Development and external validation of a prognostic nomogram for patients with gastric cancer after radical gastrectomy
Xi’e Hu,Zhenyu Yang,Songhao Chen,Jingyi Xue,Sensen Duan,Lin Yang,P. Yang,Shujia Peng,Yanming Dong,Lijuan Yuan,Xianli He,G. Bao
DOI: https://doi.org/10.21037/atm-21-6359
IF: 3.616
2021-12-01
Annals of Translational Medicine
Abstract:Background Gastric cancer (GC) is one of the most malignant diseases and threatens the health of individuals across the globe. Hitherto, the identification of prognosis risk stratification on GC has mainly depended on the TNM staging, but owing to its inaccuracy and incompleteness, the prognostic value it offers remains controversial in the current clinical setting. Thus, an effective prognostic model for GC after radical gastrectomy is still needed. Methods Patients with pathologically confirmed GC who underwent radical gastrectomy from 2 different centers were retrospectively enrolled into a training and the validation cohort, respectively. The least absolute shrinkage and selection operator (LASSO) algorithm was applied to select variables among multiple factors, including clinical characteristics, pathological parameters, and surgery- and treatment-related indicators. The multivariate Cox regression method was used to establish the model to predict 1-, 2-, and 3-year survival. Both internal and external validations of the nomogram were then completed in terms of discrimination, calibration, and clinical utility. Finally, prognostic risk stratification of GC was conducted with X-tile software. Results A total of 1,424 patients with GC were eligible in this study, including 1,010 in the training cohort and 414 in the validation cohort. Seven indicators were selected by LASSO to develop the nomogram, including the number of positive lymph nodes, tumor size, adjacent organ invasion, vascular invasion, the level of carbohydrate antigen 125 (CA 125), depth of invasion, and human epidermal growth factor receptor 2 (HER2) status. The nomogram demonstrated a robust predictive capacity with favorable accuracy, discrimination, and clinical utility both in the internal and external validations. Moreover, we divided the population into 3 risk groups of survival according to the cutoff points generated by X-tile, and in this way, the nomogram was further improved into a risk-stratified prognosis model. Conclusions We have developed a prognostic risk stratification nomogram for GC patients after radical gastrectomy with 7 available indicators that may guide clinical practice and help facilitate tailored decision-making, thus avoiding overtreatment or undertreatment and improving communication between clinicians and patients.
Medicine