LAUSR

Background The dismal prognosis of gastric signet ring cell carcinoma (GSRC) is a global problem. The current study is conducted to comprehensively evaluate clinicopathological features and survival outcomes in GSRC patients stratified by anatomic subsites. Then, predictive nomograms are constructed and validated to improve the effectiveness of personalized management. Method The patients diagnosed with GSRC were recruited from the online SEER database. The influence of anatomic subsites on overall survival (OS) and cancer-specific survival (CSS) was evaluated using multivariate Cox regression and Kaplan-Meier analysis. Then, we employed propensity score matching (PSM) technique to decrease selection bias and balance patients' epidemiological factors. Predictive nomograms were constructed and validated. Sensitivity analysis was performed to validate the conclusion. Results Multivariate Cox regression demonstrated that the patients with overlapping gastric cancer (OGC) suffered the highest mortality risk for OS (HR, 1.29; 95% CI, 1.23-1.36; P < 0.001) and CSS (HR, 1.33; 95% CI, 1.28-1.37; P < 0.001). Age, TNM stage, tumor localization, tumor size, surgery, and chemotherapy presented a highly significant relationship with OS and CSS. Following subgroup and PSM analysis, OGC patients were confirmed to have the worst OS and CSS. Then, nomograms predicting 6-month, 12-month, and 36-month survival were constructed. The area under the curve (AUC) value in ROC was 0.775 (95% CI, 0.761-0.793) for 6-month survival, 0.789 (95% CI, 0.776-0.801) for 12-month survival, and 0.780 (95% CI, 0.765-0.793) for 36-month survival in the OS group, while in the CSS group, it was 0.771 (95% CI, 0.758-0.790) for 6-month survival, 0.781 (95% CI, 0.770-0.799) for 12-month survival, and 0.773 (95% CI, 0.762-0.790) for 36-month survival. Conclusion We identified anatomic subsites as a predictor of survival in those with GSRC. Patients with OGC suffered the highest mortality risk. The proposed nomograms allowed a relatively accurate survival prediction for GSRC patients.