Construction and validation of a nomogram for cancer specific survival of postoperative pancreatic cancer based on the SEER and China database.

ABSTRACT

BACKGROUND: The recurrence rate and mortality rate among postoperative pancreatic cancer patients remain elevated. This study aims to develop and validate the cancer-specific survival period for individuals who have undergone pancreatic cancer surgery. METHODS: We extracted eligible data from the Surveillance, Epidemiology, and End Results database and randomly divided all patients into a training cohort and an internal validation cohort. External validation was performed using a separate Chinese cohort. The nomogram was developed using significant risk factors identified through univariate and multivariate Cox proportional hazards regression. The effectiveness of the nomogram was assessed using the area under the time-dependent curve, calibration plots, and decision curve analysis. Kaplan-Meier survival curves were utilized to visualize the risk stratification of nomogram and AJCC stage. RESULTS: Seven variables were identified through univariate and multivariate analysis to construct the nomogram. The consistency index of the nomogram for predicting overall survival was 0.683 (95% CI 0.675-0.690), 0.689 (95% CI 0.677-0.701), and 0.823 (95% CI 0.786-0.860). The AUC values for the 1- and 2-year time-ROC curves were 0.751 and 0.721 for the training cohort, 0.731 and 0.7554 for the internal validation cohort, and 0.901 and 0.830 for the external validation cohorts, respectively. Calibration plots demonstrated favorable consistency between the predictions of the nomogram and actual observations. Moreover, the decision curve analysis indicated the clinical utility of the nomogram, and the risk stratification of the nomogram effectively identified high-risk patients. CONCLUSION: The nomogram guides clinicians in assessing the survival period of postoperative pancreatic cancer patients, identifying high-risk groups, and devising tailored follow-up strategies.