Prognostic performance of the sarculator model in an Asian cohort of patients with extremity soft tissue sarcomas

BACKGROUND/PURPOSE: Sarculator is a prognostic nomogram for extremity soft-tissue sarcoma (eSTS), but its performance in East Asian populations has not been validated. This study externally validated the Sarculator for predicting 5- and 10-year overall survival (OS) and distant metastasis (DM) in an Asian cohort. METHODS: We retrospectively included 122 adults (median age 56 years) who underwent curative-intent resection for primary eSTS at a tertiary center in Taiwan (2010-2019). 5-year OS and DM rates were 62% (76/122) and 23% (28/122), respectively; 10-year OS and DM were 26% (16/61) and 39% (24/61). Model performance was assessed by area under the curve (AUC), calibration, Brier score, and decision-curve analysis. RESULTS: At 5 years, the Sarculator showed good discrimination (AUC 0.71 for OS; 0.73 for DM) and fair calibration, though OS was slightly overestimated (slope 0.75; intercept -0.32). It offered higher net benefit than the "treat-all" strategy at thresholds >0.27 (OS) and >0.02 (DM). For 10-year predictions, discrimination was moderate (AUC 0.71 for OS; 0.66 for DM), but OS calibration was poor (slope 0.66; intercept -1.72). Brier scores favored the Sarculator over the null model for 5-year predictions (OS 0.20 vs 0.22; DM 0.15 vs 0.17) but not for 10-year outcomes (OS 0.33 vs 0.19; DM 0.23 vs 0.23). CONCLUSION: Sarculator provided clinically useful and reasonably calibrated 5-year OS and DM estimates for Taiwanese patients with primary eSTS, supporting short-term decision-making. Its 10-year predictions were over-optimistic, warranting recalibration and validation in larger East Asian cohorts.