Evaluation of treatment efficacy in oncology using OS is confounded by survival benefit from post-progression treatment. We pooled data from the BRIM-2, -3, -7, and coBRIM studies (BRAF inhibitor–naive patients with BRAFV600-mutated metastatic melanoma) to evaluate whether DOR could be a surrogate for OS.
Time-dependent Cox proportional hazards regression was used to model the association of DOR (interval from date of first RECIST response to progressive disease [PD] or death) with OS. The risk of death for DORs of 1–10 months (in 1-month increments) was evaluated. Patients with best response of stable disease or PD [nonresponders (NR)] were assigned a DOR of zero. Models were adjusted for time-fixed baseline covariates (ECOG status, demographics, disease covariates, and first-line treatment), and time-dependent covariates (DOR and post-progression treatment [immunotherapy, targeted therapy, or other]).
This analysis included 1365 patients (DTIC = 338; V = 717; C+V = 310). Objective response was 47.5% for the overall population and 11.5%, 53.6%, and 72.9% for the DTIC, V, and C+V cohorts, respectively. Median DOR was 9.3 months in the overall population and 6.4, 7.6, and 14.6 months in the DTIC, V, and C + V cohorts, respectively. Cox proportional hazards adjusted for time-dependent covariates showed a significant and progressive reduction in the risk of death with increasing DOR vs NR. The absolute risk of death decreased by a mean of ≈6.3–7.7% per month increase in DOR in the overall population and across treatment cohorts (Table). Sensitivity analyses in responders only showed similar results.Table:
1241P Hazard of death by duration in response (excluding time-dependent PD variables)
|Patient cohort||DOR of 1 month||DOR of 10 months||Mean per month HR decrease||Range of HR decrease||P-valuea|
|HR (95% CI)||HR (95% CI)|
|All patients||0.85 (0.82–0.87)||0.19 (0.14–0.25)||0.073||0.034–0.130|
These exploratory analyses suggest that DOR is independently associated with OS outcomes regardless of treatment and merits further exploration as a surrogate endpoint to assess long-term treatment benefit.
Clinical trial identification
Legal entity responsible for the study
F. Hoffmann-La Roche Ltd.
F. Hoffmann-LaRoche Ltd.
K. Lewis: Grants from Roche/Genentech, Amgen, EMD Serono, and Incyte, and personal fees from Roche/Genentech, Incyte, and SunPharma. J. Larkin: Institutional research support from MSD, Bristol-Myers Squibb, Pfizer, and Novartis and nonremunerated consultant for GSK, Novartis, MSD, Bristol-Myers Squibb, Pfizer, and Roche/Genentech. A. Ribas: Owns stock in Kite Pharma and has received honoraria from Roche, Amgen, Pfizer, and Merck. All monies paid to Dr. Ribas are deposited into the Division Account at the David Geffen School of Medicine and do not constitute personal income. K.T. Flaherty: Consultant for Roche. G.A. McArthur: Research grant support from Pfizer, Celgene, Ventana; consultant for Provectus; uncompensated consultancy for Pfizer, Millennium, GSK, Roche‐Genentech, Novartis, Bristol-Myers Squibb, and Amgen P.A. Ascierto: Consulting or advisory role for Amgen, Array, Bristol-Myers Squibb, Genentech/Roche, Merck Serono, Merck Sharp & Dohme, Novartis, and Pierre-Fabre, and research funding from Bristol-Myers Squibb, Genentech/Roche, and Array. B. Dréno: Personal fees from Roche, Bristol-Myers Squibb, Novartis, GlaxoSmithKline, and Amgen. E. McKenna: Employee, Genentech, Inc. Q. Zhu, Y. Mun: Employment, Genentech, Inc. A. Hauschild: Personal fees from Roche, Amgen, Bristol-Myers Squibb, Celgene, Eisai, GlaxoSmithKline, MedImmune, Merck Serono, Merck Sharp & Dohme/Merck, Novartis, Oncosec, and MELA Sciences.