A propensity-score matched analysis for time to trabectedin-failure in advanced myxoid liposarcoma patients.

Abstract

11571 Background: Whether to discontinue trabectedin (T) in soft-tissue sarcoma (STS) patients (pts) with response or stable disease remains controversial. In 2015, the T-DIS trial (Le Cesne et al., Lancet Oncology) showed a progression-free survival (PFS) benefit for T-maintenance after six cycles in STS patients. However, myxoid liposarcoma (MLPS) show a unique sensitivity to T, likely due to its specific mechanism of action. We retrospectively evaluated a "stop-and-go" strategy, in terms of time to treatment (T) failure (TTF) and overall survival (OS) in advanced MLPS pts. Methods: We carried out a retrospective analysis of 79 MLPS pts treated with T at our institution from September 2002 to December 2024. Inclusion criteria required pts to be progression-free (per RECIST) after 6 cycles of T. TTF was defined as the time from the 6th cycle to secondary resistance to T or death, whichever occurred first. Statistical methods included Kaplan-Meier survival analysis and propensity-score matched Cox proportional hazards models (weighted and unweighted) with time-dependent variables. Results: 60 MLPS pts were eligible for analysis (median age: 48 years, IQR: 39–57; 37 males, 23 females). Of these, 27 (34%) pts were treated with a "stop-and-go" strategy, discontinuing T after achieving radiological response or disease stability. T was discontinued at a median of 12 cycles (range: 6-28). Median follow-up was 56.9 months (IQR: 28.1–112.5), with a median TTF of 51.4 months (95% CI: 37.1– NA) in the "stop-and-go" group versus 10.0 months (95% CI: 6.6–17.7) in the T-maintenance group. OS was 55 months (95% CI 38.5 - NA) vs 23 months (95% CI 21-27.9) in the two groups, respectively. Weighted Cox models predicted an HR of 0.32 (95% CI: 0.11–0.93) for TTF and of 0.28 (95% CI 0.07 – 1.12) for OS, thus favoring the stop-and-go approach. Additional factors, including ECOG performance status (PS), age, primary tumor site, and surgery post-T discontinuation, were adjusted in the analysis. Conclusions: Our findings raise the question whether a "stop-and-go" strategy with T is the best option in the subset of MLPS pts. How to test further this hypothesis in such a rare subset, and which predictive factors may optimize the approach, is questionable.