befochs - no idea? I can't tell from your post whether the Twitter guy posted the actual p-value that they would need to achieve to terminate early, but I can't imagine why it would be a secret - they disclosed the number of events (235) that the interim look will be based on. Maybe they'll discuss the details at JMP on Wednesday? To be honest, I have no feel for how these stats play out - I was baffled on how the 2 month benefit in median PFS in the pancreatic trial (from 3.6 to 5.6 months) was highly significant (p=0.008), while the median OS benefit of roughly the same impact (from 6.9 to ~9) wasn't even close (p=0.80). I get that the control arm was tainted by the impact of crossovers and the Kaplan Meier curves merged/crossed over for the best survivors, but after eliminating the crossover patients, the OS impact still had a p-value of about 0.20? With respect to STS, the OS benefit from the PII trial was roughly double historical averages for dox alone (21.5 vs 9.5 ), so you would hope that if they achieve a similar impact in the PIII trial with 235 patients, that it would be highly statistically significant?
Nice try idiot. THLD does NOT leak info, hence the 5 pps. It should have been 9s and 10s long ago. The recent PR would have caused the share price to reflect the analyst's forecast. Doesn't matter anymore.