Development and external validation of a prognostic nomogram for patients with gastric cancer after radical gastrectomy

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 decisionmaking, thus avoiding overtreatment or undertreatment and improving communication between clinicians and patients.

Establishment of a prognostic model for gastric cancer patients who underwent radical gastrectomy using machine learning: a two-center study

“…Table 2 further validates these findings in the verification group, demonstrating the models’ strong external applicability. Collectively, these results underscore the effectiveness of machine learning models in accurately reflecting postoperative outcomes in gastric cancer surgery ( 30 , 31 ).…”

Prognostic significance of serum CA125 in the overall management for patients with gastrointestinal stromal tumors

“…In fact, high serum CA125 level often indicates postoperative recurrence or progression of carcinomas. Many predictive models established in numerous clinical studies described CA125 as an indicator for prognosis and metastasis of gastric cancer [ 12 , 13 ]. However, the relationship between serum CA125 concentration and GIST progression received few attentions.…”

Construction of a nomogram to predict the survival of metastatic gastric cancer patients that received immunotherapy

“…Nevertheless, their prognostic factors were different from the variables used in our current study. Other studies developed a prognostic risk stratification nomogram for GC patients after radical gastrectomy ( 36 – 38 ). Here, we constructed for the first time a nomogram for immunotherapy-based therapy for MGC.…”