Wilder,
GS guy got things wrong, imo.
75% is number of the pts in BSC group that crossed over to P.
My view: at week-8 60% on P progressed (40% response rate as SD and PR, best response rate) while min. 90% of BSC pts progressed (95% confidence interval for 33% difference in PFT at any measured point was reached). Now, because they got PFS average of 46% imply that for remaining 10% BSC and 40% P arm pts rate of progression was much slower for P than for BSC, up to at last (I guess) 20 weeks. It boils down, because BSC is no longer good comparator (most likely that 10% pts progressed very quickly after week-8), how did 40% of P arm fired after week-8? Here is true medical benefit, like 6 and 12 month survival, quality of life, other medications,...
Overall, P data for third line pts may look good if there is substantial benefit for more than 10% of the pts (those with PR).
AMGN/ABGX CC did not answered clearly on this because they were comparing P data with something that has no chance in this setting, So, even 10% P survival in 6 months will generate more than 50% reduction in progression rate at this time-point. Without K-M. curve we are blind.
For instance Tarceva has great K.-M. curve relative to BSC, still market penetration isn't much greater that for Irresa before T is introduced to market (so far). Further, P will compete with C for the same pts.
Miljenko |
