Construction and testing of a nomogram for patients with pancreatic signet-ring cell carcinoma and analysis of treatment-related survival differences: a study based on the Surveillance, Epidemiology, and End Results (SEER) database.

BACKGROUND

Pancreatic signet ring cell carcinoma (PSRCC) is a rare subtype of pancreatic cancer. The primary objective of this study is to use data on PSRCC patients from the Surveillance, Epidemiology, and End Results (SEER) database to develop a prognostic model for predicting overall survival (OS) in PSRCC patients and to analyze differences in prognosis among PSRCC patients receiving different treatment regimens, thereby providing guidance for the treatment of PSRCC patients.

METHODS

Patients with pathologically confirmed PSRCC [International Classification of Oncological Diseases, third edition (ICD-O-3); code 8490/3] were identified from the SEER database and randomly assigned to a training set (n=432) and a testing set (n=186). In the training set, independent prognostic factors for PSRCC were identified using univariate and multivariate Cox regression analyses. A prognostic nomogram was constructed. Predictive performance was evaluated using receiver operating characteristic (ROC) curves and calibration curves, and clinical utility was assessed via decision curve analysis (DCA). Validation was performed on the testing set. Finally, Kaplan-Meier survival curves were used to analyze survival differences among PSRCC patients receiving different treatment modalities.

RESULTS

Cox regression analysis revealed that age, site of origin, and American Joint Committee on Cancer (AJCC) stage were independent prognostic factors for cancer-specific survival (CSS) and OS. A predictive model was constructed based on these independent prognostic factors. For CSS prediction, the C-index in the training set was 0.62 [95% confidence interval (CI): 0.59-0.66], and the area under the curve (AUC) values at 1, 3, and 5 years were 0.72 (95% CI: 0.65-0.79), 0.85 (95% CI: 0.76-0.94), and 0.87 (95% CI: 0.76-0.99), respectively. The C-index for the testing set was 0.59 (95% CI: 0.53-0.64), with 1-, 3-, and 5-year AUC values of 0.66 (95% CI: 0.56-0.76), 0.70 (95% CI: 0.49-0.91), and 0.73 (95% CI: 0.41-1.05), respectively. Regarding OS, the C-index in the training set was 0.63 (95% CI: 0.60-0.67), and the AUC values at 1, 3, and 5 years were 0.74 (95% CI: 0.66-0.83), 0.85 (95% CI: 0.75-0.94), and 0.87 (95% CI: 0.76-0.99). The C-index for the testing set was 0.59 (95% CI: 0.54-0.64), with corresponding AUC values of 0.67 (95% CI: 0.54-0.79), 0.71 (95% CI: 0.49-0.92), and 0.73 (95% CI: 0.41-1.05). The calibration curve showed that the predicted results were consistent with the actual results. Patients were classified into high-risk and low-risk groups based on nomogram scores; the Kaplan-Meier survival curve showed that the prognosis for the low-risk group was significantly better than the high-risk group. DCA indicated that the model has good clinical utility. Furthermore, our study found that preoperative neoadjuvant therapy did not prolong patients' OS; however, postoperative adjuvant therapy did prolong OS.

CONCLUSIONS

Age, site of origin, and AJCC stage are independent prognostic factors for CSS and OS. The predictive model developed in this study demonstrated good predictive accuracy and clinical utility in both the training and testing sets. Furthermore, this study also indicated that preoperative neoadjuvant therapy did not prolong OS, whereas postoperative adjuvant therapy did.