OBJECTIVES: The TOURMALINE-MM1 clinical trial assessed the efficacy of IXA+LEN+DEX vs. LEN+DEX in patients with relapsed and/or refractory multiple myeloma (RRMM; 1+ prior lines). The final data cut reflects a median follow-up of 85-months, with patients treated with IXA+LEN+DEX or LEN+DEX for a median of 18 and 16 cycles, respectively. 70.64% of patients received between 1- 12 lines of subsequent therapies. It is important to explore the potential for confounding from the subsequent therapies on overall survival (OS) outcomes given the importance of this outcome to patients, the clinical community and payers. The aim of our analysis was to estimate the relative efficacy in relation to OS after adjustment for subsequent therapies.

METHODS: The marginal structural model (MSM), the inverse probability of censoring weights (IPCW), and the rank preserving structural failure time model (RPSFTM) were utilized to attempt to adjust for the bias introduced from patients receiving subsequent therapies in the IXA+LEN+DEX (259/360) and LEN+DEX (251/362) treatment arms. Analyses were conducted for the intention-to-treat (ITT) population and the 2+ prior lines subgroup.

RESULTS: The hazard ratio for IXA+LEN+DEX vs. LEN+DEX using the unadjusted OS data was 0.939 (95% CI: 0.784 – 1.125) in the ITT population. Treatment switching methods estimated hazard ratios ranging from 0.68 to 0.89; all methods resulted in an improvement in the hazard ratio after adjustment for subsequent therapies. This trend was maintained in the subgroup analyses.

CONCLUSIONS: This study demonstrates the role of subsequent therapies in confounding the relative OS estimates in oncology clinical trials and highlights the importance of assessing the subsequent therapy profile when interpreting OS outcomes. This case study focusing on the TOURMALINE-MM1 trial indicates that addressing this confounding improves the relative OS across all methods. A key limitation to this analysis is the inability to test the assumptions underpinning each of the treatment switching methods.