Table 3.

Prospective Comparative Clinical Trials of ACTH or Cosyntropin in Infantile Spasms

Study	Type of ACTH Used	No. of Patients*	Methodology	Results and Conclusions
Lux et al., 2005	Cosyntropin/tetracosactide	107	Randomized, open-label, multicenter study Hormone† (n = 55) vs. vigabatrin (n = 52) Median age = 14 months Primary outcome measure = development at 14 months as measured by the VABS Secondary outcome measure = presence of seizures and ADRs Follow-up = 14 months	Results: VABS scores between hormone and vigabatrin treatment did not differ significantly (78.6 vs. 77.5, respectively; P = 0.73). The proportion of infants who were seizure-free at the end of the study period was similar between groups (hormone, 22/55 [40%], vigabatrin, 19/52 [37%]; P = 0.71). In a subgroup analysis, mean VABS scores were significantly higher in patients with no known etiology of infantile spasms in the hormone group (88.2) compared with the vigabatrin (78.9) group (P = 0.025). Conclusion: Treatment of infantile spasms with hormone or vigabatrin appears to produce similar developmental outcomes and cessation of spasms at 14 months. Hormonal treatment may result in better developmental outcomes than vigabatrin in patients with an unknown cause of infantile spasms, but further research is warranted.
Lux et al., 2004	Cosyntropin/tetracosactide	107	Stratified randomization, open-label, multicenter study Prednisolone (n = 30) vs. tetracosactide (n = 25) vs. vigabatrin (n = 55), for analysis of prednisolone and tetracosactide grouped into hormone treatment arm (n = 55) Median age = 6 months (range, 4–9 months) Primary outcome measure = number of patients spasm-free at 13–14 days after randomization Secondary outcome measure = number of spasm-free consecutive days before day 14; number of patients without hypsarrhythmia Follow-up = 14 days	Results: Forty patients were free of spasms at 14 days. Significantly more patients were spasm-free at 14 days in the hormone group (prednisolone, 21; tetracosactide, 19; 40/55 = 73%) than in the vigabatrin group (28/52 = 54%) (P = 0.043). Significant differences in consecutive spasm-free days before day 14 were also detected (hormonal treatment, median days = 9; vigabatrin treatment, median days = 2.5; P = 0.038). Hypsarrhythmia resolved in significantly more patients in the hormone group (81%) than in the vigabatrin group (56%) (P = 0.024). Conclusion: At 14 days, hormonal treatment might be more effective than vigabatrin for stopping infantile spasms and hypsarrhythmia; however, given the small number of patients, it is difficult to draw conclusions about the comparative efficacy of prednisolone and tetracosactide.
Yanaguki et al., 1999	Cosyntropin/tetracosactide	26	Randomized, open-label study High-dose cosyntropin (n = 13) vs. low-dose cosyntropin (n = 12) High-dose (0.025 mg/kg per day) and low-dose cosyntropin (0.05 mg/kg per day) were given daily for 2 weeks and tapered to discontinuation in the following 2 weeks Mean age = 34.5 weeks Primary outcome measure = response to treatment (cessation of spasms and hypsarrhythmia on EEG) Secondary outcome measures = long-term seizure and developmental status; ADRs Follow-up = 4 weeks; 17 responders were followed for 1 year or more	Results: Response to treatment did not differ significantly between the high dose (11/13 = 85%) and the low dose (9/12 = 75%) (P > 0.05). Long-term developmental outcomes and the presence of seizures at 1 year or more of follow-up did not differ between groups; however, the detailed methods of the follow-up protocol were not stated, so the quality of follow-up results might be suspect. Brain shrinkage, as measured by CT, and sleepiness in the first week of treatment were significantly more pronounced in the high-dose group (P < 0.05). Conclusion: Patients taking high-dose or low-dose cosyntropin appear to have similar short-term responses to treatment. A low-dose cosyntropin regimen might be more tolerable than a high-dose regimen.
Vigevano et al., 1997	ACTH	42	Randomized, open-label, response-mediated crossover study ACTH (n = 19) vs. vigabatrin (n = 23); if there was no response in 20 days (phase 1) or if drug was not tolerated, patient was switched to other drug for another 20 days (phase 2) without washout Mean age = 5.5 months (range, 2–9 months) Primary outcome measure = cessation of spasms Follow-up = 40 days	Results: For the separate phases of the study, there was no difference in cessation of spasms with ACTH or vigabatrin, but a significant difference was detected when total response to treatment (phase 1 responders + phase 2 responders) was compared between ACTH (25 responders/35) and vigabatrin (13 responders/28) (P = 0.007). In general, more ACTH patients than vigabatrin patients experienced ADRs. Conclusion: ACTH and vigabatrin may result in similar rates of cessation of infantile spasms, but patients who initially do not respond to vigabatrin and who are switched to ACTH are more likely to experience cessation of spasms than the converse. Vigabatrin may be better tolerated than ACTH.
Baram et al., 1996	ACTH	29	Randomized, single-blind study ACTH (n = 15) vs. prednisone (n = 14) Mean age = 6.3 months (range, 2–21 months) Primary outcome measure = response to treatment; patients were considered responders if they ceased having spasms at 2 weeks and had no hypsarrhythmia for a full sleep-and-awake cycle up to 24 hours, as determined by EEG. Follow-up = 2 weeks	Results: Patients receiving ACTH responded to treatment (clinical and EEG response) significantly more often than those receiving prednisone (13/15 [86.6%] vs. 4/14 [28.6%], respectively; P = 0.002). Patients were observed beyond 2 weeks, but the study was confounded by the fact that many patients originally receiving prednisone were switched to ACTH. However, most patients (25/29) still had abnormal results at their last follow-up visit (range, 2–48 months). Conclusion: The likelihood of clinical and EEG resolution of infantile spasms appears greater with ACTH than with prednisone; however, it is difficult to make conclusions about differences in long-term outcomes because the study was not designed to detect such differences.
Hrachovy et al., 1994	ACTH	59	Randomized, blinded study High-dose ACTH (n = 26) vs. low-dose ACTH (n = 24) High-dose ACTH regimen started at 150 units/m2 tapered down to zero over a period of 3 months; low-dose ACTH regimen was 20 units each day for 2 weeks Age similar between groups, but no other information provided Primary outcome measure = response to treatment (cessation of spasms and hypsarrhythmia on EEG) Follow-up = 1.5–3 months	Results: The difference in response to treatment between high and low-dose ACTH was not statistically significant (50% and 58%, respectively). There was also no statistical difference in relapse between the two groups (15% and 21%, respectively). Hypertension occurred more frequently in the high-dose group than in the low-dose group (31% and 4%, respectively; P = 0.025). Conclusion: Patients with infantile spasms appear to respond at similar rates to high-dose and low-dose regimens of ACTH; however, high-dose ACTH patients may experience hypertension more frequently than low-dose ACTH patients.
Dreifuss et al., 1986	ACTH	52	Randomized, double-blind, multicenter study ACTH (n = 25) vs. nitrazepam (n = 27) Mean age = 8.4 months (range, 2 to 23 months) Primary outcome measure = spasm frequency in 24 hours after 4 weeks of treatment Secondary outcome measure = correlation of total sleep time to change in spasm frequency Follow-up = 4 weeks	Results: There was no statistical difference in frequency of spasms between ACTH and nitrazepam (mean ± SEM = 89.7 ± 23.6, mean + SEM = 122.1 ± 20.8, respectively; P value not reported). Total sleep time was significantly correlated with change in spasm frequency (rs = −0.61, P ≤ 0.005) in patients receiving ACTH but not in patients receiving nitrazepam (rs = −0.10, P ≤ 0.63). More ACTH patients had fluctuations in systolic blood pressure compared with the nitrazepam group. Conclusion: Patients taking ACTH or nitrazepam appear to have similar frequency of spasms at 4 weeks; however, if the study had enrolled more patients, a statistically significant difference might have been detected favoring ACTH. The reduction in spasms attributed to ACTH is significantly associated with an increase in total sleep time, but the same association cannot be said for nitrazepam. Because nitrazepam is not available in the U.S., the comparative data may be less relevant than other comparative studies with ACTH.
Hrachovy et al., 1983	ACTH	24	Randomized, double-blind, response-mediated crossover study with a 1-week washout period ACTH (n = 12) vs. prednisone (n = 12), received clonazepam if no response to ACTH or prednisone Age range = 3.5 to 24 months Primary outcome measure = response to treatment (cessation of spasms and hypsarrhythmia on EEG in 24 hours at 8 weeks) Secondary outcome measure = response to treatment after crossover at 17 weeks Follow-up = 25 weeks	Results: There were no statistical differences in response between ACTH and prednisone at 8 weeks (5/12 [42%] and 4/12 [33%], respectively) or after crossover at 17 weeks (4/8 [50%] and 3/7 [43%], respectively). Adverse events appeared at similar rates between the groups. Conclusion: Patients with infantile spasms may experience similar responses to treatment between ACTH and prednisone. If they do not respond to the first treatment, they might respond to the other treatment.

^*The number of patients might differ from that reported in methodology as a result of pre-randomization exclusions.

^†Hormonal treatment = tetracosactide and prednisolone.

ADR = adverse drug reaction; CT = computed tomography; EEG = encephalography; VABS = Vineland Adaptive Behavior Scale.

Data from References 28through ³⁵.