11576 Background: We previously observed a significant improvement in post-relapse overall survival (OS) in patients with recurrent retroperitoneal soft-tissue sarcoma (RPS) treated in 2012–2021 compared with 2002–2011. This study aimed to determine whether this survival benefit reflects advances in post-relapse treatment strategies, using digital twins to isolate the impact of systemic therapy and multimodality care. Methods: Adults with resected RPS and first documented recurrence were grouped by era (pre-2012 vs post-2012). Post-recurrence management was modeled as time-dependent multimodal treatment to account for treatment timing and avoid time-related bias, including repeat surgery, radiotherapy, and systemic therapy (up to five sequential lines). Detailed treatment data (surgery, radiotherapy, number of systemic therapy lines, and use of novel agents available since 2012, within or outside clinical trials) were captured. A Cox model for post-relapse survival was fitted in the post-2012 cohort using baseline covariates and time-dependent treatments. This model was used to predict the counterfactual post-relapse survival on the Monte-Carlo simulated digital-twin trajectories for pre-2012 patients under post-2012 management. The effect of post-2012 systemic therapies was isolated and applied to the pre-2012 cohort to estimate their counterfactual impact. Restricted mean survival time (RMST) at 2 and 5 years was standardized to the pre-2012 baseline distribution. Results: Among 872 operated patients, 417 developed recurrence; complete data were available for 86 pre-2012 and 263 post-2012 patients. Post-relapse OS significantly improved over time, with median post-relapse OS of 14.9 months pre-2012 versus 43.8 months post-2012. Observed RMST differences favored the post-2012 cohort by +5.6 months at 2 years, and +18.0 at 5 years (p < 0.001). Counterfactual simulation predicted that applying post-2012 management to pre-2012 patients would increase median OS to 33.8 months and RMST by +4.8 and +12.2 months at 2 and 5 years, respectively. Isolating systemic therapy alone reduced the gain by −2.2, and −5.7 months, indicating a partial but not exclusive contribution of systemic treatment advances (p<0.001). Conclusions: Digital-twin target trial emulation suggests that post-2012 management strategies explain most of the early and a substantial portion of the long-term survival improvement after RPS recurrence. Advances in systemic therapy contribute meaningfully but do not fully explain the observed OS gains, supporting the importance of integrated multimodality post-relapse care.
Sanfilippo et al. (Wed,) studied this question.