USE OF MULTIPLE SURROGATE ENDPOINTS IN ADVANCED COLORECTAL CANCER

OBJECTIVES: Regulatory and reimbursement agencies increasingly base their decisions on surrogate outcomes especially if they can be measured early compared to the final clinical endpoint. Progression free survival (PFS) has been investigated as a surrogate endpoint for overall survival (OS) in advanced/ metastatic colorectal cancer, however the outcome of these investigations has not been conclusive. In recent years, meta-analytic methods allowing for use of multiple surrogate endpoints jointly have been proposed. The aim of this research was to assess if PFS and response to treatment used jointly as surrogate endpoints to OS improve their predictive value. METHODS: Data were obtained from a systematic review of randomised control trials in advanced/metastatic colorectal cancer on effectiveness of pharmacological therapies (systemic chemotherapy (SC), anti-EGFR and antiangiogenic). Multivariate meta-analysis was used to model the association between treatment effect on both surrogate (PFS and response) and final (OS) endpoints, on all data and in subgroups of subclass therapy. RESULTS: Bivariate meta-analysis showed significant association between treatment effects on PFS and OS, which was only minimally improved in the trivariate analysis modelling the effect on two surrogate outcomes jointly. For subclass therapies, there was a moderate improvement in the association for SC, with increased precision by 9% of the regression coefficient between effects on OS and PFS, but not for the other two subclasses. Predicted treatment effects on OS were obtained with higher precision only for SC and antiangiogenics (reduction in uncertainty on average 1.7% and 2.4% respectively) when using both surrogates jointly. CONCLUSIONS: Joint use of two surrogate endpoints did not lead to much improvement in the association between treatment effects on surrogate and final endpoint but in some subclasses led to improved precision of the predicted effects on OS, likely due to the more accurate estimation of PFS when both surrogates were modelled jointly.