614 Archives ofDisease in Childhood 1993; 69: 614-618

REGULAR REVIEW Arch Dis Child: first published as 10.1136/adc.69.5.614 on 1 November 1993. Downloaded from

Infantile spasms

Richard E Appleton

Infantile spasms represent a disorder disorders that may be mistaken as spasms,8 with unique clinical and electroencephalo- including benign infantile sleep myoclonus, graphic () features and a poor opisthotonic posturing due to spasticity, and prognosis including chronic intractable gastro-oesophageal reflux. At the onset, and psychomotor retardation. The spasms may be infrequent and occur singu- association of spasms and hypsarrhythmia, larly; however, within days, the spasms occur with or without mental retardation, defines in clusters particularly upon awakening or on West's syndrome. West's syndrome is not falling asleep. The clusters may consist of as uncommon; the incidence is considered to be few as five, or as many as 100 spasms with 0 16-0-42 per 1000 live births.' In 1991, over between three and 30 seconds between each 760 000 live births occurred in England, spasm and may be followed by irritability and Wales, and Scotland (Office of Population crying, lethargy, and drowsiness.5 This cluster- Censuses and Surveys, personal communica- ing of spasms is almost unique to this age tion) giving an estimated number of 122 to 319 group and tends to diminish over a period of new cases ofWest's syndrome each year. In the months, disappearing in most children by the 150 years since the original description, there age of 2 years. In older children each spasm has been little progress in the understanding of lasts longer and represents more a tonic the pathophysiology of the spasms, although seizure, one of the characteristic some advance has been made in their classifi- seen in the Lennox-Gastaut syndrome'.3 cation and aetiology. Treatment has remained essentially empirical, but there is increasing evidence that the newer antiepileptic drugs The electroencephalogram (EEG) and even surgery may be of benefit. This Hypsarrhythmia is the characteristic EEG paper reviews the current 'understanding' and pattern seen in children developing spasms in areas of future development in this seizure the first year of life, and was recognised in the http://adc.bmj.com/ disorder. early 1950s.9 It consists ofhigh voltage (greater than 200 microvolts) and multifocal spikes, spike and wave discharges, chaotic slowing, The spasms and asynchrony. This activity may occur con- Spasms were originally reported in 1841 by Dr tinuously or in bursts, may be absent in the West in his own 4 month old son; his detailed waking state appearing only in non-rapid eye description remains unsurpassed and should movement (non-REM) sleep, and disappear- on September 23, 2021 by guest. Protected copyright. be read.2 Spasms may develop between 1 day ing in REM sleep. A sleep recording should and 5 years of age,3 but the vast majority of therefore be obtained in infants suspected of children have an onset between the ages of 4 having spasms particularly if an initial, or and 9 months4 5; 90% begin within the first 'awake' EEG is non-specifically abnormal. The year of life. Boys appear to be more frequently occurrence of a spasm during an EEG is fre- affected by a ratio of approximately 1 3:1.5 6 quently associated with a relative flattening Synonyms include lightning or jacknife convul- (attenuation) or suppression of the trace, sions and salaam . A spasm is due to a rather than by a spike discharge.7 Hypsar sudden muscular contraction that is usually rhythmia is usually generalised but may be generalised but may be asymmetrical, involves asymmetrical or even unilateral.'0 1 It is rarely muscles of the neck, trunk, and limbs and is seen after the age of 3 years. The pattern of accompanied by a brief loss of consciousness. hypsarrhythmia may be used prognostically in The spasms are usually very brief and transient certain children, although there is no clear rela- (myoclonic), or less commonly more sustained tionship.7 Unilateral or grossly asymmetrical (tonic), lasting a few seconds. The most recordings (which may represent a persistent common seizure pattern is mixed flexor and focal lesion) are associated with a poorer out- extensor.7 Where the spasm is massive, and come, in contrast to 'typical' symmetrical The Roald Dahl EEG flexor the infant may double over and cry out, patterns, where the response to treatment and Unit, Royal Liverpool which may lead to a misdiagnosis of colic.7 In prognosis are more favourable. Less abnormal Children's NHS Trust other infants the spasms may be subtle, such as patterns of hypsarrhythmia are also reported (Alder Hey), Eaton Road, Liverpool L12 a slight head nod, or even a momentary to have a better outcome. The degree of 2AP upward deviation of the eyes and eyelids, 'abnormality' of a hypsarrhythmic record is which may also difficult to is and Correspondence to: delay diagnosis. Conversely, define, frequently subjective Dr Appleton. there are other, non-epileptic paroxysmal may simply reflect the patient's age; in the Infantile spasms 615

author's experience there is no correlation neurological dysfunction without a demon-

between outcome and degree of abnormality. strable cause. The much smaller idiopathic Arch Dis Child: first published as 10.1136/adc.69.5.614 on 1 November 1993. Downloaded from group is defined by normal development before the onset of symmetrical spasms, Associated clinical features normal examination, normal computed Mental retardation is common in West's syn- tomography and magnetic resonance imaging, drome and may precede the onset of spasms in hypsarrhythmia, and absence of any focal EEG approximately 70% of patients12 13; this figure abnormality.'5 The acceptance of this idio- may be greater as mild developmental delay pathic group is not universal and probably may be difficult to identify, particularly retro- constitutes only a very small proportion of all spectively. Eventually almost 90% of children infants with West's syndrome. Clearly, the will manifest some degree of mental retarda- definition of the symptomatic and cryptogenic tion.13 In previously well or 'normal' children, groups is dependent upon the number and there may be a plateau, or even regression in degree of sophistication of investigations. The development. Children are hypotonic and advent of first computed tomography and then visually inattentive and may therefore be con- magnetic resonance imaging has clearly sidered to be visually impaired. Even in resulted in an altered proportion of sympto- children with abnormal development before matic versus cryptogenic cases. The develop- the onset of spasms, the development of ment of functional brain imaging such as seizures is frequently reflected in a definite positron emission tomographyl8 may, theoreti- regression. An arrest in development may be cally, identify additional symptomatic cases, the presenting feature in those infants with although this technique is obviously not subtle spasms. Once the spasms resolve, devel- ubiquitous, and is restricted to only a few opmental progress may be resumed, although centres within the UK. Finally, the use of this is not invariable and depends on both the simultaneous video-EEG monitoring has led to underlying cause and response to treatment. a much more precise characterisation of infan- Disorganised sleep is also frequently associ- tile spasms in terms of clinical and EEG symp- ated with infantile spasms,5 14 as manifest by a tomatology. It is now realised that at least in reduction in both total sleep but also REM some children spasms may in fact have a partial sleep.'4 The disturbed REM sleep pattern onset with rapid secondary generalisation, appears to be related directly to the frequency rather than being always generalised in onset, of spasms at least in some patients, as electro- as previously believed and which is reflected in clinical suppression of the spasms by steroid the 1989 ILAE epilepsy classification. 16 treatment is accompanied by an increase in REM sleep. 14 This apparent association between spasms and REM sleep has led to Aetiology and pathophysiology speculation about a common pathogenesis Infantile spasms have multiple aetiologies that involving the pons, or other brainstem struc- are commonly separated into prenatal, peri- tures. 14 natal, and postnatal,5 12 depending on the tim- http://adc.bmj.com/ ing of the cerebral insult; prenatal causes are the most common.19 There is no pathognomic Classification of spasms lesion for West's syndrome. It is inappropriate It is in the terminology and classification of to discuss all causes in this paper; common spasms that some progress has recently been aetiologies include cerebral dysgenesis or mal- made.'5 Firstly, the seizure type classically formations (including tuberous sclerosis and termed infantile spasms typically occurs in chromosomal disorders) and as a sequel of on September 23, 2021 by guest. Protected copyright. infancy where it is usually associated with neonatal hypoxic-ischaemic encephalopathy7 hypsarrhythmia (West's syndrome) but may (table). Metabolic or toxic disorders and also occur in later childhood where it may intrauterine infections are not common causes occur as part of a chronic epilepsy syndrome. This has implications for the International Aetiology ofinfantile spasms League Against Epilepsy (ILAE) 1989 classifi- Dysgenesis cation where infantile spasms is synonymous Tuberous sclerosis, neurofibromatosis, incontinenta pigmenti, with West's syndrome and listed in the Aicardi's syndrome, Sturge-Weber syndrome, cortical dysplasias (including hemimegalencephaly), heterotopias, generalised cryptogenic or symptomatic holoprosencephaly, lissencephaly, and schizencephaly (age related) 16; it is proposed that in Hypoxic-ischaemic insult addition to this syndrome classification, Prenatal (multicystic encephalomalacia, porencephalies, hydranencephaly) 'spasms' should be included as a specific Perinatal (hypoxic-ischaemic encephalopathy) seizure type which may occur outside West's Postnatal (cardiac arrest, near drowning) syndrome. Secondly, 'infantile spasms' (in this Haemorrhage and trauma Perinatal (intraventricular and periventricular haemorrhage) context referring directly to spasms occurring Postnatal (subarachnoid and subdural haemorrhage) in infancy with hypsarrhythmia, that is West's Infections syndrome) may be classified according to Prenatal (cytomegalovirus, toxoplasmosis, rubella, syphilis) aetiology, into symptomatic, cryptogenic, and Postnatal (purulent meningitis, cerebral abscess, encephalitis) Metabolic and toxic idiopathic cases.17 The symptomatic group is Prenatal (phenylketonuria, non-ketotic hyperglycinaemia, characterised by the existence of neurological hyperomithinaemia, homocitrullinaemia, histidinuria, Leigh's syndrome, pyridoxine (vitamin B6) dependency, sul- dysfunction (for example psychomotor retar- phite oxidase deficiency) dation, or earlier seizures) before the onset of Perinatal/postnatal (hypoglycaemia, lead toxicity) spasms with a known aetiology, and the cryp- Miscellaneous togenic group by the prior existence of Cerebral tumour (rarely) 616 Appleton

of spasms, but need to be excluded in the However, this belief may be ill founded as the

initial diagnostic evaluation. The association ultimate prognosis depends primarily on the Arch Dis Child: first published as 10.1136/adc.69.5.614 on 1 November 1993. Downloaded from with pertussis immunisation is unproved, con- underlying aetiology; additional prognostic troversial, and probably coincidental.2021 factors such as pre-existing neurodevelopment When all children with spasms are considered, status, age of onset, and response to treatment between 70 and 75% will be found to have a are also in fact dependent upon the aetiology specific aetiology (symptomatic cases.)'9 A and can therefore not be regarded as signifi- decade ago this number would have been cantly influencing the outcome per se. Most of closer to 60%5; the improved detection of an the studies have not assessed response to treat- underlying aetiology relates primarily to the ment (with whatever agent) in relation to advent of improved anatomical imaging whether the spasms were symptomatic or cryp- methods. Magnetic resonance imaging has togenic, or to the underlying aetiology. significantly increased the detection of subtle Although there is limited evidence that at least cerebral dysgeneses such as neuronal hetero- in the less common cryptogenic (and perhaps topias and other disorders of migration, to the particularly in idiopathic) cases, early treat- point that it is considered an essential investi- ment of spasms is associated with a more gation in the evaluation of children with favourable outcome,'3 27 this is not necessarily spasms, particularly if an initial computed cause and effect; infants with cryptogenic tomogram is normal.15 spasms may be treated earlier than patients A family history ofspasms is uncommon12 13; with symptomatic spasms because the spasms monozygotic twins may be both concordant are recognised earlier. The effect of prompt and discordant for infantile spasms.22 How- treatment is thus difficult to separate from that ever, a family history ofany epilepsy may occur due to the 'severity' of the disorder. Finally in 6-17% of cases.'2 No specific genetic there has been little attempt to assess the marker has yet been identified, and is unlikely influence of different periods of treatment lag; to be, but as already stated, infantile spasms most studies report lag periods of months may occur in a number of chromosomal/ which could arguably be associated with a genetic disorders including Aicardi's and less favourable response to treatment and out- Miller-Dieker syndromes and tuberous come. However, a treatment lag of between sclerosis. one and two weeks is unlikely to influence out- The pathogenesis ofinfantile spasms, as well come and could therefore be reasonably incor- as the regression in psychomotor development, porated into a placebo controlled protocol. remains unclear. It is generally believed that There is also an opinion that in some patients spasms are a non-specific response of an spasms may remit spontaneously.28 All ofthese immature brain to any insult; however certain factors militate against a rational evaluation of insults may be more likely to cause them. treatment of spasms. Although most insults are multifocal or diffuse However, it does appear to be true that in character, unifocal or even unilateral cere- spasms are resistant to many 'conventional' bral disorders (for example hemimegalen- antiepileptic drugs. Most success has been http://adc.bmj.com/ cephaly) 10 may cause spasms. The peak age of achieved using the benzodiazepines (particu- occurrence of West's syndrome coincides with larly nitrazepam),29 sodium valproate,30 31 the critical period of formation of dendritic corticosteroids and adrenocorticotrophic spines and myelination that may contribute to hormone (corticotrophin, ACTH).32 High the pathogenesis,23 for which there is some dose pyridoxine33 (vitamin B6), intravenous evidence.24 Additional hypotheses include a gammaglobulin,34 and a benzodiazepine- structural or functional disturbance in sub- carbamazepine cocktail35 have also been on September 23, 2021 by guest. Protected copyright. cortical (pontine or lenticular nuclei) neuro- reported to be effective. Vigabatrin, one of the transmitter (dopaminergic, serotonergic) newer antiepileptic drugs has been used to pathways; the observation of spasms in an treat spasms initially as 'add-on' treatment36 infant with hydranencephaly would support where it appears to be particularly successful in such a mechanism.25 Focal cerebral lesions symptomatic cases (usually the more resistant have also been implicated, and a cortical- type), and more recently, as monotherapy,37 subcortical interaction has been proposed to where the response has been rapid. explain how a focal lesion could be respon- ACTH was first used to treat spasms in the sible for producing generalised, symmetrical late 1950s38; a recent survey among child spasms.'8 The neuronal circuitry involved is neurologists in the United States indicated that interesting but perhaps simplistic and has of those who responded (only 45%), over 85% implications for both the medical and surgical used ACTH as the agent of first choice.39 This treatment of spasms. is in sharp contrast to at least one European country (Dr W F Arts, Professor A C B Peters, the Netherlands, personal communication) Treatment and the practice of this author where ACTH is It is in this area that most has been written rarely if ever prescribed; others advocate the about spasms, although largely on the basis of more selective use of ACTH, in cryptogenic or retrospective, anecdotal, and predominantly idiopathic cases. There is much anecdotal uncontrolled studies.26 Placebo controlled evidence but little scientific basis for the use of studies have been considered to be unjustified ACTH or corticosteroids; a recent report has in view of the belief that early diagnosis and valiantly attempted to explain and rationalise treatment determines the long term outcome the multiple aetiologies, efficacy of hormonal of late epilepsy and mental retardation. treatment, and even spontaneous resolution Infantile spasms 617

of infantile spasms.40 In addition there is cases and is often accompanied by motor and

considerable variation in the preparation, dose, behaviour difficulties. Sixty five per cent of Arch Dis Child: first published as 10.1136/adc.69.5.614 on 1 November 1993. Downloaded from and duration of treatment and relapses are children will develop chronic epilepsy, usually common, whether on or off treatment.26 41-45 the Lennox-Gastaut syndrome or cryptogenic More importantly these drugs are associated localisation related epilepsy (temporal lobe with with frequent, severe, and potentially fatal side complex partial seizures). effects. Drug related mortality may be as high as 5%.46 ACTH and corticosteroids have also not been shown conclusively to improve the Conclusion long term outcome of either the development The clinical description of spasms has not been of chronic epilepsy (which occurs in over 65% improved upon since 1841; the pathogenesis of children) or mental retardation.28 42 Sodium remains unclear and the classification has been valproate and nitrazepam are additional 'rec- revised to include symptomatic, cryptogenic, ommended' drugs but are also associated with and idiopathic cases. The development of variable success, relapses, and frequent side improved anatomical, and, to a lesser extent effects; nitrazepam is limited by the develop- functional, brain imaging has identified an ment of early tolerance and tachyphylaxis. increasing number of symptomatic cases. Until Vigabatrin appears to have considerably fewer recently, the recommended treatment has and less severe side effects than these recom- relied upon drugs that have an unknown mech- mended treatments and its efficacy may anism of action and serious side effects. have some recognised neuropharmacological Infantile spasms (particularly in the context basis47; however, there are relatively few data of West's syndrome) demands a more scien- on the use ofvigabatrin in infantile spasms and tifically precise study of its pathogenesis and specifically the long term effects of this drug. It treatment. This will almost certainly require a is likely that other, newer antiepileptic drugs multicentre, even multinational collaboration may also be used to treat spasms, particularly if in view ofthe relatively small numbers of child- the initial treatment is unsuccessful. ren involved. This practical difficulty should Finally, surgical resection could (? should) not preclude the evaluation of what is arguably be considered for those children with the most malignant epilepsy syndrome/seizure intractable spasms who are shown to have a type of childhood. In the interim antiepileptic unifocal cortical abnormality.'5 48 Although treatment should be used more rationally and this should be demonstrated by computed selectively on the basis of the electroclinical tomography/magnetic resonance imaging and pattern and underlying aetiology,'5 and in view an EEG it may, rarely, require functional of the recent and promising response to the (positron emission tomography) rather than newer antiepileptic drugs. anatomical imaging to identify the abnor- The author is grateful to Linda Finnegan for her patience and mality. assistance in the preparation of this review. The psychological impact of the spasms, http://adc.bmj.com/ associated mental retardation, and disturbed 1 Cowan LD, Hudson LS. The epidemiology and natural sleep on the families of these children may history of infantile spasms. J7 Child Neurol 1991; 6: 355-64. be immense, is always stressful and is fre- 2 West WJ. On a peculiar form ofinfantile . Lancet quently devastating. Support for these families 1841; i: 724-5. 3 Kurokawa T, Goya N, Fukuyama Y, Suzuki M, Seki T, is vital but it is usually inadequate and reliant Ohtahara S. West syndrome and Lennox-Gastaut syn- upon the relevant voluntary epilepsy support drome: a survey of natural history. Pediatrics 1980; 65: 81-8. associations. 4 Jeavons PM, Bower BD, Dimitrakoudi M. Long-term prog- nosis of 150 cases of 'West syndrome'. Epilepsia 1973; 14: on September 23, 2021 by guest. Protected copyright. 153-64. 5 Lombroso CT. A prospective study of infantile spasms: Prognosis clinical and therapeutic correlations. Epilepsia 1983; 24: 135-8. Persistence or recurrence of spasms may occur, 6 Riikonen R, Donner M. Incidence and aetiology of infantile particularly in those of symptomatic origin (for spasms from 1960 to 1976: a population study in Finland. Dev Med Child Neurol 1979; 21: 333-43. example cerebral dysgenesis, Aicardi's and 7 Aicardi J. Infantile spasms and related syndromes. In: Miller-Dieker syndromes). Aicardi J, ed. Epilepsy in children. New York: Raven Press, 1986: 17-38. Factors determining outcome are multiple 8 Donat JF, Wright FS. Clinical imitators of infantile spasms. and are ofvariable significance.7 12 27 In general, J Child Neurol 1992; 7: 395-9. 9 Gibbs FA, Gibbs EL. In: Gibbs FA, Gibbs EL, eds. Atlas of cryptogenic/idiopathic cases, a symmetrical . Volume 2: epilepsy. Cambridge, hypsarrhythmia, and rapid response to treat- Mass: Addison Wesley, 1952: 24. 10 Tjiam AT, Stefanko S, Schenk VWD, de Vlieger M. ment tend to predict a good prognosis; sympto- Infantile spasms associated with hemihypsarrhythmia and matic cases, a grossly asymmetrical EEG and hemimegalencephaly. Dev Med Child Neurol 1978; 20: 779-88. poor, or no, response to treatment are usually 11 Alvarez LA, Shinnar S, Moshe SL. Infantile spasms due to associated with a bad prognosis. The influence unilateral cerebral infarcts. Pediatrics 1987; 79: 1024-6. 12 Matsumoto A, Watanabe K, Negoro T, et al. Infantile of treatment lag is unclear; a delay in treatment spasms: etiological factors, clinical aspects and long-term of one to two weeks is unlikely to have a detri- prognosis in 200 cases. Eur J Pediatr 1981; 135: 239-44. 13 Riikonen R. A long-term follow-up study of 214 children mental effect on either the short, or long term with the syndrome of infantile spasms. Neuropediatrics outcome, whereas a delay of two or more 1982; 13: 14-23. 14 Hrachovy RA, Frost JD, Kellaway P. Sleep characteristics in months may arguably adversely affect both the infantile spasms. Neurology 1981; 31: 688-94. response to treatment and neurological status. 15 Commission on Pediatric Epilepsy of the International League Against Epilepsy. Workshop on infantile spasms. The overall prognosis of infants with West's Epilepsia 1992; 33: 195. syndrome is poor. Mortality is approximately 16 Commission on Classification and Terminology of the International League Against Epilepsy. Proposal for 5% (both drug and non-drug related), mental revised classification of epilepsies and epileptic syn- retardation occurs in 90% (severe in 70%/o) of dromes. Epilepsia 1989; 30: 389-99. 618 Appleton

17 Dulac 0, Plouin P, Jambaque I, Motte J. Spasmes infatiles 33 Blennow G, Starck L. High dose B treatment in infantile epileptiques benins. Revue d'Electroencephalographie et spasms. Neuropediatrics 1986; 17: 7-10. 16: 371-82. B, Dulac MJ, et al. Treatment Neurophysiologie Clinique 1986; 34 Echenne 0, Parayre-Chanez Arch Dis Child: first published as 10.1136/adc.69.5.614 on 1 November 1993. Downloaded from 18 Chugani HT, Shewmon DA, Sankar R, Chen BC, Phelps of infantile spasms with intravenous gamma-globulins. ME. Infantile spasms: II. Lenticular nuclei and brain stem Brain Dev 1991; 13: 313-9. activation on positron emission tomography. Ann Neurol 35 Tatzer E, Groh C, Muller R, Lischka A. Carbamazepine 1992; 31: 212-9. and benzodiazepines in combination - a possibility to 19 Dulac 0. Convulsions et epilepsies du nouveau ne et du improve the efficacy of treatment of patients with nourrisson. In: Arthuis M, Pinsard N, Ponsot G, eds. 'intractable' infantile spasms? Brain Dev 1987; 9: 415-7. Neurologie pediatrique, medecine-sciences. Paris: 36 Chiron C, Dulac 0, Beaumont D, Palacios L, Pajot N, Flammarion, 1990: 163-95. Mumford J. Therapeutic trial of vigabatrin in refractory 20 Melchior JC. Infantile spasms and early immunisation infantile spasms. J Child Neurol 1991; 6 (suppl 2): against whooping cough. Danish study from 1970 to S52-9. 1975. Arch Dis Child 1977; 52: 134-7. 37 Appleton RE, Montiel-Viesca F. Vigabatrin in infantile 21 Bellman MH, Ross EM, Miller DL. Infantile spasms and spasms; why add-on? Lancet 1993; 341: 962. pertussis immunisation. Lancet 1983; i: 1031-4. 38 Sorel L, Dusaucy-Bauloye A. A propos de 21 cas d'hypsar- 22 Elian M, McKenzie S. Infantile spasm - enigma of the rhythmia de Gibbs; son traitment spectculaire par 1'ACTH. genes. Seizure 1992; 1 (suppl A): P9/06. Acta Neurologica Psychiatrica Belgica 1958; 58: 130-1. 23 Adams RD, Victor M. Normal development and deviations 39 Bobele GB, Bodensteiner JB. The treatment of infantile in development of the nervous system. In: Adams RD, spasms by child neurologists in the United States: a 1991 Victor M, eds. Principles of neurology. 4th Ed. New York: survey. Ann Neurol 1991; 3: A461. McGraw-Hill, 1989: 457-61. 40 Baram TZ. Pathophysiology of massive infantile spasms: 24 Huttenlocher PR. Dendritic development of neocortex of perspective on the putative role of the brain adrenal axis. children with mental defect and infantile spasms. Ann Neurol 1993; 33: 231-6. Neurology 1974; 24: 203-10. 41 Singer WD, Rabe EF, Haller JS. The effect of ACTH 25 Neville BGR. The origin of infantile spasms: evidence from therapy upon infantile spasms. J Pediatr 1980; 96: a case of hydranencephaly. Dev Med Child Neurol 1972; 485-9. 14: 644-7. 42 Glaze DG, Hrachovy RA, Frost JD, Kellaway P, Zion TE. 26 Snead OC. Treatment of infantile spasms. Pediatr Neurol Prospective study of outcome of infants with infantile 1990; 6:147-50. spasms treated during controlled studies of ACTH and 27 Matsumoto A, Watanabe K, Negoro T, et al. Prognostic prednisone. Jf Pediatr 1988; 112: 389-96. factors of infantile spasms from the etiological viewpoint. 43 Parekh HU, Dunn DW. Infantile spasms. Ann Neurol 1990; Brain Dev 1981; 3: 361-4. 28: A473-4. 28 Hrachovy R, Glaze DG, Frost JD. A retrospective study of 44 Lopes I, Troncoso L, Troncoso M, Japaz 0, Novoa F. spontaneous remission and long-term outcome in patients Infantile spasms: brief and intermittent treatment with with infantile spasms. Epilepsia 1991; 32: 212-4. synthetic ACTH. Pediatr Neurol 1992; 8: A391. 29 Dreifuss F, Farwell J, Holmes G, et al. Infantile spasms; 45 Willig RP, Lagenstein I. Use of ACTH fragments in comparative trial of nitrazepam and corticotrophin. Arch children with infantile spasms. Neuropediatrics 1982; 13: Neurol 1986; 43: 1107-10. 55-8. 30 Siemes H, Shohr HL, Michael Th, Nau H. Therapy of 46 Riikonen R, Donner M. ACTH therapy in infantile spasms: infantile spasms with valproate; results of a prospective side effects. Arch Dis Child 1980; 55: 664-72. study. Epilepsia 1988; 29: 553-60. 47 Ring HA, Trimble MA, Costa DC, George MS, VerhoeffP, 31 Prats JM, Garaizar C, Rua MJ, Garcia-Nieto ML, Madoz P. Ell PJ. Effect ofvigabatrin on striatal dopamine receptors: Infantile spasms treated with high doses of sodium evidence in humans for interactions of GABA and valproate; initial response and follow-up. Dev Med Child dopamine systems. J' Neurol Neurosurg Psychiatry 1992; Neurol 1991; 33: 617-25. 55: 758-61. 32 Hrachovy RA, Frost JD, Kellaway P, Zion TE. Double- 48 Shields DW, Shewmon DA, Chugani HT, Peacock WJ. blind study of ACTH vs prednisone therapy in infantile Treatment of infantile spasms: medical or surgical? spasms. _J Pediatr 1983; 103: 641-5. Epilepsia 1992; 33 (suppl 4): S26-31. http://adc.bmj.com/ on September 23, 2021 by guest. Protected copyright.