Validation of the postoperative nomogram for 12-year sarcoma-specific mortality

Abstract

Background: On the basis of a prospectively followed cohort of adult patients with primary soft tissue sarcoma (STS) who were treated at Memorial Sloan-Kettering Cancer Center (MSKCC; New York, NY), a nomogram for predicting sarcoma-specific mortality was developed. Although this nomogram was found to be accurate by internal validation tests, it had not been validated in an external patient cohort, and thus its universal applicability remained unproven.

Methods: Between 1975 and 2002, 1167 adult patients (age > or = 16 years) underwent treatment for primary STS at the University of California-Los Angeles (UCLA; Los Angeles, CA). All patients treated with an ifosfamide-based chemotherapy protocol (n = 238) were excluded from the current analysis. The remaining 929 patients constituted the population on which the validation study was performed. The nomogram validation process comprised two activities. First, the extent of discrimination was quantified using the concordance index. Second, the level of calibration was assessed by grouping patients with respect to their nomogram-predicted mortality probabilities and then comparing group means with observed Kaplan-Meier estimates of disease-specific survival.

Results: With median follow-up intervals of 48 months for all patients and 60 months for surviving patients, the 5-year and 10-year disease-specific survival rates were 77% (95% confidence interval [CI], 74-80%) and 71% (95% CI, 67-75%), respectively. Application of the nomogram to the UCLA data set yielded a concordance index of 0.76, and the observed correspondence between predicted and actual outcomes suggested a high level of calibration.

Conclusions: In the current study, the MSKCC Sarcoma Nomogram was found to provide accurate survival predictions when it was applied to an external cohort of patients who were treated at UCLA.