By Matthew Willett

Repligen Corp. said an initial analysis of results from its Phase II trial of human synthetic secretin in autistic children indicated the treatment produces statistically significant improvement in symptoms, at least according to parents.

The trial, however, failed to meet its primary endpoint, an improvement in symptoms measured by a psychologist's evaluation using the Childhood Autism Rating Scale (CARS). Repligen's stock (NASDAQ:RGEN) dropped 30 percent Wednesday on the news, or 81 cents, closing at $1.843.

The 136-child, multicenter, double-blind, placebo-controlled, randomized trial evaluated the safety and efficacy of secretin administered three times and at three-week intervals to 3- to 6-year-olds, and measured changes in symptoms over an eight-week period through parent and psychologist evaluations.

It found a statistically significant symptomatic improvement in the secretin-treated group compared to placebo (p=0.02) according to parental assessment. Among the subgroup of responders, 12 of 66 evaluable patients (18 percent) were rated by their parents at the highest level of improvement.

But the primary endpoint of the trial, a psychological examination using CARS, showed treated patients did not differ significantly from the placebo group. That news, which Wall Street saw as a failure, Repligen President and CEO Walter Herlihy saw as encouraging enough to move into Phase III studies.

He pointed out the trial was the first Phase II study of any therapeutic in autism, and that as such his company is still evaluating trial design.

"What we found was that there was a good deal of variation in autism symptomology," Herlihy told BioWorld Today. "Children's symptoms waxed and waned, and because of that the evaluations done by a psychologist in an one-hour session prior to dosing and a one-hour session after dosing were subject to a lot of variation and noise. That was the reason it did not yield a statistically significant measure of symptomatic change. By contrast, parental assessment over the eight-week period did, in a statistically significant way, demonstrate the activity of secretin."

Secretin is a hormone produced by a part of the intestine called the duodenum, which acts upon the pancreas to assist in digestion. The road toward marketing secretin as a treatment for autistic children hasn't been the smoothest for Repligen.

Repligen acquired the patent to use secretin for the treatment of autism from Victoria Beck, a New Hampshire mother who discovered the potential secretin-autism connection when her autistic son was treated with the hormone for severe gastrointestinal problems, and the Autism Research Institute of San Diego. (See BioWorld Today, March 11, 1999.)

In 1999 the company took a hit when the New England Journal of Medicine published findings from a study on secretin in autism sponsored by the National Institutes of Health's National Institute of Child Health and Human Development. The 56-patient study found a single dose of secretin failed to benefit children with autism. Repligen's stock fell 32 percent, closing at $3, though the company criticized both the scope and design of that trial. (See BioWorld Today, Dec. 9, 1999.)

Herlihy said Repligen will move secretin, which researchers believe works through stimulation of the vagus nerve, into Phase III testing upon FDA approval of the trial's design. He added the trial will be similar in design, but that the company will make changes to the psychological evaluation measurement standards. Parental assessment, he said, is a better evaluation of symptomatic improvement.

"The parents were blinded. They didn't know who got the drug and who got placebo, yet their assessments, though optimistic, weren't biased," he said. "The data recorded clearly, and indicated in a statistically significant way for the whole population, and with very much statistical significance in a subgroup, that the children who got secretin did significantly better than the children who got placebo." n