Surrogacy of Recurrence-free Survival for Overall Survival as an Endpoint of Clinical Trials of Perioperative Adjuvant Therapy in Hepatobiliary-pancreatic Cancers: A Retrospective Study and Meta-analysis.

ABSTRACT

OBJECTIVE: To assess the correlation between recurrence-free survival (RFS) and overall survival (OS) in the hepato-biliary-pancreatic (HBP) surgical setting in order to validate RFS as a surrogate endpoint. SUMMARY BACKGROUND DATA Reliable surrogate endpoints for OS are still limited in the field of HBP surgery. METHODS: We analyzed patients who underwent curative resection for HBP disease (986 patients with pancreatic ductal adenocarcinoma [PDAC], 1168 with biliary tract cancer [BTC], 1043 with hepatocellular carcinoma [HCC], and 1071 with colorectal liver metastasis [CRLM]) from September 2002 to June 2022. We also conducted meta-analyses of randomized controlled trials of neoadjuvant or adjuvant therapy to validate the surrogacy in PDAC and BTC. RESULTS: Correlation coefficients between RFS and OS were low for HCC (ρ = 0.67) and CRLM (ρ = 0.53) but strong for PDAC (ρ = 0.80) and BTC (ρ = 0.75). In a landmark analysis, the concordance rates between survival or death at 5 years postoperatively and the presence or absence of recurrence at each time point (1, 2, 3, and 4 y) were 50%, 70%, 74%, and 77% for PDAC and 54%, 67%, 73%, and 78% for BTC, respectively, both increasing and reaching a plateau at 3 years. In a meta-analysis, the correlation coefficients for the RFS hazard ratio and OS hazard ratio in PDAC and BTC were ρ = 0.88 (P < 0.001) and ρ = 0.87 (P < 0.001), respectively. CONCLUSION: Three-year RFS can be a reliable surrogate endpoint for OS in clinical trials of neoadjuvant or adjuvant therapy for PDAC and BTC.