AUSTIN, Texas — The primary endpoint was missed in a phase II trial of arbaclofen in autistic children and adolescents, but positive signs in secondary analyses still give hope for the drug, a researcher said here.
Among 150 patients ages 5 to 21 with autism spectrum disorder (ASD), no difference in change from baseline in lethargy and social withdrawal scores from the Autism Behavior Checklist (ABC) — the study’s primary outcome measure — was seen in those assigned to arbaclofen versus placebo (-5.4 points, SD 0.78 with arbaclofen; -6.0 SD 0.75 with placebo; P=0.518), reported Walter Kaufmann, MD, of Boston Children’s Hospital.
Kaufmann attributed the missed endpoint to an unexpectedly strong placebo response while presenting at the Child Neurology Society’s annual meeting
But arbaclofen, the R- enantiomer of the spasticity drug baclofen, did show some significant advantages over placebo on some secondary measures. These included severity scores on the Clinical Global Impressions (CGI) scale and, among certain subgroups, Vineland II socialization scores, Kaufmann said.
Arbaclofen’s developer, Seaside Therapeutics of Cambridge, Mass., said it planned to conduct another trial to confirm the results of these secondary analyses.
Diane Treadwell-Deering, MD, of Texas Children’s Hospital in Houston, told MedPage Today in an email that the missed primary endpoint was a major problem for the drug.
Nevertheless, the secondary outcomes “are enticing, and should lead to some study redesign, using different primary outcome measures that are more likely to capture the improvements suggested by the findings of this study,” said Treadwell-Deering, who is leading a trial of arbaclofen for social deficits in fragile X syndrome patients.
“Finding a robust outcome measure that adequately captures the improvements found in this study will be the challenge for the next trials,” she said.
Arbaclofen is believed to be responsible for most of racemic baclofen’s clinical effects, with a relative potency of 10- to 100-fold, Kaufmann said. The main effect is to stimulate GABA (gamma-aminobutyric acid) neurotransmission.
This mechanism has only recently emerged as a potentially beneficial approach to autism spectrum disorders, owing to findings from electroencephalographic studies in ASD patients suggesting that GABA modulation could help normalize electrical activity in the brain.
Also, findings in animal and clinical studies of arbaclofen in fragile X syndrome, which resembles ASD in some ways, supported a benefit for the drug.
Currently, social deficits in ASD are treated mainly with behavioral therapies, Treadwell-Deering said.
“While recommended, these therapies are seldom as successful as we would like; also, they are costly and time-consuming,” she said. “The importance of the discovery of a medication that successfully targeted social deficits in ASD could not be overemphasized.”
In the current trial, Kaufmann and colleagues enrolled 130 patients meeting DSM-IV criteria for autistic disorder, 17 with Asperger’s syndrome, and three diagnosed with pervasive developmental disorder “not elsewhere classified” in the DSM-IV system.
Half the patients were 5 to 11 years old and nearly 80% were male.
They were randomized 1:1 to arbaclofen or placebo for 12 weeks. Drug doses ranged from 5 to 15 mg two or three times daily, with doses titrated to maximize CGI improvement. Most patients in the 5 to 11 age range ended up at 10 mg three times daily or, in the placebo group, the equivalent number of pills; most older patients received 15 mg three times daily or the placebo equivalent.
Kaufmann said that, in a post-hoc analysis, the research group decided to examine whether outcomes were different in participants with relatively higher versus lower IQ, with 70 as the cutoff.
There was still no advantage or disadvantage for arbaclofen in the primary outcome of ABC lethargy and social withdrawal score. But Vineland II socialization and communication scores emerged as suggesting greater mean improvement from baseline with arbaclofen in the higher-IQ group, as follows:
- Communication: +4.0 (SD 11.23) with arbaclofen; +0.8 (SD 6.82) with placebo
- Socialization: +5.9 (SD 14.19) with arbaclofen; +2.6 (SD 9.19) with placebo.
Also, across both IQ strata, socialization scores improved markedly with arbaclofen relative to placebo when the analysis was limited to 93 patients who had the same raters at baseline and at the final evaluation. Those receiving arbaclofen had a mean increase of 7.1 points (SD 1.38) compared with an increase of 1.8 points (SD 1.26) with placebo (P=0.006).
A responder analysis showed that considerably more patients receiving arbaclofen with single-rater evaluations achieved improvements of at least 5, 7.5, and 10 points in Vineland II socialization scores. For example, 51% of this arbaclofen subgroup had increases of at least 5 points, compared with 28% of the corresponding placebo subgroup.
Kaufmann described arbaclofen as “generally well tolerated,” but withdrawals due to adverse events were somewhat more common with the drug (eight patients versus two in the placebo group). Moreover, one of the patients assigned to arbaclofen had serious suicidal ideation.
Other events prompting withdrawal included aggression, emotional upset, sleep problems, skin rash, dyskinesia, oculogyric crisis. One patient in the placebo group also withdrew because of suicidal ideation, and another because of insomnia.
Treadwell-Deering said the apparent outcome differences in the IQ strata “should be controlled for in future trials.”