Staff Writer

How do you develop a drug for a disease that has never been successfully treated? How do you convince the FDA that your drug works without validated endpoints?

These types of questions often plague biotechs attempting to blaze a trail against a complex and intractable disease like lupus, or like muscular dystrophy. And Seaside Therapeutics Inc. is no exception as it works to design a pivotal program for STX209 in fragile X syndrome.

Fragile X, a genetic mental disorder, is the most common known cause of autism. Yet while autism is estimated to occur in as many as one in 150 children, Fragile X is found in just one in 4,000 males and one in 6,000 to 8,000 females. Accordingly, there are no drugs specifically approved for fragile X, although the antipsychotics Abilify (aripiprazole, Bristol-Myers Squibb Co.) and Risperdal (risperidone, Johnson & Johnson) are approved for autism and often are used off-label.

Abilify and Risperdal gained their autism approvals by demonstrating improvement both on the Aberrant Behavior Checklist Irritability subscale (ABC-I) and the Clinical Global Impressions of Improvement scale (CGI-I). Thus when Seaside planned its Phase II trial of STX209 in fragile X, it selected ABC-I as the primary endpoint, reasoning that at least the FDA would be familiar with it.

"We had to pick a primary endpoint because if we got lucky and hit it, this could be a registrational trial rather than an exploratory trial," Seaside president and CEO Randall Carpenter explained. Yet the company had doubts that ABC-I was really the best way to evaluate fragile X patients. The fact was that nobody knew the best way to evaluate fragile X patients.

"We've been at the forefront of trying to develop novel endpoints for fragile X," Carpenter said, working closely with the FDA and the NIH. To that end, Seaside included just about every neurodevelopmental endpoint it could think of in its Phase II trial, pulling them from autism, Alzheimer's disease and other fields. The company also cross-checked and validated the endpoints by including multiple scales that measured the same outcome. And the trial used a crossover design, testing the feasibility of the endpoints in the placebo arm and their sensitivity in detecting the drug's effect in the treatment arm.

All in all, Seaside's 63-patient Phase II trial was the largest randomized, placebo-controlled trial in fragile X patients to date. And as the Cambridge, Mass.-based biotech expected, some of the endpoints worked and others didn't.

Data presented this month at the National Fragile X Foundation's 12th International Fragile X Conference showed that STX209 didn't meet the primary endpoint of significantly improving ABC-I. And while patients showed trends toward improvement on CGI-I and other global scales, the changes didn't hit statistical significance.

Yet in a subpopulation of 15 patients with severe social withdrawal at baseline, STX209 treatment significantly improved the Aberrant Behavior Checklist Social Withdrawal subscale (ABC-SW, p=0.05). The drug also improved all global measures, including CGI-I (p<0.01), Clinical Global Impressions of Severity (CGI-S, p<0.05), clinician preference (p<0.01) and parent/caregiver preference (p<0.01).

Carpenter said that if Seaside were to screen for social withdrawal rather than irritability at baseline, the company could enroll a new trial about as big as the Phase II in about the same amount of time. But before moving forward, Seaside plans to discuss its Phase II data with the FDA and with expert clinicians to get their feedback.

STX209, a derivative of the spasticity drug baclofen, has been developed "from bedside to bench," Carpenter said. Baclofen is sometimes used to relax the esophagus in patients with resistant gastroesophageal reflux disease, and its use in autism patients led to feedback that the drug was impacting other aspects of the disease. Seaside isolated the R-isomer of baclofen, eliminating side effects thought to be associated with the S-isomer, and began testing the drug for fragile X.

Yet Seaside also has a more traditional "bench to bedside" approach in the works. STX209 is a selective gamma-amino butyric acid type B (GABA-B) receptor agonist that is thought to indirectly inhibit the metabotropic glutamate receptor (mGluR) pathway implicated in fragile X. Seaside's second compound, STX107, works directly to antagonize mGluR5. Phase I trials are underway.

Also working in the fragile X field in start-up Afraxis Inc., which is in preclinicals with p21-activated kinase inhibitors designed to fix misshapen dendritic spines and form better synaptic connects in the brains of fragile X patients.

The autism field is only slightly more crowded. Seaside is testing STX209 in Phase II for autism, while Curemark LLC is in Phase III with an enzyme replacement approach. Meanwhile Addex Pharmaceuticals Inc. has a number of mGluR5 programs for schizophrenia, Parkinson's disease, anxiety, gastroesophageal reflux disease and migraines, but Seaside said it has intellectual property covering the autism and fragile X space.