Validation of the personalized sarcoma care prognostic tool in Turkish patients with soft tissue sarcomas

Abstract

Objective: To validate the personalized sarcoma care (PERSARC) tool for predicting overall survival (OS) in a cohort of Turkish patients with soft tissue sarcomas (STS) and to assess its accuracy and clinical utility in a non-Western population. Methods: A retrospective analysis was performed on 100 patients diagnosed with STS between 2017 and 2023. Demographic, clinical, pathological, and treatment-related data were collected. Personalized sarcoma care predictions for OS and disease-free survival were compared with actual outcomes using calibration plots. Predictive performance was evaluated using receiver-operating characteristic analysis and Cox regression models. Results: Of the 100 patients, 59% were male and 57% were older than 50 years. Most tumors were deep-seated (75%) and >10 cm in size (52%). High-grade (Grade 3) tumors were present in 59% of cases. R0 resection was achieved in 95%, with neoadjuvant and adjuvant radiotherapy administered in 46% and 34% of patients, respectively. Local recurrence occurred in 14%, and distant metastases in 27%. At 3-year follow-up, 72% were alive. Personalized sarcoma care predicted 5-year OS as <= 50% in 33%, 51%-75% in 35%, and >75% in 32% of patients. Receiver-operating characteristic analysis demonstrated high predictive performance (area under the curve = 0.844; accuracy = 0.848; sensitivity = 0.607; specificity = 0.944). Personalized sarcoma care score was significantly associated with survival (P < .001), with a C-index of 0.857 (95% CI: 0.769-0.947). Conclusion: The PERSARC model demonstrates high predictive accuracy and calibration for Turkish STS patients, supporting its generalizability beyond Western populations. It offers a robust tool for personalized prognostication in clinical sarcoma care. Prospective validation in larger, diverse cohorts is recommended.