Prognostic nomogram for colorectal cancer liver metastasis treated with tumor resection and chemotherapy based on SEER database

Background: Colorectal cancer liver metastasis (CRCLM) remains a leading cause of mortality in colorectal cancer (CRC) patients. While primary tumor resection combined with chemotherapy has shown promise, robust prognostic models for these patients remain scarce. This study aimed to develop and validate a nomogram to predict overall survival (OS) in CRCLM patients undergoing primary tumor resection and chemotherapy. Methods: Data from 3,252 CRCLM patients (2010-2015) were obtained from the Surveillance, Epidemiology, and End Results (SEER) database. Patients were randomly divided into training (70%, n=2,276) and validation (30%, n=976) cohorts. Independent prognostic factors were identified using Cox regression. A nomogram was constructed to predict 1-, 3-, and 5-year OS, with performance assessed via receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA). A web-based calculator was developed for clinical use. Results: Multivariate analysis identified race, age, marital status, tumor site, grade, carcinoembryonic antigen (CEA) level, tumor deposits, regional nodes examined, regional nodes positive, and liver metastases surgery as independent prognostic factors. The nomogram demonstrated good discrimination, with area under the curves (AUCs) of 0.729 (1-year), 0.710 (3-year), and 0.714 (5-year) in the training cohort and 0.717, 0.736, and 0.737 respectively in the validation cohort. Calibration curves showed strong agreement between predicted and observed outcomes. Through 1,000 bootstrap resampling iterations, the Cox model demonstrated superior discriminative accuracy with a mean C-index of 0.657 compared to the Fong clinical risk score (0.609), Basingstoke index (0.558), and tumor-node-metastasis (TNM) staging system (0.629) (all P<0.001). DCA indicated clinical utility across risk thresholds. Patients were stratified into low-, intermediate-, and high-risk groups (scores <220, 220-301, >301). The web calculator was accessible at https://lxt134520.shinyapps.io/output/. Conclusions: This nomogram provides individualized OS prediction for CRCLM patients treated with resection and chemotherapy. However, limitations include the lack of external validation, exclusion of chemotherapy regimens/molecular markers, and potential socioeconomic confounders affecting race associations. Future studies should incorporate treatment-specific variables and validate the model in diverse cohorts to enhance generalizability.