<p>Cancer clinical trials often encompass various time-to-event (TTE) outcomes, such as overall survival, progression-free survival, and time-to-treatment failure, with each shedding light on related but different aspects of treatment efficacy. However, given the intricate relationships and mechanisms by which survival is extended for patients with cancer, making suitable treatment decisions for patients becomes challenging. To address this challenge and enhance precision in treatment selection, identifying patient subgroups before commencing treatment is important. To achieve this objective, concurrently considering two TTE outcomes and employing clustering analysis to determine the patient subgroups, thereby facilitating the selection of the most appropriate treatment, might be a more suitable approach. In this paper, on the basis of the random partition model with regression on covariates (PPMx), we propose a PPMx-based clustering model (DbTTE-PPMx) with Gaussian copula to jointly model two TTE outcomes while accounting for their correlations. Importantly, we account for the presence of semi-competing risks to consider the complicated interplay between these two TTE measures. In addition, we examine the operating characteristics of the proposed method in simulation studies. For illustration, we apply the proposed method to data from a clinical study in patients with advanced breast cancer.</p>