Updated and corrected: Oncothyreon conducted the second interim analysis of the Stimuvax trial after 75% of events (patients deaths) occurred, not after 75% of the patients in the trial died.
) -- This week's Biotech Stock Mailbag opens with an email from Orbee98. He writes, "I follow you on Twitter, so I know you've been negative on
, but I wonder if you have any thoughts about the Stimuvax trial now that it's moving to final analysis. Did you think it would fail already? The fact that it continues perhaps is a good sign that we could see positive results later, possibly?"
Possibly. Likely? I don't believe so. Oncothyreon's best shot at a positive result in the Stimuvax lung cancer trial was at this just-completed second interim analysis conducted after 75% of the events (539 patient deaths) required to conduct the analysis.
I don't want to bog down the Mailbag in biostatistics, but
(Oncothyreon's partner) designed the study with enough statistical power to demonstrate a Stimuvax benefit at the second interim analysis. It didn't, which is a problem as the study continues to the final analysis in 2013.
Oncothyreon shares are down 40% this week because supporters weren't just hoping for a win, they were expecting a win based on all their predictive modeling of patient enrollment and prior survival data from similar lung cancer patients.
I suspect that if we got a peek at the survival curves in this study right now, the Stimuvax curve is sitting right on top of the control-arm curve. No difference in survival or very small. With 75% of patient deaths already reported, Stimuvax is running out of time and patients to demonstrate a benefit.
What we do know for certain today is that the assumptions upon which Merck and Oncothyreon built the Stimuvax phase III trial were wrong. Control group lung cancer patients were expected to live around 20 months, which would allow Stimuvax to demonstrate a six-month survival advantage, or a 33% relative reduction in the risk of death.
Survival times in the control arm of the study are well beyond that now, which doesn't bode well for Stimuvax. In fact, it strongly suggests that Stimuvax is having no effect at all. Under the best circumstances, Stimuvax may reduce the relative risk of death by a small amount -- let's say 10% -- compared with the control arm. If that is the final outcome, it's going to be exceedingly difficult for the study as designed to reach statistical significance. A tiny survival benefit may also not be deemed clinically significant.