Population-Adjusted Indirect Treatment Comparisons of Repotrectinib Among Patients with ROS1+ NSCLC

BACKGROUND: locally advanced or metastatic non-small-cell lung cancer (aNSCLC). METHODS: Using evidence from a systematic literature review, unanchored matching-adjusted indirect comparisons (MAICs) were used to estimate population-adjusted hazard ratios (HRs) for PFS and DoR and odds ratios (ORs) for ORR for repotrectinib vs. crizotinib and vs. entrectinib among patients with TKI-naïve aNSCLC. The MAICs were adjusted for imbalances in baseline patient characteristics that were pre-specified as being prognostic or predictive of treatment effects. Weighted Cox (for PFS and DoR) and logistic (for ORR) regression models were fit. Supplementary analyses (SAs) explored the impact of missing data and modeling assumptions on effect estimates. RESULTS: The evidence base was formed by TRIDENT-1 EXP-1 (repotrectinib; N = 71), a pooled set of five trials involving crizotinib (N = 273), and the pooled ALKA-372-001/STARTRK-1 and -2 trials (entrectinib; N = 168). After population adjustment, repotrectinib was associated with statistically significant improvements in PFS relative to crizotinib (HR = 0.44; 95% confidence interval [CI]: 0.29, 0.67) and entrectinib (HR = 0.57; 95% CI: 0.36, 0.91). Differences in ORR and DoR were not statistically significant but numerically favored repotrectinib. SAs were consistent with the main analyses across all comparisons. CONCLUSIONS: aNSCLC.