OBJECTIVES

: The most common measure used in the outcome reporting of oncology clinical trials is the median survival, however it lacks statistical value. The most scientifically robust endpoint is the Hazard Ratio (HR), however its interpretation is difficult for patients, physicians and policy makers. There is a need for a tool which can serve both as a clear and statistically valid measure. The objective of this study was to evaluate alternative endpoints based on the mean benefit of novel oncology treatments.

METHODS

: We reviewed all FDA approvals for oncology agents between 2013-2017. We digitized survival curves and implemented statistical transformations to calculate for each trial the Restricted Mean Survival Time (RMST), and the mean survival using Weibull distribution. We assessed whether bonus points would be awarded for durable survival based on the ASCO framework. We calculated the improvement in mean survival in each trial and compared this to reported median survival benefit.

RESULTS

: The FDA approved 83 solid tumor indications for oncology agents between 2013-2017. 27 approvals were based on response rates while 49 approvals were based on survival endpoints, of which 27 were Progression Free Survival (PFS) and 22 were Overall Survival (OS). The average improvement in median OS/PFS was 4.6 months, while the average improvement in RMST was 3.6 months and the average improvement in mean survival using Weibull distribution was 6.1 months.

CONCLUSIONS

: Mean survival provides valuable information for different stakeholders – patients, physicians and policy-makers. Its inclusion should be considered in the reporting of clinical trials.