Beyond Event-Free Survival: A Multi-State Analysis of Local Recurrence, Metastasis, and Death in Soft Tissue Sarcoma

Background/Objectives: Event-free survival (EFS) is widely used in soft tissue sarcoma (STS) research but combines clinically distinct outcomes such as local recurrence (LR), distant metastasis (DM), and death into a single endpoint. We aimed to decompose conventional EFS and characterize STS progression as a longitudinal process of transitions between clinically distinct disease states. Methods: Prospectively and retrospectively collected registry data from 662 patients with STS at two tertiary sarcoma centers were retrospectively analyzed. EFS was decomposed by first-event type, and longitudinal disease evolution was assessed using multi-state modeling and time-dependent Cox regression. Results: Among 585 patients included in the EFS analysis, first events were isolated LR in 114 (19.5%), DM in 106 (18.1%), synchronous LR and DM in 21 (3.6%), and death without documented recurrence in 19 (3.2%). In the full cohort, 151 deaths occurred. After multivariable adjustment, isolated LR showed a moderate but statistically imprecise association with mortality compared with the no documented LR or DM state (HR 2.55, 95% CI 0.93–7.02), whereas substantially greater mortality hazards were observed for isolated DM (HR 17.56, 95% CI 9.46–32.60) and combined LR and DM (HR 32.63, 95% CI 16.86–63.17). Conclusions: STS progression is characterized by clinically distinct, evolving disease states with markedly different prognostic implications. A disease-state framework complements conventional EFS by preserving the nature and longitudinal evolution of recurrence and may provide a more informative representation of STS progression.