In 1841, West first described infantile spasms in his own son as “a peculiar form of infantile convulsions” (13). Later, their association with the sequelae of severe mental deficiency became clear. The EEG manifestations of infantile spasms—the high-voltage chaotic slowing with multifocal spikes and marked asynchrony—were identified in 1952 by Gibbs and Gibbs, who called the disorder hypsarrhythmia (14). Thus the creation of the definition of West syndrome: a triad of infantile spasms, hypsarrhythmia, and developmental regression or mental retardation.

Published studies on the efficacy of ACTH and corticosteroids in infantile spasms display considerable variability in design, complicating the establishment of research-based recommendations for optimal treatment. A few observations are generally accepted. The cumulative spontaneous remission rate over the first 12 months of seizures is about 25% (15). Seizures are almost always intractable to treatment with traditional anticonvulsant drugs. ACTH or oral steroid therapy should significantly reduce seizures in 50% to 75% of patients, but ACTH protocols, particularly those employing high-dose, long-acting synthetic formulations, are associated with a significantly high rate of side effects (16,17). The best chance for a treatment response is probably between 4 and 12 months of age in children who are neurologically normal when spasms that have no demonstrable cause begin (11,12,16,18, 19, 20). The ultimate prognosis is dismal for most patients and depends heavily on the cause of the spasms, preexisting neurologic and developmental status, the presence or absence of other seizures concomitant with the spasms, and the patient’s age at seizure onset (8,12,16,21, 22, 23, 24).

The controversies surrounding the treatment of infantile spasms outnumber the areas of agreement. Which is the most effective therapy: ACTH or steroids; other anticonvulsants such as vigabatrin, valproic acid, benzodiazepines, topiramate, or zonisamide; pyridoxine; some or all of these in combination; or yet another treatment? What is the impact of treatment with ACTH compared with steroids or anticonvulsants on long-term outcome in recurrence of spasms, evolution into other forms of intractable epilepsy,
and cognitive or behavioral function? Does treatment change the outcome for a patient with preexisting mental retardation and a structurally abnormal brain? What is the optimal dosage of these drugs, and how long should treatment last? Does the ultimate outcome depend on timing of treatment? Does the efficacy of ACTH depend on the formulation (natural vs. synthetic, sustained vs. short-acting)? Most of these questions remain unanswered.

The pathogenesis of infantile spasms and therefore the mechanism of action of ACTH and steroids in this condition are unknown, principally because an animal model for this disorder is lacking (25). Infantile spasms begin within a narrow range of ages, and various abnormalities are causally linked; however, infantile spasms may also occur without apparent cause. The effect of ACTH and corticosteroids is frequently all or nothing, and the steroid-induced seizure-free state is often sustainable even after drug withdrawal. These observations support the theory that the developing brain experiences a significant stress response to various etiologies that results in this age-dependent epileptic encephalopathy. Within this very narrow developmental window, ACTH and steroids may be able to reset the deranged homeostatic mechanisms of the brain, thereby reducing the convulsive tendency and improving the developmental trajectory.

Evidence suggests that the effects of ACTH on infantile spasms may be independent of steroidogenesis. Efficacy studies have demonstrated the superiority of ACTH to corticosteroids in treating infantile spasms and its efficacy in adrenal-suppressed patients (26, 27, 28, 29). Substantial physiologic and pharmacologic data indicate that ACTH has direct effects on brain function: increasing dendrite outsprouting in immature animals (30); stimulating myelination (31); regulating the synthesis, release, uptake, and metabolism of dopamine, norepinephrine, acetylcholine, serotonin, and γ-aminobutyric acid; regulating the binding at glutamatergic, serotonergic, muscarinic type 1, opiate, and dopaminergic receptors (32,33); and altering neuronal membrane lipid fluidity, permeability, and signal transduction (30). These neurobiologic effects can influence synaptic function and neurotransmission and may reside in fragments of the peptide devoid of corticotropic activity.

ACTH has a down-regulatory effect on corticotropin-releasing hormone (CRH), and low ACTH levels have been found in the cerebrospinal fluid of children with infantile spasms (34,35). CRH is an excitatory neuromodulator, with potent age-specific convulsant effects demonstrated in animal models (34,36). ACTH reduces CRH gene expression in specific brain regions, an effect demonstrated in the absence of adrenal steroids and achieved with use of only the 4 to 10 fragment of ACTH, which does not release adrenal steroids (28). Melanocortin-receptor antagonists blocked this effect, suggesting that these are the targets of ACTH action (28).

An hypothesis can therefore be generated, in which a stress response enhances CRH expression, leading to neuronal hyperexcitability and seizures. By suppressing CRH expression, possibly through the action of peptide fragments of ACTH on melanocortin receptors, hyperexcitability may be reduced, ameliorating infantile spasms. Clinical trials of ACTH fragments without activity on adrenals have yielded disappointing results (37,38), but these studies used the 4 to 9 rather than the 4 to 10 peptide fragment studied in animal models (28). The events that precipitate this proposed endocrine abnormality remain unclear.

Table 67.1 lists the formulations of ACTH. The biologic activity, expressed in international units (IU), permits a comparison of potency but represents the relative ability of the peptide to stimulate the adrenals and may not reflect its ability to affect brain function. The biologic activity of natural ACTH in the brain may differ from that of synthetic ACTH (12) as a result of ACTH fragments and possibly other pituitary hormones with neurobiologic activity in the brain that are present in the pituitary extracts. These compounds could enhance the therapeutic efficacy of natural ACTH (39). Any differences in the biologic effects of sustained ACTH levels provided by the depot formulations, as opposed to those of the short-acting preparations, are unknown. Given in high doses, however, long-acting depot preparations are associated with an increased incidence of severe side effects, including death from overwhelming infection (17).

Although most efficacy studies of ACTH and steroids are retrospective, an expanding body of prospective data is available (12,23,40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50). Most published literature supports the hypothesis that the natural ACTH 1 to 39 peptide (p-ACTH) is superior to oral steroids. In randomized, controlled trials, spasms ceased in 42% to 87% of children treated with ACTH, compared with 29% to 33% of children treated with prednisone (43, 44, 45, 46, 47). Respective relapse rates were 15% to 31% and 29% to 33%.

The most effective dose of p-ACTH for remission of spasms is, however, controversial. Notably, compared with prednisone, no major advantage was demonstrated by low doses of ACTH, whereas high doses were superior (43,44). High-dose p-ACTH (60 IU per day or 150 IU/m2 per day) has produced excellent short-term response rates (87% to 93%) in prospective studies (43,48). In the only randomized, prospective comparison of p-ACTH, however, Hrachovy and associates (45) found no difference between high-dose and low-dose therapy. A prospective
study of synthetic ACTH (47) by Yanagaki and colleagues compared very-low-dose (0.2 IU/kg per day) and low-dose (1 IU/kg per day) ACTH and found equivalent efficacy, with response and relapse rates comparable to those in other studies. Describing a stepwise increase in dosage, Heiskala and coworkers (50) demonstrated that while some patients can be controlled on lower doses of carboxymethyl-cellulose ACTH (3 IU/kg per day), others required high doses (12 IU/kg per day). Spasms were controlled initially in 65% of patients, but the rate of relapse was high.

A good response to ACTH appears to be associated with better long-term outcome (20). Some evidence supports high-dose ACTH over low-dose ACTH or oral steroids in cognitive outcome (9,11). Glaze and colleagues (40) found no difference between low-dose p-ACTH (20 to 30 IU per day) and prednisone (2 mg/kg per day). In a comparison of high-dose p-ACTH (110 IU/m² per day) and steroids, however, Lombroso (12) showed a higher rate of normal cognitive outcome in cryptogenic patients treated with ACTH than in those treated with prednisone alone (55% versus 17%). In a retrospective comparison of different ACTH dosage regimens (51), Ito and coworkers also noted a positive correlation between dose and developmental outcome.

Although data support high-dose ACTH as being more effective than low-dose ACTH, the precise dosage and duration are undetermined. The optimal dose may lie between 50 and 200 IU/m² per day. Doses of 400 IU/m² per day or higher are contraindicated because of a high incidence of life-threatening side effects (16,17,50).