About a year after secretin failed in a Phase III trial for autism, Repligen Corp. reported mixed results for a Phase II trial in schizophrenia.

While secretin showed no statistically significant improvement in patient symptoms, it did show a trend in elevating patient mood and it offered some promise in treating a cognitive deficit symptom of the disease.

"If you look strictly at the endpoints that were defined for the study, it didn't meet statistical significance," said Laura Whitehouse, vice president of market development for Waltham, Mass.-based Repligen. "However, if you do some further analysis, there were some interesting new findings."

The initial Phase II results of secretin in refractory schizophrenia indicated the product did not produce a statistically significant improvement in patient symptoms as measured by the Clinical Global Impression (CGI) or the Positive and Negative Syndrome Scale (PANSS).

The trial evaluated secretin vs. a placebo over a two-week time frame. Patients were assessed at baseline and six other times during the treatment period. A total of 44 evaluable patients received placebo, secretin at 2 CU/kg or secretin at 5 CU/kg.

Results showed that the mean change in CGI in the secretin group was not significantly different than placebo. A CGI responder was someone who was "very much improved" or "much improved" in at least four of the six evaluations. The placebo group had two responders, while the low-dose and high-dose secretin groups had three and four responders, respectively.

"We found that there was a site effect for the placebo response and that one of the sites in the study had all of the placebo responders," Whitehouse told BioWorld Today. "Because it's an early study and a small number of patients, that sort of thing tends to push the results in one direction or another."

The PANSS measurement also did not show a significant treatment effect, but it did indicate a trend toward improvement on the dysphoric mood scale in the secretin group (p=0.06). The scale looks at a patient's levels of anxiety and tension.

In addition, clinicians noticed transient changes in the social interaction of patients at the time of the infusion.

There were no serious adverse events reported from the trial.

Repligen plans to gather more observations from the trial site before unblinding clinicians. A few patients participated in a cognitive testing paradigm within two hours of the first dose. Results support further study of secretin's impact on one of the cognitive deficits found in schizophrenic patients. The company is discussing the design for a follow-up study.

"Although it's a quantitative observation, it was only in a small number of patients," Whitehouse said. "So there's no way at this point to say whether it's real or not, but it certainly does warrant further investigation."

If the company moves into a follow-up study with secretin, the results should be available by the end of this year, she said.

The Phase II schizophrenia study was designed to confirm results from a Phase I study conducted by investigators at the University of North Carolina at Chapel Hill (UNC). In that trial, 22 patients with refractory schizophrenia received either a single intravenous dose of 1 or 2 CU/kg of secretin or a placebo. Several of the secretin-treated patients had marked, but transient, improvements in symptoms, while the placebo patients showed no improvements. A CGI responder analysis showed that three of 11 patients in the secretin group were responders, compared to none in the placebo group. However, the PANSS evaluation showed no differences between the treatment and placebo groups.

Repligen's Phase II trial would have confirmed that earlier study if not for the two placebo responders.

"If you take the placebo responders out, and you compare the result with the UNC result," Whitehouse said, "it's basically the same rate of response in the drug-treated group."

Repligen also is studying secretin, or RG1068, in a Phase I trial in patients with obsessive-compulsive disorder. The product failed in a Phase III trial last year, showing little difference from placebo when evaluated in 132 autistic children. (See BioWorld Today, Jan. 6, 2004.)

"We have no additional studies in autism planned," Whitehouse said. "That could change in the future."

In a separate release, Repligen reported results for its third quarter of fiscal 2005 ended Dec. 31. The company posted revenues of $2.3 million, an increase of 71 percent over the previous year's quarter. Gross profit was $1.2 million.

Repligen's stock (NASDAQ:RGEN) rose 6 cents on Friday to close at $2.11.