Is progression-free survival a rational surrogate endpoint in front-line ovarian cancer clinical trials?

Selection of clinical trial endpoints is a critical decision for investigators. Choosing a poor endpoint can result in false discovery or thwart regulatory approval. Overall survival (OS) is the most objective clinical trial endpoint; however, use of this endpoint is impractical in some tumors. A prime example is in epithelial ovarian cancer where the case for a surrogate endpoint such as progression-free survival (PFS) is best exemplified in the front-line setting where post-progression survival (PPS) intervals are long and even exceed the primary PFS. Furthermore, when a pivotal trial leads to regulatory approval of a new agent, in later lines of therapy, extensive cross-over to the experimental agent and/or to other active agents may obscure any OS effect. OS proponents will cite the subjective nature of the assessment of PFS; nevertheless, PFS correlates well with OS and with true patient benefit when assessed in a rigorous manner. In a therapeutically responsive disease such as ovarian cancer, OS improvements are frequently derived from a cobbling together of PFS gains from multiple interventions. In fact, insisting upon OS as the primary endpoint would have denied patients the 12 new regulatory approvals in the past six-plus years in ovarian cancer, many of which have demonstrated PFS hazard ratios <0.50, a remarkable clinical benefit for patients.