Randomized trial of vigabatrin in patients with infantile spasms

Study design.

The design was a 2-week, randomized, single-blind, multicenter study with a long-term, open-label, dose-ranging follow-up study. After a baseline evaluation between days −7 and −3, study visits were scheduled for weeks 0 (randomization) and 2. During long-term follow-up, efficacy data were collected at week 4 and months 2 and 3, although patients could remain in the study for as long as 3 years. A final visit occurred 2 weeks after treatment with vigabatrin was discontinued.

Patients.

Patients were eligible for the study if they were younger than 2 years of age, weighed at least 3.5 kg, and had a diagnosed duration of infantile spasms of no more than 3 months. Each study site had Institutional Review Board approval for the research, and enrolled patients had parental written informed consent. Diagnostic criteria included infantile spasms and an EEG pattern typical of hypsarrhythmia or modified hypsarrhythmia or multifocal spike wave discharges or a video EEG recording capturing an event confirming the diagnosis of infantile spasms electroencephalographically. Eligible patients were receiving stable doses of antiepilepsy medication, if applicable; had unremarkable ophthalmologic examinations or examinations reflecting static problems; were unresponsive to a 50-mg IV bolus of pyridoxine, when appropriate; and had MR images or CT scans within the previous 12 months confirming the absence of progressive lesions.

Patients excluded from the study were those with treatable or progressive causes of seizures (e.g., metabolic disturbance, neoplasm, phenylketonuria, nonketotic hyperglycinemia, or active infection); aspartate aminotransferase, alanine aminotransferase, or total bilirubin concentration greater than twice the upper limit of normal; currently active significant medical disorders that would interfere with the safe completion of the study (e.g., renal, endocrine, hepatic, respiratory, or cardiovascular problems); clinically diagnosed Lennox–Gastaut syndrome; or a history of generalized tonic-clonic seizures. Patients also were excluded if they had received an investigational drug within 30 days of enrollment (1-week washout for ganaxolone); had been or were currently taking corticosteroids, ACTH, or valproic acid (brief exposure to low-dose corticosteroids for inflammatory conditions or replacement therapy was acceptable but required a 1-week washout); had been taking felbamate within 60 days of enrollment; or were receiving more than two standard antiepilepsy drugs to treat noninfantile spasm seizures. Caregiver inability to reliably record seizures or recall adverse events or history of poor medication adherence also resulted in patient exclusion.

Treatment.

Patients were randomly assigned to receive low-dose (18–36 mg/kg/day) or high-dose (100–148 mg/kg/day) vigabatrin. As outlined in table 1, the dose was based on the patient’s weight and, in the high-dose group, was titrated over the first 7 days of the study. Vigabatrin was given twice daily unless the total dose was less than 250 mg per day in which case once daily dosing was permitted. After 14 days of therapy, all patients entered the flexible dosing, follow-up phase of the study. At that time, patients in the low-dose group who were still having infantile spasms began the high-dose regimen. Patients in either group who continued to have infantile spasms after 7 days of the full dose of vigabatrin specified for their weight could have the dose increased at a rate not to exceed 25 to 50 mg/kg/day on a weekly basis until the patient became free of infantile spasms, a maximum tolerated dose was reached, or a maximum allowable dose of 200 mg/kg/day was achieved. An investigator also could decrease or discontinue vigabatrin therapy at any time during the flexible dosing phase, although abrupt discontinuation was avoided. Dosages of concomitant antiepilepsy drugs could not be changed during the first 21 days of the study; subsequently, adjustment, withdrawal, or addition of such drugs could be undertaken at the discretion of an investigator.

Vigabatrin was supplied as 500 mg tablets in open-label bottles. Caregivers were masked to whether their children were assigned to the high- or low-dose group.

Efficacy evaluations.

Caregivers kept a seizure diary in which they maintained a daily record of the number of seizures (clusters per day for infantile spasms) that occurred during the periods between study visits; adverse events and concurrent medications also were recorded. At each visit, a study coordinator reviewed the diary and entered data in a case report form. In patients reported to have been free of infantile spasms for 7 consecutive days beginning within the first 14 days of therapy, an 8-hour closed circuit television EEG recording that included at least one sleep–wake cycle was performed to detect evidence of infantile spasms or hypsarrhythmia. The test was conducted within 3 days of the end of the 7-day period during which the patient was deemed to be free of infantile spasms by caregiver observation (i.e., between 7 and 24 days of starting treatment).

The primary efficacy variable was the proportion of patients in each treatment group who were free of infantile spasms for 7 consecutive days beginning within the first 14 days of vigabatrin therapy (treatment responders). Treatment responders must have remained free of infantile spasms according to caregiver response to direct questioning regarding spasm frequency and have had no indication of infantile spasms or hypsarrhythmia during 8 hours of closed circuit television EEG recording. The secondary efficacy variable was time to response to therapy during the entire 3-month study period. Time to response was defined as the time from the beginning of therapy to the seventh day of a 7-day consecutive period of freedom from spasms and was based solely on caregiver assessment of infantile spasm frequency in response to direct questioning regarding frequency (weeks 2 and 4) or as recorded in the seizure diary (months 2 and 3). A patient who relapsed was one who responded during the 3-month period and subsequently had at least one episode of infantile spasms by the end of month 3.

Safety evaluations.

Safety was evaluated on the basis of physical and neurologic examinations, vital signs, ophthalmologic examinations, clinical laboratory tests, and adverse events. Investigators assessed whether adverse events were related or unrelated to therapy vigabatrin.

Statistical analyses.

The primary efficacy analysis compared proportions of treatment responders (by caregiver report and closed circuit television EEG confirmation) in the two treatment groups using the Mantel–Haenszel χ² statistic.10 Kaplan–Meier10 estimates of time to response distributions (by caregiver report only) during the 3-month study were computed for both dose groups combined and for the low- and high-dose groups separately; the significance of differences between dose groups was assessed using the log-rank test. All analyses were intent-to-treat and were done using SAS software, version 6.12 (SAS Institute, Cary, NC).

The cause of infantile spasms was classified as dysgenetic, tuberous sclerosis, postnatal (hypoxic-ischemic encephalopathy, infections, hemorrhage or trauma, and metabolic or toxic), or idiopathic or cryptogenic. Proportions of treatment responders stratified by cause were compared by the Cochran–Mantel–Haenszel test for general association; the Mantel–Haenszel χ² statistic was used to compare proportions of treatment responders among those with tuberous sclerosis versus patients with other causes. Distributions of times to response were compared between patients grouped by cause using Kaplan–Meier survival analyses and log-rank tests.10

Patient characteristics.

A total of 179 patients were randomized to receive high-dose or low-dose vigabatrin (figure 1); 167 patients were exposed to study medication and were able to be evaluated for safety. Of the 142 patients with complete data for efficacy analyses, 75 received low-dose vigabatrin and 67 received high-dose therapy. Demographic characteristics for patients who were able to be evaluated for efficacy are summarized in table 2. There were no significant differences between treatment groups with respect to any measured characteristic.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Trial profile. See table 1 foe vigabatrin dosage. HD = high-dose vigabatrin; LD = low-dose vigabatrin.

Protocol violations.

Six of the 142 patients (three in each group) started to be treated with an incorrect dose of vigabatrin but were analyzed as randomized in intent-to-treat efficacy analyses.

Primary efficacy variable.

Overall, 32 of 142 patients were treatment responders, including 8 of 75 patients in the low-dose group and 24 of 67 in the high-dose group (p < 0.001) (figure 2). More patients with tuberous sclerosis responded to therapy than did those with other conditions (13/25 vs. 19/117; p < 0.001); only 7 of 72 patients in the postnatal or dysgenetic cause groups responded within the first 14 days.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Treatment responders by vigabatrin dose and etiology. Response beginning within the first 14 days of vigabatrin therapy determined by cessation of infantile spasms for seven consecutive days and results of 8-hour EEG showing no evidence of infantile spasms or hypsarrhythmia. p < 0.001 for difference between low and high dosages, p < 0.001 for difference between tuberous sclerosis vs all other etiologies combined.

Secondary efficacy variable.

Vigabatrin dosage at the time of response ranged from 16.9 to 198.9 mg/kg/day (median, 106.4 mg/kg; mean, 95.3 mg/kg). Eight responders (9%) received doses of at least 150 mg/kg/day.

The Kaplan–Meier plot of the estimated nonresponse rate by duration of vigabatrin therapy (figure 3) shows that response increased dramatically after approximately 2 weeks of vigabatrin therapy. The estimated proportion of responders increased from 8% at 2 weeks, to 42% at 4 weeks, 55% at 2 months, and 65% at 3 months. Time to response was shorter in patients receiving high-dose vigabatrin than in those receiving low doses (p = 0.04) (figure 4), and among patients with tuberous sclerosis compared with those in other groups (p < 0.001) (figure 5). In fact, within 3 months of initiating vigabatrin therapy, 23 of 25 patients with tuberous sclerosis were free of infantile spasms by caregiver report. Excluding two patients with tuberous sclerosis who discontinued therapy before the end of the 3 months, response rates among these patients were 19 of 23 and 23 of 23 at 37 and 71 days, respectively. Patients with tuberous sclerosis in the high-dose vigabatrin group showed a slightly greater early response to therapy; however, by day 71, response rates were approximately the same (i.e., 14 of 15 patients started on high-dose vigabatrin vs. 10 of 10 patients started on low-dose therapy [p = 0.2]). Of the 87 patients who responded to vigabatrin therapy within 3 months, 14 (16%) relapsed.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3. Percent nonresponders overall. Response determined by caregiver report.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4. Percent nonresponders among patients receiving high-dose (dotted line) vs low-dose (dashed line) vigabatrin. Response determined by caregiver report. p < 0.04 for difference between response curves.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5. Percent nonresponders among patients with tuberous sclerosis (dashed line) vs other etiologies (dotted line). Response determined by caregiver report. p < 0.001 for difference between response curves.

Safety.

In the 167 patients exposed to vigabatrin and able to be evaluated for safety, 150 (90%) had at least one adverse event. Most events were not deemed to be related to vigabatrin therapy. The most common treatment-related adverse events were sedation (42/167), insomnia (15/167), and irritability (15/167). No dose-related adverse events were identified, and adverse event rates were similar in the low- and high-dose groups. Fifty-six patients reported serious adverse events; 53 patients had a total of 123 hospitalizations, 7 of whom were admitted to the hospital for reasons possibly related to vigabatrin. Two patients experienced cardiorespiratory arrest deemed unrelated to vigabatrin. Nine patients discontinued vigabatrin because of an adverse event, and two of these patients died. In one case, the cause of death could not be determined, whereas the other death was attributed to hemorrhage secondary to congenital pulmonary angiomatosis and was judged to be unrelated to vigabatrin. Clinical laboratory assessments showed no serious adverse events. A total of 103 patients had follow-up ophthalmologic examinations; six had clinically significant changes, including detection of cortical blindness, esotropia of the left eye and decreased visual acuity, findings consistent with retinopathy of prematurity, detection of homonymous hemianopia, improvement in cortical visual loss with exotropia, and change from poor visual processing at baseline to normal visual processing at follow-up.

Members of the US Infantile Spasms Vigabatrin Study Group:

Martina Bebin, MD, University of Alabama at Birmingham, Birmingham, AL; Joan A. Conry, MD, Children’s National Medical Center, Washington, DC; Patricia K. Crumrine, MD, Children’s Hospital of Pittsburgh, Pittsburgh, PA; Larry E. Dollar, RPh, MS, Aventis Pharmaceuticals, Inc., Kansas City, MO; Roy D. Elterman, MD, Medical City Dallas Hospital, Dallas, TX; Charles W. Gordetzky, MD, PhD, Quintiles, Inc., Kansas City, MO; Christina M. Gullion, PhD, Medical City Dallas Hospital, Dallas, TX; Karen A. Mansfield, Medical City Dallas Hospital, Dallas, TX; Warren A. Marks, MD, Fort Worth, TX; Wendy Mitchell, MD, Children’s Hospital Los Angeles, Los Angeles, CA; Jo Anne Nakagawa, Mattel Children’s Hospital at UCLA, Los Angeles, CA; Sue Ruckh, RPh, MBA, Quintiles, Inc., Kansas City, MO; W. Donald Shields, MD, Mattel Children’s Hospital at UCLA, Los Angeles, CA; Edwin Trevathan, MD, St. Louis Children’s Hospital, St. Louis, MO; Elaine Wyllie, MD, The Cleveland Clinic Foundation, Cleveland, OH.