic trend


    Less than 6 months Length of partial remission 16

    3 I Length of complete remission I More than 6 months 4 1 Early in course Mid-course Late in course

    Timing of remission 8 13

    *Data concerning remissions were often incomplete.

    tween exacerbation and remission have characterized a few cases (Christensen et al. 1956, Worsley et al. 1965, Dunn and Dolman 1969). While remissions may occur early or late in the course of the disease, they are exceedingly rare once respiratory failure develops, particularly when it lasts more than a day. Only two such cases have been reported: in one, remission was in- complete and lasted for 24 weeks (Thieffry et al. 1965); in the other, remission was virtually complete, lasted for several months and was associated with thiamine and TPD therapy (Pincus et al. 1971).

    While more than half the cases lived for less than a year after the onset of symp- toms, 30 per cent of the reported cases have lived for more than two years. The longest reported survivals of 15 years (Namiki 1965), 12 years (Hardman et al. 1968) and 9 years (Peterson and Alvord 1964, Montpetit et al. 1971) suggest that the disease is not wholly incompatible with a prolonged life. It is conceivable, too, that the milder forms of this disease are even more common than the fatal ones which thus far have been the only ones described.

    Exacerbations The disease progresses in a step-wise

    fashion, with exacerbations often associa-

    ted with fever (Christensen et al. 1956, Anderson 1966, Ford 1966, Procopis et al. 1967, Guazzi et al. 1968, Kamoshita et al. 1968, Richter 1968, Dunn and Dolman 1969, de Villard 1970, Montpetit et al. 1971). These may be followed by a period of stabilization, during which no further deterioration and occasionally some im- provement occurs. Death often occurs during an episode of exacerbation and is often associated with fever. It is not easy to determine whether fevers are a mani- festation of the primary central nervous system problem or whether viral illnesses or other exogenous problems cause neuro- logic deterioration. Sudden exacerbations have been noted in 47 cases, most of which were associated with fever. Crome (1970) noted deterioration after a DPT inocula- tion. In case 2 of Ebels t t al. (1969, a pneu- moencephalogram was followed by apnea and death; in the first case of Robinson et al. (1967), sudden deterioration followed an operation; in case 1 of Peterson et al. (1964) upper respiratory infections led to worsen- ing, and in Case 4 of Reye (1969), sudden worsening followed an attack of rubeola. Seizures have preceded episodes of pro- longed apnea (Lewis 1965). Thus, acute exacerbations in a child with a progressive central nervous system problem must suggest the possibility of SNE.



    Variations Within a Family There is less variability of the age of

    onset within families with more than one child involved. In only three of 23 such fam- ilies has the age of onset varied by more than six months. In the cases of Worsely et nl. (1965) and Montpetit et al. (1971) the span was 14 months, and in the case of Peteison and Alvord (1964) it was 23 years. This is important in establishing the generally good prognosis of normal siblings of affected children who have remained in good health for six months beyond the age at which their affected siblings first manifested symptoms.

    The same has not been true of the course of the disease, which has varied by more than 23 years in five of the 23 families reported. The differences in course among siblings have been 23 years (de Villard 1970), 3 years (Feigin and Wolf 1954), 5 years (Peterson and Alvord 1964), 6 years (Montpetit et al. 1971) and 12 years (Namiki 1965). Thus any evaluation of therapeutic measures in which the length of illness in a sibling is used as a control must be regarded with extreme caution.

    Unfortunately, the onset and course has also varied significantly in a single family. Ebels et al. (1965) reported two siblings, of whom one had an unremitting course and the other had partial remissions following acute exacerbations. Feigin and Wolf (1954) reported a similar family in which the boy died during an acute exacerbation at 17 months after an in- sidious illness of five months duration; his sisters illness also began insidiously at the end of her first year, but she enjoyed a six-month period of weight-gain and in- creased strength before relapsing. More dramatic improvement in siblings has seemed to result from certain therapeutic measures such as the use of lipoic acid (Clayton et nl. 1967). The use of thiamine and TPD has more often been associated with marked improvement in patients


    compared with their untreated siblings (Pincus et a/. 1 9 7 1 ~ ; Sheramata et a / . 1971, Weil et al. 1971), but against the general background of variability in this disease, these reports fall short of proof.

    Heredity Data concerning siblings have been in-

    cluded in reports of 48 autopsy-proven cases of SNE. Eighty-six of the siblings have been normal, 15 have had autopsy-proven SNE and in 20 more the diagnosis was almost certain on clinical grounds. Serious medical problems existed in ten other siblings, but a diagnosis of SNE did not seem justified: these consisted of five neonatal deaths, one child who died of pertussis, two children with schizophrenia, one with cerebral palsy and mental retardation and one with a cerebrovascular accident at one year of age. If these ten are included in the group which was unaffected by SNE, the incidence of SNE in siblings of affected children is 26 .8 per cent. This is close to the theoreti- cal 25 per cent expected in an autosomal recessive disorder. That this is indeed the mode of inheritance is further suggested by the high rate of consanguinity in the parents: in five of the 28 families, con- sanguinity was reported in the parents (Feigin and Wolf 1954, Ebels et a / . 1965, Namiki 1965, Robinson et a/. 1967, Guazzi er a]. 1968). Also, the disease has never been reported in two successivegenerations. The male: female ratio of the disease is 3:2; there are 48 males and 34 females reported. The reason for this is obscure, and may well be chance.

    Symptoms and Signs The general incidence of various symp-

    toms and signs is shown in Table I. Two excellent review articles on this disease have appeared recently, in which the symptoms and signs were discussed in terms of the neuropathology of the disorder (de Villard 1970, Montpetit et a/. 1971); I

  • will therefore limit myself to certain points likely to be helpful in clinical diagnosis. The respiratory symptoms are particularly important in SNE. Disturbances of respira- tion occurred in more than three-quarters of the cases, most often in the latter part of the illness, and constituted the major terminal problems in 60 of the 86 cases included in this review. The respiratory abnormalities varied in kind and included such problems as dyspnea, tachypnea and apnea. Cheyne-Stokes respirations and ataxic respirations were also seen.

    Of great interest were episodes of intermittent hyperventilation or sobbing, the diagnostic importance of which has been stressed by de Villard. He pointed out that such episodes seemed to exist more frequently than had been realized and that episodes of sighing or sobbing may often be brought out by blowing in the childs face or startling the child in some way. He said that it is possible that this reaction might be confused with a simple Moro reflex, a fear response to an unpleasant stimulus or a spasm of sobbing. These interpretations of such symptoms may have been responsible for the low frequency with which they have been reported. Even so, episodes of deep sighing respirations, hyperventilation, rapid irregular breathing and sobbing unassociated with crying were reported in 26 of the 86 cases. The specifi- city of these symptoms is open to question, but they appear to be infrequent in other diseases of the brain-stem, even where they were looked for by de Villard, and this is consistent with my experience.

    Ocular manifestations of the disease have been reviewed (Grover et al. 1970, Borit 1971).

    Although hypotonia is a characteristic of the disease (having been specifically re- ported in 51 of the 86 cases) peripheral neuropathy has not often been well docu- mented, though it is undoubtedly present in many cases Absent deep tendon

    reflexes have been recorded in 21 cases; spinal fluid protein of more than I00 mg/ lOOml has been recorded five times (Peterson and Alvord 1964, Procopis et al. 1967, Robinson et al. 1967, Pincus et al. 1969, Grover et al. 1970). In three cases the presence of peripheral neuropathy has been documented by delayed conduction velocity or sural nerve biopsy (Namiki 1965, Clayton et al. 1967, Robinson et al. 1967). Focal demyelination of peripheral nerves has been reported from time to time when it has been looked for at autopsy. Spinal cord involvement may be another factor in the development of hypotonia (Robinson et al. 1967). The combination of a progres- sive disorder of the central nervous system in association with evidence of peripheral neuropathy, hypotonia or loss of reflexes, often suggests to pediatric neurologists the possibility of metachromatic leukodystro- phy or neuraxonal dystrophy (Wohlwill and Paine 1958); SNE must also be included in the differential diagnosis.

    Although cerebellar signs may be promi- nent in SNE, the disease does not very often seem to have been mistaken for the spino- cerebellar degenerations (Ule 1959). In the case reported by Dunn and Dolman (1969), many of the features associated with Freidreichs ataxia were present, including progressive ataxia, nystagmus, absent deep tendon reflexes, bilateral Babinskis, scolio- sis and some degree of mental retardation. Episodes of hyperventilation and a remit- ting course, together with the absence of a family history of the disease, helps in the differentiation of this form from the usual spino-cerebellar degenerative diseases. Acute cerebellar ataxia was considered in the differential diagnosis in the second case reported by Kamoshita et al. (1968). SNE is more often confused with intracra- nial tumors (such as pontine gliomas) as these tumors characteristically present subacutely with cerebellar, cranial nerve and long tract signs, without increased




    intracranial pressure (Garcin et al. 1956, Richter 1957, Tuthill and Henn 1964, Lakke et al. 1967, Robinson et al. 1967).

    Seizures are usually generalized when they occur, but focal and myoclonic seizures have been reported and infantile spasms have also been seen (Kamoshita et al. 1970).

    Routine laboratory tests have seldom been helpful in diagnosing this disease. Cerebrospinal fluid protein has been elevated in 13 cases (Peterson and Alvord 1964, Ebels et al. 1965, Thieffrey et al. 1965, Procopis et al. 1967, Robinson et al. 1967, Feigin et al. 1968, Kamoshita et al. 1968, Pincus et al. 1969, Crome 1970, Grover etal. 1970), but it is usually normal. It has not been emphasized that a mild lymphocytic reaction may be seen in the spinal fluid (Garcin et al. 1956, Clayton et al. 1967, Greenhouse and Schneck 1968, Hardman et al. 1968). This suggests the possibility of confusion with viral en- cephalitis if a patient with SNE develops fever and seizures and has a mild pleocy- tosis in the spinal fluid. Seventeen patients have been reported with non-specifically abnormal electroencephalograms with generalized slow waves; 15 have had no noted abnormality. Pneumoencephalo- graphy has usually been normal, but this procedure is not without risk in SNE patients and seems to have caused acute exacerbation leading to death in at least one case (Ebels et al. 1965). Urine amino- acid chromatography has been normal in 15 cases. Generalized aminoaciduria has been seen in at least five cases.

    Of much wider interest has been the occasionally reported elevation in blood lactate and pyruvate values (Worsley et al. 1965, Clayton et al. 1967, Hommes et al. 1968, de Villard 1970, Grover et al. 1970, Borit 1971). In the cases reported by Clayton et al., elevations in pyruvate and lactate were always associated with epi- sodes of hyperventilation. Although refer-

    ence is made by the authors to simultaneous acidosis, no reports of blood pH, bicar- bonate or CO, were recorded. In the case of Hommes et al., elevation of blood lactate and pyruvate was associated with a mild alkalosis; blood pH was 7.54. Slightly reduced plasma bicarbonate levels in that case suggest that the patient was hyper- ventilating, though this was not mentioned by the authors. In both of Worsley and his colleague's cases, elevation in lactate and pyruvate, low bicarbonate concentrations and acid blood pH were seen. In both cases hyperventilation lasting for several days preceded chemical estimations. de Villard also reported several such cases : in his case 3, blood lactate and pyruvate were elevated during a glucose tolerance test; venous pH was not abnormal but pCOz ranged from 23 to 28 and bicar- bonate from 12.8 to 16 mEq/L. i n Borit's case, elevated lactate was associated with low blood CO, levels. Normal blood lactates have been found in eight cases (Greenhouse and Schneck 1968, Pincus et al. 1971a, Montpetit et al. 1971). Low blood CO, has been seen in several cases (Procopis et al. 1967, Kamoshita et al. 1968, Crome 1970, David et al. 1970, de Villard 1970). Periodic respiratory alka- losis was reported by Dunn and Dolman (1969).

    Etiologic Considerations Because of the marked variability in

    clinical and biochemical findings, some have felt that the etiology of SNE is also varied, and that the pathologic findings of the disease could be the iesult of more than one dynamic process. Many have felt that there may be a deficiency of pyruvate dehydrogenase or one of its cofactors in this disease, at least in some cases. This would explain the elevations in pyruvate and lactate and, since the enzyme is thiamine pyrophosphate dependent, it might also explain why central nervous



    system lesions resemble those of Wernickes encephalopathy. Although this hypothesis cannot be ruled out at present, it has long been known that pontine tegmental lesions can give rise to central neurogenic hyper- ventilation (Plum and Swanson 1959).

    Hyperventilation itself consistently re- sults in elevated blood lactate and pyruvate levels during passive overbreathing under anesthesia (Chamberlain and Lis 1968), as well as during conditions associated with pathologic hyperventilation (Hucka- bee 1961). The breathing of 5 per cent carbon dioxide during hyperventilation prevents the development of hyperlacta- temia (Eichenholz et al. 1962). One of the theoretical mechanisms by which this could be accomplished is that low carbon dioxide tension could lead to an inhibition of pyruvate dehydrogenase with a subse- quent increase in pyruvate and lactate (Oliva 1970). Eichenholz et al. (1962). showed that during mechanical over- ventilation of anesthetized dogs for a period of up to seven hours, respiratory alkalosis was reversed as lactic and pyruvic acid concentrations rose. Ultimately, in the last hours of their experiment, acidosis ensued. Thus, continued loss of bicar- bonate during the rise in lactic and pyruvic acids did not end at the stage of complete compensation. When hypocapnea was sustained, loss of bicarbonate continued and metabolic acidosis resulted due to accumulation of lactic and pyruvic acids. Such a sequence of events has not been well documented in human beings (whereby respiratory alkalosis progressed to lactic acidosis), though marked lactic acidemia has been seen in two patients with pro- longed hyperventilation caused by brain- stem lesions in Wernickes encephalopathy (Dossetor et al. 1965). Montpetit et al. (1971) have suggested that this is the basic mechanism for occasional elevations of lactate and pyruvate in SNE, and 1 am inclined to agree. Whether or not some

    such cases may also represent a separate metabolic problem remains to be seen. But, suffice it to say that hyperpyruvatemia and hyperlactatemia in hyperventilating patients with pontine disease are not necessarily the result of a primary deficit in the metabolism of pyruvate.

    For the same reason, it would be difficult to know how to interpret the relative inactivity of enzymes which con- cern decarboxylation 01 CO, fixation in patients who have been hyperventilating before the enzyme assays were performed. Blass et a!. (1970) have reported a de- ficiency in pyruvate dehydrogenase in a child with recurrent ataxia, but blood gases and the respiratory status of the patient were not mentioned. Ebadi et al. (1969), in a case of SNE, demonstrated a diminution of glutamate decarboxylase in the basal ganglia and considered the low levels of cerebral amines to reflect a relative in- activity of amino acid decarboxylating systems. Unfortunately, the patients res- piratory status and blood gas determina- tions were not recorded. Hommes et al. (1968) reported relative inactivity of pyruvate carboxylase in a patient with SNE who had evidence of respiratory alkalosis (blood pH 7.54, bicarbonate 21 mEq/L) and elevated blood lactate and pyruvate. Carbon dioxide fixation has been shown to vary with the concentration of CC, in inspired air (Pincus 1969) and it is not unreasonable to suppose that the activity of decarboxylases might also relate to tissue CO, concentrations.

    Estimation of thiamine pyrophosphate- dependent enzymes has revealed no ab- normality of red blood cell transketolase levels (Clayton et al. 1967, Greenhouse and Schneck 1968, Pincus et al. 1969). Similar- ly, no abnormality was found in the pyruvate metabolism of the white blood- cells of an affected patient, as measwed by C1*O, evolution when incubated with pyruvate- I-Cl4 or pyruvate-2-el4. Brain-



    tissue of affected patients has been ana- lyzed for pyruvate dehydrogenase, trans- ketolase and alphaketoglutarate dehydro- genase and the levels were found to be the same as in controls (Pincus et al. 1969).

    At this point, it might be worthwhile to review some of the basic aspects of the function of thiamine in the nervous system which have been undertaken recently. Over the past several years, information has accumulated which supports the hypothesis that thiamine has a specific role in nervous tissue which is independent of its r81e as co-enzyme. Some of these studies are summarized in Table V. There is a

    TABLE V Evidence in support of a coenzyme-independent r6le

    of thiamine in the nervous tissue

    1. Inconsistencies in the correlation between neurological signs in thiamine deficiency and activity of TPP-dependent enzymes in nervous tissue.

    2. Electrophysiological effects of pyrithiamine, an antimetabolite of thiamine, are due to displace- ment of thiamine from the nerve, rather than to enzyme inhibition.

    3. Thiamine is localized in nerve membranes, as opposed to axoplasm.

    4. Both electrical stimulation and neuroactive drugs release thiamine from a variety of intact nerve preparations. Neuroactive drugs also specifically release the vitamin from nerve membrane fractions.

    5. As nerve membranes are purified from a brain homogenate, the percentage of the thiamine triphosphate (TTP) form of the vitamin in- creases.

    6. Thiamine restores the action potential in an ultra-violet irradiated nerve.*

    * Eichenbaum and Cooper (1971) for references.

    suggestion fiom this work that the neurophysiologically active form of thia- mine may be triphosphate (TTP), in contrast to the co-enzyme form thiamine pyro- phosphate (TPP). The enzyme which is responsible for the formation of TTP (TPP- ATP phosphoryl transferase) has been of particular interest. This enzyme is found in a crude mitochondria1 fraction of the brain and catalyzes the following reaction :


    G 95

    This reaction is very rapid, but to date the enzyme has resisted purification and complete characterization. The reaction is assayed spectrophotometrically in either direction by assaying either the formation of ADP in the forward reaction or ATP in the reverse reaction : alternatively, the synthesis of TTP can be determined directly by fluorometry following its isola- tion by paper electrophoresis (Cooper et al. 1970). In our own metabolic studies on SNE, specimens of blood, urine and spinal fluid in a child with SNE inhibited this enzyme. However, they had no inhibitory effect on other thiamine-related enzymes such as thiamine pyrophosphatase or thiamine triphosphatase, nor could any effect be noted on the release of thiamine from perfused nerve preparations. Normal blood and urine had no inhibitory effect on this enzyme, nor did the blood of the patients mother. Curiously, the patients fathers blood did have an inhibitory effect; this was particularly interesting since the father was not neurologically normal. At eight years of age he was diagnosed as having progressive muscular dystrophy ; however, his weakness did not progress, and at 39 years of age he is normal in every way except for weakness in the shoulder and pelvic girdle muscles.

    When this particular child died, samples of brain, liver and kidney were examined to determine the thiamine content of the tissues and the form of thiamine phosphate esters. As might have been predicted from the assay, thiamine triphosphate was absent from the brain, but present in her liver and kidney.

    A preliminary experiment indicated that the liver enzyme which makes TTP is not inhibited by the factor found in patients with SNE. This finding would explain the normal concentrations of TTP in extra- neural tissues. Attempts to characterize the inhibitory material indicate that it is a small protein, perhaps a glycoprotein, with


    a molecular weight somewhere in the range of 30,000.

    Since the first case was described, assays have been performed on the brains of five other patients who died of SNE, in whom TTP was absent or in abnormally low con- centrations in some parts of the brain. Unfortunately, the thiamine content of the most heavily involved portions of the nervous system were not determined in all cases; however, in the three cases where it was possible, there was a highly sugges- tive correlation between regions of maxi- mal neuropathologic involvement and lowest thiamine triphosphate levels (Pincus et al. 1971b). The proportion of total thiamine in the triphosphate form has been the same in children and adult control cases. It is particularly interesting that one of the control values was from a child who died of malnutrition and, although the total value of thiamine in the childs brain was quite low, thiamine triphosphate was still present in fairly normal proportions.

    Another finding of some importance is that the mammillary bodies contain from three to ten times the amount of total thiamine found in other regions of the brain (unpublished data). This was true in control brains as well as in a patient with SNE, and it may explain the sparing of the mamillary bodies in SNE. Experiments with rats and other animals have shown that if they are placed on thiamine-deficient diets, the brain is the last organ to become depleted of thiamine (Rindi 1963). It is possible that there is a lack of restriction in the transport of thiamine between blood and mamillary bodies, so that when human beings are placed on thiamine-deficient diets the mammillary bodies may become deficient long before the rest of the nervous system. This would explain why the mam- millary bodies are almost uniformly in- volved in Wernickes encephalopathy. If this is true, it is possible that when thiamine levels are high, the mammillary

    bodies also become highly enriched with the vitamin. If portions of the nervous system that have high concentrations of total thiamine have a better chance of having some TTP, and this were so in the mammillary bodies, it could explain their infrequent involvement in SNE. If this explanation is correct, it might be construed as evidence that high levels of total thiamine tend to mitigate against the possibility of lesions in a particular region, although the total thiamine level is clearly not the only factor involved. The presence and concentration of the inhibitor are likely to be critically important.

    The use of thiamine in high doses (1 to 3g a day), or large amounts of the thiamine derivative thiamine propyldisulfide, have been tried in 12 cases of SNE known to us. The latter compound and another thiamine derivative, thiamine tetrafurfuryl disulfide (TTFD), in contrast to regular thiamine, are transported by a passive diffusion process so that much higher blood and tissue levels of the vitamin can be achieved: both TPD and TTFD are rapidly converted to thiamine in the body (Pincus et al. 1971a, Thomson et al. 1971). At least temporary remissions have seemed to correlate with the use of these preparations in nine children and they have appeared completely ineffective in only two. In one case, the clinical course has been milder and more prolonged than in older untreated siblings. Indeed, in all cases with untreated siblings, thiamine- treated patients have had milder involve- ment than their siblings.

    A major problem has been the inability to maintain high spinal fluid levels of thiamine, despite continued therapy. This seemed to be the cause of deterioration in one well-documented case (Pincus et al. 1971a) and has also been seen in another case (unpublished data). A curious finding has been that the urine of affected patients ceases to inhibit the phosphoryltransferase within two to three weeks following the





    initiation of thiamine therapy. In several instances relapse has correlated with the reappearance of inhibitory activity of urine. Because of the variability of the natural course of SNE as demonstrated above, it is impossible to state absolutely that treatment with thiamine or its deriva- tives has a beneficial effect; however, initial reports are somewhat encouraging and the ease of administration and the low toxicity of the compounds involved justify further clinical trials.

    Since the initial reports, we have had the opportunity to test many urine specimens received from different medical centers from patients with unusual neurological diseases. The results of these tests are shown in Table VI. There are several points of interest. The disease, though rare, does not seem to be as uncommon as was originally believed : since Leighs descrip- tion (1951), 85 cases have been reported in the literature and in the past 18 months we have detected 10 cases of proven SNE (confirmed by autopsy or family history). Also, as can be noted from this Table, this inhibitor assay is fairly specific. The urines



    of patients with a wide variety of neuro- logical diseases have been tested and shown to be negative with respect to the inhibitor assay. There have been no false negatives and only one (a case of metachromatic leucodystrophy) was definitely false posi- tive. Another child with this same disease had a negative test. Four children with degenerative diseases who had positive urines finally came to autopsy and were found to have SNE, even though there was no family history of the disease. On the other hand, the absence of SNE has been successfully predicted by negative urine tests in two children born to families in which the disease was present, and in a third in which the diagnosis of SNE had been made on clinical grounds but which was proven to be mistaken at autopsy. In some instances the normal parents and siblings of affected children have had positive urine tests: it has been assumed that this repre- sents the carrier state which cannot yet be reliably distinguished from the disease state on the basis of the urine test.

    The urines of a number of patients with unusual neurological syndromes have had



    TABLE V I An appraisal of the specificity of the TPP-ATP

    phosphoryl transferase assay for SNE



    Autopsy-proven cases of SNE

    Living patients with SNE (autopsy in a sibling)

    Patients suspected of SNE but proven to have other conditions

    Patients once suspected of SNE, now felt to have other conditions

    Patients suspected of SNE, no diagnosis yet established







    Inhibitory Assay

    Positive Negative

    1 , 12* *These have included Werdnig Hoffmans disease, cerebral arteritis,

    Guillian-Barre syndrome, multiple sclerosis, metachromatic leukodystrophy, Barters syndrome, tuberous sclerosis, progressive supranuclear ophthal- moplegia, Wernickes encephalopathy and carbamyl phosphate transferase deficiency.



    an inhibitory effect and these may therefore represent aberrant cases of SNE. These have included two cases with Undines curse (alveolar hypoventilation); in one of these cases, the mother of the patient also had the inhibitor in her urine. This is particu- larly interesting since three of the five cases of this respiratory syndrome in which neuropathologic findings were reported have shown pathology closely resembling that of SNE, both in the nature of the lesions and in their distribution (Naeye 1965, Seriff 1965). Positive urines have also been seen in several children with cerebellar disease. One such case has been reported by Lonsdale el af . (1969). This patient had intermittent cerebellar ataxia associated with high blood pyruvate levels. We have also found positives in a family in which the children have cerebellar ataxia and

    myoclonic seizures. We are continuing this screening in order

    to appraise further the specificity of this inhibitor assay. It is conceivable that there are several variants of SNE which have yet to be identified. In particular, we hope to investigate patients with a variety of progressive diseases which appear to be transmitted as autosomal recessives, in- cluding cerebellar, muscular and extra- pyramidal disorders. Also children with encephalitis in which no infectious agent can be identified may offer a fruitful field for research. Ultimately, we hope to purify the inhibitor, prepare an antibody to it, and attempt therapy with the antibody.

    Acknowledgements: The author wishes to express his gratitude to his colleague, Dr. Jack R. Cooper, for his assistance in the vrevaration of this vaver. This study was supported inpart by USPHS Giant NS4962241.

    SUMMARY Eighty-six cases of subacute necrotizing encephalomyelopathy (SNE) have been reviewed

    in order to establish asaccuratelyabpossiblethe natural course andcharacteristic symptoma- tology of the disease. In most cases the diagnosis was confirmed by autopsy, which introduces a bias toward a poor prognosis. Laboratory tests have not usually been helpful diagnostically, but low blood C02 content, elevated lactate and pyruvate and mild pH changes have been reported in several cases. These abnormalities may be related to the hyperventiltion which results from neurogenic respiratory disturbances which have been seen in a high proportion of reported cases. The finding that the urine of affected patients contains an inhibitor of the enzyme which catalyzes thiamine triphosphate (TTP) formation may be the basis of a specific diagnostic test. TTP was either totally absent or was present in abnormally low concentrations in the brains of six children who died of SNE and these find- ings have led to therapeutic trials with high doses of thiamine, thiamine propyldisulfide and thiamine tetrafurfuryldisulfide. Initial results of therapy have been encouraging and further clinical trials seem justified. Ultimately, however, elimination of the inhibitor is likely to prove to be the only completely successful mode of therapy.

    R G S U M ~ Encephalomydlopathie nicrosante subaigue (maladie de Leigh):

    Considdrations sur les symptomes cliniques et Idtiologie 86 cas denc6phalomyClopathie nkcrosante subaigue (SNE) ont Ct6 revus dans la litterature

    pour Ctablir 16volution et les caractdristiques de la maladie. Dans la plupart des cas rapportbs, le diagnostic a CtC confirm6 par lautopsie, ce qui indique un tendence A un mauvais pronostic. Les examens de laboratoire nont pas Ct6 dune grande aide pour le diagnostic mais dans plusieurs cas, il a 6t6 signal6 une faible pression partielle de CO, dans le sang, des taux 61ev6s de lactate et de piruvate et de 16gkres modifications du pH.



    Ces anomalies peuventstre relites A Ihyperventilation qui provient de troubles respiratoires neuroghes, observts dans une forte proportion de cas.

    Un test spkcifique de diagnostic peut reposer sur la dkcouverte que lurine des malades contient un inhibiteur de lenzyme qui catalyse la formation de la thiamine triphosphate (TTP). La TTP dtait totalement absente ou prtsente B des concentrations anormalement basses dans les cerveaux de 6 enfants morts de SNE, et ces dtcouvertes ont conduit A des essais thkrapeutiques avec de fortes doses de thiamine, de thiamine propyldisulfide et thiamine tetrafurfuryldisulfide. Les premiers rtsultats de ce traitement sont encourageants et justifient dautres essais cliniques.

    ZUSAMMENFASSUNG Subakute nekrotisierende Encephalomyelopathie (Leighsche Erkrankung):

    klinische Merkmale und A tiologie 86 Falle mit subakuter nekrotisierender Encephalomyelopathie (SNE) wurden aus der

    Literatur zusammengetragen, um den Verlauf und die charakteristischen Merkmale der Krankheit festzustellen. Bei den meisten beschriebenem Fallen wurde die Diagnose erst durch die Autopsie gesichert, was beweist daD die Prognose sehr schlecht ist. Laborunter- suchungen haben wenig zur Diagnose beigetragen, in mehreren Fallen allerdings wurden niedrige C0,-Werte in Blut, erhohte Laktat-und Pyruvat-Spiegel und geringe pH- Veranderungen beschrieben. Diese Veranderungen sind moglicherweise durch die Hyper- ventilation bedingt, welche durch neurogene respiratorische Storungen, die bei einer Reihe der Falle beobachtet wurden, hervorgerufen werden.

    Der Befund, da8 der Urin dieser Patienten einen Inhibitor des Enzyms enthalt, das die Bildung von Thiamintriphosphat (TTP) katalysiert, ist moglicherweise die Grundlage eines spezifischen diagnostischen Testes. TTP war in den Gehirnen von 6 Kindern, die an SNE starben, entweder iiberhaupt nicht oder nur in extrem niedrigen Konzentrationen vorhan- den, und diese Befunde veranlal3ten therapeutische Versuche mit hohen Dosen von Thiamin, Thiaminpropyldisulfid und Thiamintetrafurfuryldisulfid. Die ersten Ergebnisse dieser Therapie waren Erfolg versprechend und es scheint gerechtfertigt, weitere klinische Versuche zu unternehmen.

    RESUMEN Encefalomielopatia subaguda necrotizante (enfermedad de Leigh) : una consideracidn sobre 10s caracteres clinicos y la etiologia

    Se revisan ochenta y seis casos de encefalomielopatia necrotizante subaguda de la literatura con el objeto de establecer el curso y caracteristicas de la enfermedad. En la mayoria de 10s casos estudiados el diagnbsticb a la autopsia lo que indica un tendencia a un pobre pron6stico.Los tests de laboratorio no han sido de gran ayuda para eldiagnbstico, si bien en varios casos se meciona un contenido bajo de CO, en sangre, niveles altos de lactato y piruvato y cambios moderados en el pH. Estas anormalidades pueden estar relacionadas con la hiperventilacibn que resulta de las alteraciones respiratorias neurb- genas vistas en una alta proporcibn de casos.

    El hallazgo de que la orina de pacientes afectos contiene un inhibidor del ensima que cataliza la formaci6n de trifosfato de tiamina (TTP) puede constituir la base de un test diagnbstico especifico. El TTP estaba totalmente ausente o presente s610 en concentraciones anormalmente bajas en 10s cerebros de 6 niiios que murieron de la enfermedad. Estos hallazgos han conducido a tentativas terape6ticas con altas dosis de tiamina, tiamina propildisulfido y tiamina tetrafurfurildisulfido. Los resultados iniciales son alentadores y parecen justificar nuevas pruebas clinicas.




    Anderson, R. M. (1966) Four cases of subacute necrotizing encephalomyelopathy in childhood (Leighs

    Aronson, S. M., Okazaki, H. (1963) Clinical Neuropathological Conference. Diseases of the Nervous

    Bignami, A,, Castello, M., Colloridi, V., Zappella, M. (1966) Encefalite necrotizzante subacuta. Acta

    Blass, J. P., Avignan, J., Uhlendorf, B. W. (1970) A defect in pyruvate decarboxylase in a child with an

    Borit, A. (1971) Leighs necrotizing encephalomyelopathy. Neuro-ophthalmological abnormalities.

    syndrome). Proceedings of the Australian Association for Neurology, 4, 97.

    System, 24,630.

    Paediatrica Latina, 19, 396.

    intermittent movement disorder. Journal of Clinical Investigation, 49, 423.

    Archives of Ophthalmology, 85,438. Chamberlain, J. H., Lis, M. T. (1968) Observations of blood lactate and pyruvate levels and excess lactate

    uroduction during and after anaesthesia with and without hyperventilation. British Journal of Anaesthesia, 40, 315.

    Christensen, E., Melchior, J., Plum, P. (1956) Combined lesions of basal ganglia, medulla oblongata, and spinal cord in a 10 year-old-boy. Acta Paediatrica (Uppsala), 45,396.

    Clayton, B. E., Dobbs, R. H., Patrick, A. D. (1967) Leighs subacute necrotizing encephalopathy: clinical and biochemical study, with special reference to therapy with lipoate. Archives of Disease in Childhood, 42, 467.

    Cooper, J. R., Itokawa, Y., Pincus, J. H. (1969) Thiamine triphosphate deficiency in subacute necrotizing encephalomyelopathy. Science, 164, 72. - Pincus. J. H., Itokawa, Y., Hashitani, Y., Piros K. (1970) Experience with thiamine phosphoryltrans-

    ferase inhibition in subacute necrotizing encephalomyelopathy. New EnglandJournalof Medicine, 283,793. Crome, L. (1970) Subacute necrotizing encephalomyelopathy associated with renal and arterial lesions.

    Brain, 93, 709. David, R. B., Gomez, M. R., Okazaki, H. (1970) Necrotizing encephalomyelopathy (Leigh). Developmental

    Medicine and Child Neurology, 12,436. Dossetor, J. B., Zborowski, D., Dixon, H. B., Pare, J. A. P. (1965) Hyperlactemia due to hyperventilation:

    use of CO, inhalation. Annals of the New York Academy of Sciences, 119, 1153. Dunn, H. G., Dolman, C. L. (1969) Necrotizing encephalomyelopathy. Report of a case with relapsing

    DolvneuroDathy and hyperalaninemia and with manifestations resembling Friedreichs ataxia. Neiirdogy


    {Mhneapdlis), 19, 536: Ebadi, M. S , Bostad, R., Pellegrino, R. J. (1969) Impairment of pyridoxal phosphate dependent metabolic

    reactions in a child with subacute necrotizing encephalopathy. Journal of Neurology, Neurosurgery and Psychiatr.v, 32, 393.

    Ebels, E. J., Blokzijl, E. J., Troelstra, J. A. (1965) A Wernicke-like encephalomyelopathy in children (Leigh), an inborn error of metabolism? Report of 5 cases with emphasis on its familial incidence. Helvetica Paediatrica, 20, 3 10.

    Eichenbaum, J. W., Cooper, J. R (1971) Restoration by thiamine of the action potential in ultraviolet irradiated nerves. Brain Research, 32,258.

    Eichenholz, A., Mulhausen, R. O., Anderson, W. E., MacDonald, F. M. (1962) Primary hypocapnia: a cause of metabolic acidosis. Journal ~f Applied Physiology, 17,283.

    Feigin, I., Goebel, H. H. (1969) Infantile subacute necrotizing encephalopathy in the adult. Neurology (Minneapolis), 19,748.

    - Wolf, A. (1954) A disease in infants resembling chronic Wernickes encephalopathy. Journal of Pediatrics, 45, 243. - Pena, C. E., Budzilovich, G. (1968) The infantile spongy degenerations. Neurology (Minneapolis), 18,

    153. Ford, F. R. (1966) Diseases of the Nervous System in Infancy, ChildhoodandAdolescence. 5th edn. Springfield,

    Ill. : C. C. Thomas. Garcin, R., Gruber, J., Godlewski, S. (1956) Spongiose dissCminCe de Ienckphale Cvoluant cliniquement

    par pousseks chez un enfant malgache. Ses rapports Cventuels avec Ienkphalopathie subaigue de Wer- nicke. Revue Neurologique, 95, 273.

    Greenhouse, A. H., Schneck, S. A. (1968) Subacute necrotizing encephalomyelopathy. A reappraisal of the thiamine deficiency hypothesis. Neurology (Minneapolis), 18,. 1 .

    Grover, W. D., Green, W. R., Pileggi, A. J. (1970) Ocular findings in subacute necrotizing encephalomyel- itis. American Journal of Ophthalmology, 70, 599.

    Guazzi, G. C., Martin, J. J., Brucher, J. M., Taper, H. S., Macken, J., Neetens, A., Van Haegenborgh, J. (1968) Sur Iimportance de Iatteinte vasculaire et de la dystrophie gliale dans lenkphalomyelopathie nkcrosante de Leigh. Etude de deux familles et de trois observations anatomiques. Journal of Neurological Science, 7 , 357

    Hardman, J. M., Allen, L. W., Baughman, F. A,, Waterman, D. F. (1968) Subacute necrotizing encephalo- pathy in late adolescence. Archives of Neurology, 18,478.

    Hommes, F. A., Polman, H. A., Reerinck, J. D. (1968) Leighs encephalomyelopathy: an inborn error of gluconeogenesis. Archives of Disease in Childhood, 43,423.

    Huckabee, W. E. (1961) Abnormal resting blood lactate. I. The significance of hyperlactemia in hospitalized patients. American Journal of Medicine, 30, 833.

    Kamoshita, S., Aguilar, M. J., Landing, B. H. (1968) Infantile subacute necrotizing encephalomyelopathy. American Journal of Diseases of Children, 116, 120.



    - Mizutani, I., Fukuyama, Y . (1970) Leighs subacute necrotizing encephalomyelopathy in a child with infantile spasms and hypsarrhythmia. Developmental Medicine and Child Neurology, 12,430.

    Lakke, J. P. W. F., Ebels, E. J., Thye, 0. J. (1967) Infantile necrotizing encephalomyelopathy (Leigh). Archives of Neurology, 16, 227.

    Leigh, D. (1951) Subacute necrotizing encephalomyelopathy in an infant. Journal of Neurology, Neuro- surgery and Psychiatry, 14, 216.

    Lewis, A. J. (1965) Infantile subacute necrotizing encephalopathy. Canadian Medical Association Journal, 93, 878.

    Lonsdale, D., Faulkner, W. R., Price, J. W., Smeby, R. R. (1969) Intermittent cerebellar ataxia associated with hyperpyruvic acidemia, hyperalaninemia, and hyperalaninuria. Pediatrics, 43, 1025.

    Mpntpetit, V. J. A., Andermann, F., Carpenter, S., Fawcett, J. S., Zborowska-Sluis, D., Giberson, H. (1971) Subacute necrotizing encephalomyelopathy-a review and a study of two families. Brain, 94, 1.

    Naeye, R. L. (1965) Alveolar hypertension and cor pulmonale secondary to damage to the respiratory center. American Journal of Cardiology, 8,416.

    Namiki, H. (1965) Subacute necrotizing encephalomyelopathy. Case report with special emphasis on associated pathology of peripheral nervous system. Archives of Neurology, 12,98.

    Oliva, P. B. (1970) Lactic acidosis. American Journal qfMedicine, 48, 209. Peterson, H. deC., Alvord, E.C. (1964) Necrotizing encephalopathy with predilection for the brain stem.

    Subacute infantile and chronic juvenile familial forms. Transactions of the American Neurological Associa- tion, 89, 104.

    Pincus, J. H. (1969) Carbon dioxide fixation in rat brain: relationship to cerebral excitability. Experimental Neurology, 24, 339. - ltokawa, Y. , Cooper, J. R. (1969) Enzyme inhibiting factor in subacute necrotizing encephalomyelo-

    pathy. Neurology (Minneapolis), 19, 841. - Cooper, J. R , Itokawa, Y., Gumbinas, M. (1971a) Subacute necrotizing encephalomyelopathy. Effects

    of thiamine and thiamine propyl disulfide. Archives of NeuroloRy, 24,511. - Solitaire, G. B., Itokawa, Y. , Hashitani, Y. , Cooper, J. R. (1971 b) Thiamine, thiamine triphosphateand

    nervous system lesions in subacute necrotizing encephalomyelopathy. Neurology (Mineapolis), 21,444. Plum, F., Swanson, A. G. (1959) Central neurogenic hyperventilation in man. Archives of Neurology and

    Psychiatry, 81,535. Poser, C., van Bogaert, L. (1960) Leuco et polio enckphalopathies sym6triques nkrosantes. Revue Neuro-

    logique, 103, 3. Procopis, P G., Turner, B., Selby, G. (1967) Subacute necrotizing encephalopathy in an acidotic child.

    Journal of Neurology, Neurosurgery and Psychiatry, 30, 349. Reye, R. D. K. (1960) Subacute necrotizing encephalomyelopathy. Journal of Pathology and Bacteriology,

    79, 165. Richter, R. B. (1957) Infantile subacute necrotizing encephalopathy with predilection for the brain stem.

    Journal of Neuropathology and Experimental Neurology, 16,281. - (1968) Infantile subacute necrotizing encephalopathy (Leighs disease). Neurology (Minneapolis), 18,

    1125. Rindi, G. (1963) Recent progress on the mechanism of action of some structural analogues of thiamine.

    Bolletino Chimico-Farmaceutico, 102, 363. Robinson, F., Solitaire, G. B., Lamarche, J. B., Levy, L. L. (1967) Necrotizing encephalomyelopathy of

    childhood. Neurology (Minneapolis), 17,472. Seriff, N. S. (1965) Alveolar hyperventilation with normal lungs: the syndrome of primary or central

    alveolar hyperventilation. Annals of the New York Academy of Sciences, 121,691. Sheremata, W. A,. Cooper, J. R., Joseph, J. (1971) Leighs disease: response to high dose thiamine therapy.

    Neurology (Minneapolis), 21,456. Thieffry, S., Farkas-Bargeton, E., Martin, C., Lyon, G. (1965) Endphalopathie nkcrosante subaigue de

    Ienfant. Revue Neurologique, 113, 105. Thomson, A. D., Frank, O., Baker, H., Leevy, C. M. (1971) Thiamine propyl disulfide: absorption and

    utilization. Annals of Internal Medicine, 74, 529. Tom, M. I., Rewcastle, N. B. (1962) Infantile subacute necrotizing encephalopathy. Neurology (Minneapolis),

    12, 624, Tuthill, C. R. (1960) Der morphologische Wernicke-Komplex in friihem Kindesalter. Familiare Erkrankung

    bei 7 Monate alten Zwillingen. Archiv ,fur Psychiatrie und Nervenkrankheiten, 200, 520. - Henn, R. (1964) Wernicke-syndrom in Kindesalter ohne Anzeichen von Mangelernahrung. Archiv f i r

    Psychiatrie und Nervenkrankheiten, 205, 116. Ule, G. (1959) Uber eine der Wernickeschen Pseudoencephalitis entsprechende Encephalopathie bei

    Kindern. Virchows Archiv fur pathologische Anatomie und Physiologie, 332, 204. devillard, R. (1970) LEnckphdomy6lopathie necrotique subaigue infantile. M.D. Thesis, University of

    Lyons. Weil, M. L., Menkes, J. H., Shaw, K. N. F., Pincus, J. H. (1971) Pyridoxine responsive inhibition of

    thiamine metabolism. Clinical Research, 19. 229. WohlwiU, F. J., Paine, R. S. (1958) Progressive demyelinating leukoencephalopathy. Neurology (Minnea-

    polis), 8, 285. Worsley, H. E., Brookfield, R. W., Elwood, J. S., Noble, R. C., Taylor, W H (1965) Lactic acidosis with

    necrotizing encephalopathy in two sibs. Archives of Disease in Childhood, 40,492.