Re: YMI phase-3 Tesmilifene data:
Here’s the summary from the JCO publication referenced by poorgradstudent:
>>
Although study accrual closed early after no difference in RR was observed at the planned interim analysis, the number of patients accrued was close to the planned total sample size. All patients were observed for toxicity, PFS, and OS. The final analysis was conducted as planned. For the primary outcome measure, PFS, there was no statistically significant difference, although a trend toward a positive hazard ratio (0.85; 95% CI, 0.67 to 1.09) was observed, with an apparent separation of PFS curves after 10 months of follow-up[emphasis added]; however, because there were only 73 patients still at risk in both arms, this observation should be interpreted cautiously… There was a significant prolongation in OS (hazard ratio, 0.66; 95% CI, 0.48 to 0.91.
<<
In other words, there was some benefit seen in PFS as measured by the hazard ratio (0.85), but it was not statistically significant at the p=.05 level. (The fact that the median PFS on active therapy was no better at all than the median on placebo appears to be the kind of anomaly I was talking about earlier.)
How to reconcile the modest increase in PFS with the stellar increase in OS? That’s the question that the second phase-3 trial will answer. However, it’s less startling that the stellar OS was seen with a “miss” in the PFS endpoint now that we know that there was a trend toward significant efficacy in PFS as measured by the 0.85 hazard ratio.
FWIW: on the Rodman & Renchaw webcast, the CEO said that it was the failure in response rate (not the miss in PFS) that caused BMY to pull out of the development program.
Market Data 
Markets 
AI
