Development and internal validation of a prognostic nomogram for cancer-specific survival among elderly unresected colorectal cancer patients receiving chemotherapy: a retrospective SEER database study.

BACKGROUND: Colorectal cancer (CRC) is common among older adults, and a substantial proportion of elderly patients are unable to undergo surgical resection because of advanced disease, comorbidities, or poor physical condition. Although chemotherapy is an important treatment option for these patients, their survival outcomes are highly heterogeneous. Therefore, this study aimed to identify prognostic factors and develop an internally validated nomogram to predict cancer-specific survival (CSS) in elderly patients with unresected CRC receiving chemotherapy. METHODS: Clinical variables were sourced from the Surveillance, Epidemiology, and End Results (SEER) database, encompassing 2005 elderly CRC (ECRC) patients who underwent chemotherapy without surgery from 2004 to 2019. Cox regression analysis was conducted to identify prognostic factors, which were then utilized to construct the nomogram. The nomogram was validated using receiver operating characteristic (ROC) curves and calibration plots. RESULTS: The study identified seven prognostic factors, including tumor grade, bone metastasis, metastasis stage (M stage), race, tumor location, carcinoembryonic antigen (CEA) levels, and tumor stage (T stage), which were incorporated into the nomogram. The ROC curves demonstrated areas under the ROC curve (AUCs) of 0.705, 0.756, and 0.797 for predicting CSS at 1, 3, and 5 years in the training cohort, and 0.714, 0.822, and 0.807 in the validation cohort. Calibration plots revealed a high degree of concordance between observed and predicted CSS at 1, 3, and 5 years. CONCLUSIONS: The nomogram for predicting CSS in elderly patients with unresected CRC undergoing chemotherapy significantly enhances the precision of treatment.