gms | German Medical Science

Background: Progression-free survival (PFS) is a frequently used endpoint in oncological clinical studies. In case of PFS, potential events are progression and death. Progressions are usually observed delayed as they can be diagnosed not before the next study visit. For this reason potential bias of treatment effect estimates for progression-free survival is a concern. For relative treatment effects as for example hazard ratios, bias-correcting methods have been proposed before. The goal is to derive unbiased absolute treatment effect measures like median survival times. Interval-censoring approaches were used in the past.

Methods: This paper proposes two new methods for correcting the assessment time bias of median progression-free survival when derived from parametric posterior distributions. Two different methods are proposed. The first one applies a modified likelihood function when deriving the a posteriori distribution. The second one directly approximates the unknown posterior distribution by bayesian computation.

Results: In particular the latter one leads to substantial reduction of the bias and thus allows a fair comparison of median PFS times between treatment groups. Both methods are compared with the interval-censoring and the naive right-censoring approach.

Conclusion:

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.