Background: Many pts treated with ICIs can discontinue treatment and experience ongoing disease control and toxicity. We previously proposed a novel outcome, TFS, as the time between ICI therapy cessation and subsequent therapy initiation or death, with integrated graphical analysis to better characterize the unique effects of ICIs (ASCO 2018 #9531). In the CheckMate 067 and 069 trials of ipilimumab (IPI) and nivolumab (NIVO) alone or in combination (NIVO+IPI) in pts with MEL, the 36-month restricted mean TFS was 8.7, 4.6 and 11.1 mo, respectively. We explored the sensitivity of TFS to be informative of overall survival (OS) differences through subgroup analysis. Methods: Data from MEL pts in the CheckMate 067 and 069 trials were pooled and analyzed. TFS was defined as the area between the Kaplan-Meier curves for two endpoints, from randomization: (A) time to ICI therapy cessation; and (B) time to subsequent therapy initiation or death. TFS was estimated by restricted mean survival time until 36 mo since randomization. Differences in restricted mean TFS between subgroups of pts (such as PD-L1 status, lactate dehydrogenase [LDH], performance status [PS], gender) were calculated with bootstrapped 95% CIs. OS from randomization was also estimated. Results: Among 407 pts treated with NIVO+IPI, restricted mean TFS ranged from 8-13 mo across subgroups (Table). Subgroup differences in mean TFS were larger for prognostic factors LDH and PS and smaller in non-prognostic PD-L1 status and gender subgroups. Conclusions: Clinically meaningful TFS was seen across all subgroups. TFS was sensitive to prognostic subgroup differences. These results provide further support for TFS as a relevant clinical endpoint. Additional subgroups and comparisons with single agent NIVO and IPI arms will be presented at the meeting.