Nomogram predicting cancer-specific survival in elderly patients with stages I-III colon cancer.

ABSTRACT

Aim: This study aims to establish and validate an effective nomogram to predict cancer-specific survival (CSS) in elderly patients with stages I-III colon cancer.Methods: The data of elderly colon cancer patients with stages I-III were enrolled from the Surveillance, Epidemiology, and End Results database (SEER) between 2010 and 2015. The eligible patients were randomly divided into a training cohort and a validation cohort (ratio 11). All predictors of cancer-specific survival were determined by Cox regression. The concordance index (C-index) and calibration curves were used for validation of nomograms. Decision curve analysis (DCA) was performed to evaluate the clinical net benefit of the nomogram.Results: Cox hazard analysis in the training cohort indicated that grade, tumor stage, node stage, colectomy, and CEA were independent predictors of CSS. Nomogram was constructed based on these predictors. The C-index of nomograms for CSS was 0.728 (95%CI 0.7133-0.7427), and were superior to that of AJCC TNM Stage (C-index 0.625, 95%CI 0.6093-0.6406). The calibration curves showed satisfactory consistency between actual observation and nomogram-predicted CSS probabilities. The validation cohort demonstrated similar results. The DCA showed high net benefit of nomogram in a clinical context. The population was divided into three groups based on the scores of the nomogram, and the survival analysis showed that this prognostic stratification was statistically significant (p < 0.01).Conclusion: The nomograms showed significant accuracy in predicting 1-, 3-, and 5-year CSS in elderly patients with stages I-III colon cancer and may be helpful inpatient counseling clinical decision guidance.