RESUMEN
AIM: A nomogram for prediction of 12-year sarcoma-specific survival has been developed based on patients with soft tissue sarcomas treated in Memorial Sloan Kettering Cancer Centre (MSKCC). We aim to evaluate the predictive accuracy of the MSKCC sarcoma nomogram in a cohort of patients treated at an Asian institution. This has not been validated in an Asian population and thus its universal applicability remains unproven. MATERIALS AND METHODS: Between 1990 and 2013, 840 adult patients underwent treatment for primary soft tissue sarcoma (STS) at the National Cancer Centre Singapore. Patients who presented with locally recurrent or metastatic disease were excluded from the analysis. The variables included in the MSKCC nomogram included age at diagnosis, tumor size, histologic grade, histologic subtype, depth and site. A total of 399 patients were left for analysis. The nomogram was validated by assessing its extent of discrimination and level of calibration. RESULTS: All patients had deep tumors. Disease occurred most commonly in the lower extremity (n = 149 [37.3%]), the most common histologic subtype was "Others" (angiosarcoma, ewing's sarcoma, endometrial stromal sarcoma, sarcoma NOS [not otherwise specified] and rhabdomyosarcoma). Sixty-four percent of all patients had high-grade tumors while 36% had low-grade tumors. The median patient age at diagnosis was 54 years (range: 17-88 years). The median follow up time for all patients and surviving patients were 29 (range: 1-174) and 33 (range: 1-157) months, respectively. The observed 5- and 10-year sarcoma-specific survival were 55% and 33%, respectively. The concordance index was 0.71. For level of calibration, the observed correspondence between predicted and actual outcomes suggest that the MSKCC nomogram generally predicts well for patients with higher survival probability, but consistently overpredicts survival for the other groups, in our cohort of patients. CONCLUSION: The MSKCC sarcoma nomogram was found to be accurate in terms of extent of discrimination. In terms of level of calibration, it generally predicts well for patients with higher survival probability, but consistently overpredicts survival for the other groups in our population.