Seizures and Steroids

Arch Dis Child: first published as 10.1136/adc.60.2.94 on 1 February 1985. Downloaded from

Archives of Disease in Childhood, 1985, 60, 94-95 Annotations Seizures and steroids

For over 40 years we have known that some children nuria, subdural haematomas, and tuberose sclerosis with otherwise intractable fits improve with elicit this stereotypic response by precisely the same steroids,' and for over 30 years that a similar mechanisms. Once this heterogeneity of the condi- response may be obtained with adrenocortico- tion is accepted, it is easier to see that more trophic hormone.2 Since that time a considerable appropriate questions might be why some patients amount of work has centred on infantile spasms, respond to one drug and not another, and does this particularly on the relative efficacies of prednisone tell us anything about what is actually going on? The and adrenocorticotrophic hormone. For each report nature of the disease process, particularly in the asserting the superiority of one, it is possible to find cryptogenic cases, raises as big a research challenge another giving the opposite view. Close comparison as when West described its devastating course 143 is confounded by the fact that drugs, dosage, and years ago. duration of treatment have varied between the Work has begun in this direction. Serotonin has trials. One recent paper3 was unable to find any been of interest since the observation that 5 OH convincing difference of effect between the two. tryptophan induced infantile spasms in infants with Despite this, certain facts have emerged; more Downs syndrome.6 The same phenomenon can be patients respond to adrenocorticotrophic hormone reproduced in young guinea pigs, after which the than to prednisone;4 there are a few, however, who spasms can be abolished entirely by steroids or by will respond to the latter but not the former.5 The serotonin antagonists. Other myoclonic disorders in relation between the spasms and hypsarrhythmia is man including dyssinergia cerebellaris myoclonica not a close one as the electroencephalographic (Ramsey Hunt syndrome), paramyoclonus multi- changes may appear considerably later than the plex, and I dopa induced myoclonus respond to the onset of spasms. Recurrence of spasms after hormone serotonergic blocking agent, methysergide. To ex- treatment is rarely attended by reappearance of pect that infantile spasms are simply caused by a

hypsarrhythmia and neither the spasms nor the central serotonin excess, however, underestimates http://adc.bmj.com/ hypsarrhythmia by themselves are likely to affect the complexity of even the little we know of long term development. What conclusions can be neurotransmitter circuitry. Indeed postanoxic myo- drawn from all this work? clonus responds to 5 OH tryptophan. It may come as The first and most obvious conclusion is that no surprise, therefore, that a recent study reported adrenocorticotrophic hormone acts independently lower serotonin metabolite concentrations (5HIAA) of its effect on the adrenal glands. It is just in the cerebrospinal fluid of children with infantile conceivable that adrenocorticotrophic hormone acts spasms than in age matched controls.7 These con- via steroid metabolites intermediate on the biosyn- flicting pieces of evidence have been rather neatly on October 4, 2021 by guest. Protected copyright. thetic pathway; it is difficult otherwise to see how reconciled by suggesting that the primary abnor- some patients do better on adrenocorticotrophic mality may lie in supersensitive serotonin receptors hormone and a few on prednisone. This also makes that consequently suppress presynaptic serotonin the clinical observation that pharmacological doses turnover and release.7 A similarly recent and of adrenocorticotrophic hormone (80 IU) may be perplexing observation is that although, as might be more effective than lower doses (20 IU), more expected, children with infantile spasms have lower plausible. Perhaps, most convincingly, adrenocorti- concentrations of the inhibitory neurotransmitter cotrophic hormone can control fits in adrenal ot-aminobutyric acid in cerebrospinal fluid, the suppressed patients, including those concurrently on effect of adrenocorticotrophic hormone which prednisone, even in the presence of documented low abolished spasms was to reduce concentrations of serum cortisol concentrations. The most simple a-aminobutyric acid still further.8 alternative is that adrenocorticotrophic hormone It seems clear that we are unlikely to advance far has a direct effect on neurons or glial cells, or both. from an empirical approach until we understand A second conclusion must be that infantile spasms more about the site and mode of action of adreno- are associated with almost every condition affecting corticotrophic hormone in the brain. Confined not infants known to the paediatric neurologists. It is only to the anterior pituitary, it is also found along unlikely that conditions as disparate as phenylketo- with beta endorphin in cell bodies in the hypothala- 94 Arch Dis Child: first published as 10.1136/adc.60.2.94 on 1 February 1985. Downloaded from

Seizures and steroids 95 mus which project to the limbic system, thalamus, 120 U/day. I usually start with a dose of 40 U/day and brain stem. It is probable, therefore, that it acts and then increase it to 80 U/day if after three to four as a neurotransmitter in its own right in specific weeks there has been no response. I then wait a subcortical circuits. It would be of considerable further four weeks on this higher dose before interest to know if there are functional connections concluding that this treatment is ineffective. If it is between these cells and those of the ponto- effective, I continue it for three to six months. If a medullary raphe complex, which contains a high child is normal initially and responds rapidly to concentration of serotonergic neurons and which, it adrenocorticotrophic hormone, I would opt for the has been suggested, may modulate infantile spasms. shorter time. Prednisone is usually started at a One only has to consider the gap between these dosage of 2 mg/kg/day, and in the event of response, speculations and our lack of understanding of the this is reduced to a 10 to 20 mg daily maintenance mechanism for ensuing mental handicap to see how dose. Switching to alternate day steroids does not many unanswered questions remain. seem to result in loss of control. It must be realised Another steroid responsive movement disorder in that as none of these suggestions have been sub- children is that of the opsoclonus myoclonus or jected to controlled clinical trials, they arise from 'dancing eyes' syndrome. This is frequently but not rank empiricism. invariably a paraneoplastic effect associated with Finally, however, one empirical claim deserves neuroblastoma. Whether the relation between the consideration. For some time there has been an neuroblastoma and the brain rests upon a cross anecdotal impression in the published reports that immunological basis, on an undescribed neurotrans- early, urgent adrenocorticotrophic hormone treat- mitter secreted by the tumour, or some other ment of cryptogenic patients may preserve them mechanism is unknown, but once again this is a from dementia. The key to this approach lies in the condition that responds dramatically to steroids. early recognition of infantile spasms for what they Despite the increasing development of more are. It is not always realised that before classic effective anticonvulsants, it is worth remembering 'salaam' attacks develop, the fits may merely consist that older children with other types of seizure of a slight loss of postural tone, producing a disorder may also respond to adrenocorticotrophic momentary drooping of the trunk and dropping hormone in particular, or sometimes to steroids.5 forward of the head. Seemingly innocuous, they Adrenocorticotrophic hormone may often be effec- rarely draw an urgent response from the parents, tive in children with somewhat delayed development nor always from the general practitioner. Paediatri- who display a wide repertoire of seizure types- cians have an important educational role in this myoclonus, akinetic, absence, complex, and major situation. http://adc.bmj.com/ motor. It is in these children who have failed to respond adequately to a variety of anticonvulsant References McQuarie I, Andersson JA, Ziegler MR. Observations on 'cocktails' that either adrenocorticotrophic hormone antagonistic effects of posterior pituitary and cortico-adrenal or a ketogenic diet should be considered. While hormones in the epileptic subject. J Clin Endocrinol Metab each has its disadvantages, it should be borne in 1942;2:406-10. mind that apart from growth retardation, a Cushing- 2 Klein R, Livingston S. The effect of adreno-corticotrophic oid and the hormone in epilepsy. J Pediatr 1950;37:733-42. face, general obesity caused by former 3 Hrachovy RA, Frost JP, Kellaway P, Zion TE. Double blind on October 4, 2021 by guest. Protected copyright. treatment, more troublesome side effects such as study of ACTH vs. prednisone therapy in infantile spasms. hypertension and diabetes are most uncommon. It is J Pediatr 1983;103:641-5. sensible to keep an eye on electrolyte concentrations 4 Lacy JR, Penry JK. Infantile spasms. New York: Raven Press, as severe hypokalaemia occurs. 1976:79-105. occasionally It is 5 Snead OC, Benton JW, Myers GJ. ACTH and prednisone in also good to know that children who are neurologi- childhood seizure disorders. Neurology (Minneap) 1983;33: cally normal at the onset of seizures and who 966-70. respond to adrenocorticotrophic hormone rarely 6 Coleman M. Infantile spasms associated with 501O tryptophan administration in patients with Down's syndrome. Neurology relapse after a three month course. (Minneap) 1971;21:911-8. It is, however, all too frequent that children who 7 Silverstein F, Johnston MV. Cerebrospinal fluid metabolites in are already neurologically impaired and develop patients with infantile spasms. Neurology (Minneap) infantile spasms relapse on withdrawal of steroids. 1984;34: 102-5. Since one is 8 Ito M, Kikawa H, Taniguchi T. Cerebrospinal fluid GABA essentially buying time with a course of levels in children with infantile spasms. Neurology (Minneap) steroids in these patients, it may be argued that a 1984;34:235-8. more rational approach in the first instance is to R 0 ROBINSON attempt to control seizures with benzodiazepines or Guy's Hospital, sodium valproate. The recommended dose of adre- St Thomas's Street, nocorticotrophic hormone varies between 20 and London SE] 9RT