mRECIST to predict survival in advanced hepatocellular carcinoma: Analysis of two randomised phase II trials comparing nintedanib vs sorafenib

Background & Aims: Response Evaluation Criteria in Solid Tumors (RECIST) has been shown to be a poor surrogate for survival benefit with targeted therapy in advanced hepatocellular carcinoma (HCC). Methods: We investigated whether response evaluated using modified RECIST (mRECIST) predicted overall survival (OS) using data from two Phase II clinical trials. Analyses were conducted on pooled data from 188 patients with advanced HCC treated with nintedanib or sorafenib, of whom 180 were evaluable for response. Cox regression and Kaplan-Meier survival analyses were used to explore differences in OS between the responders and non-responders according to RECIST 1.0 and mRECIST criteria. Multivariate Cox proportional hazards models, including factors known to influence survival, were used to compare survival according to RECIST and mRECIST response. Results: Discordance between RECIST and mRECIST evaluation was most common for assessment of partial response (12.2%) and stable disease (13.3%). OS was significantly longer in patients with response compared to patients without response—RECIST: hazard ratio (HR) 0.325 (95% confidence interval [CI] 0.130-0.815), P=.0122; mRECIST: HR 0.544 (95% CI 0.335-0.881), P=.0122. HRs from the multivariate models used to evaluate response by RECIST or by mRECIST as predictors of OS approached significance for RECIST (0.40 [95% CI 0.16–1.01]; P=.053) and for mRECIST (0.62 [95% CI 0.38–1.01]; P=.053). Conclusions: Response according to RECIST or mRECIST is associated with improved survival and should be considered as a valid endpoint for use in HCC clinical trials.