Replies to post #2090 on Dendreon Corp fka DNDNQ

[p3analyze]

9901 Cox model was described in the JCO 9901 publication - I think if this is the model adopted without knowledge of 9902a dataset, then whether it was prospectively specified for 9901 or not, we should be fine, because the 9902a result validated the model.

"To test the robustness of the survival findings, the individual effect of 20 potentially important prognostic factors that have been described in the literature was evaluated. Using exploratory Cox regression models, each one having treatment and a single covariate as main effects, the treatment effect remained significant. Covariates that were significant at the .05 level were then selected to build a predictive multicovariate Cox regression survival model. A backwards stepwise selection method (P = .05 for entry and P = .10 for removal, likelihood ratio test) was used to identify significant prognostic factors and their impact on treatment effect."

[p3analyze]

Walldiver, what were those two covariates that remained stat significant? Thanks.

"The fact that the 9902A Cox regression was done using the same five final prognostic factors that were stat sig for 9901, even though only two of them were stat sig individually, indicates to me that 9902A's Cox analysis was also on the up and up. "