147P Population-adjusted indirect comparisons of repotrectinib and entrectinib in ROS1+ locally advanced or metastatic non-small cell lung cancer (NSCLC)

Repotrectinib demonstrated durable clinical activity among patients with ROS1 fusion-positive (ROS1+) locally advanced or metastatic NSCLC in the TRIDENT-1 non-randomized phase I/II clinical trial. However, no head-to-head evidence is available comparing repotrectinib against other approved ROS1 tyrosine kinase inhibitors (TKIs). The objective was to indirectly compare progression free survival (PFS) and objective response rate (ORR) between repotrectinib and entrectinib in patients with TKI-naïve ROS1+ locally advanced or metastatic NSCLC. Using evidence from a systematic literature review, unanchored matching adjusted indirect comparisons (MAIC) were used to estimate population-adjusted hazard ratios (HR) for PFS and odds ratios (OR) for ORR, between repotrectinib and entrectinib. TKI-naïve cohorts from registrational trials for repotrectinib (individual patient-level data [IPD]) and entrectinib (aggregate-level data and digitized Kaplan-Meier curves) formed the evidence base. The MAIC adjusted for imbalances in pre-specified baseline prognostic and effect modifying factors (age, sex, race, smoking, performance status, brain metastasis, prior lines of therapy). Weighted Cox (for PFS) and logistic (for ORR) regression models were fit. Supplementary analyses (SA) explored the impact of missing data and modeling assumptions on effect estimates. After MAIC weighting, patient characteristics were balanced between TRIDENT-1 (repotrectinib; n=71) and ALKA-372-001, STARTRK-1&-2 (entrectinib; n=168). The population-adjusted PFS HR favoured repotrectinib over entrectinib (0.52; 95% confidence interval [CI]: 0.31-0.87); SAs were consistent with the main findings (SA HR range, 0.35 to 0.53). The adjusted ORR OR numerically favoured repotrectinib (1.53; 95% CI: 0.64-3.67), though the 95% CI included the null value; SAs were consistent (SA OR range, 1.08 to 1.84). Estimates from this MAIC suggest that repotrectinib may be associated with about half the risk of progression or death compared with entrectinib in TKI-naïve patients with ROS1+ NSCLC. Comparisons with other ROS1 TKIs, and of other endpoints (e.g., safety) are underway.