Integrated exposure–response analysis of efficacy and safety of lurbinectedin to support the dose regimen in small-cell lung cancer

Abstract

Purpose: These exposure–response (E–R) analyses integrated lurbinectedin effects on key efficacy and safety variables in relapsed SCLC to determine the adequacy of the dose regimen of 3.2 mg/m2 1-h intravenous infusion every 3 weeks (q3wk). Methods: Logistic models and Cox regression analyses were applied to correlate lurbinectedin exposure metrics (AUCtot and AUCu) with efficacy and safety endpoints: objective response rate (ORR) and overall survival (OS) in SCLC patients (n = 99) treated in study B-005 with 3.2 mg/m2 q3wk, and incidence of grade 4 (G4) neutropenia and grade 3–4 (G ≥ 3) thrombocytopenia in a pool of cancer patients from single-agent phase I to III studies (n = 692) treated at a wide range of doses. A clinical utility index was used to assess the appropriateness of the selected dose. Results: Effect of lurbinectedin AUCu on ORR best fitted to a sigmoid-maximal response (Emax) logistic model, where Emax was dependent on chemotherapy-free interval (CTFI). Cox regression analysis with OS found relationships with both CTFI and AUCu. An Emax logistic model for G4 neutropenia and a linear logistic model for G ≥ 3 thrombocytopenia, which retained platelets and albumin at baseline and body surface area, best fitted to AUCtot and AUCu. AUCu between approximately 1000 and 1700 ng·h/L provided the best benefit/risk ratio, and the dose of 3.2 mg/m2 provided median AUCu of 1400 ng·h/L, thus maximizing the proportion of patients within that lurbinectedin target exposure range. Conclusions: The relationships evidenced in this integrated E–R analysis support a favorable benefit-risk profile for lurbinectedin 3.2 mg/m2 q3wk. Trial registration: Clinicaltrials.gov: NCT02454972; registered May 27, 2015.