Park HJ, et al. Korean J Radiol. 2019 Oct;20:e101. https://doi.org/10.3348/kjr.2019.0470

Hwang SH, et al. J Gastric Cancer. 2022 Apr;22(2):94-106. https://doi.org/10.5230/jgc.2022.22.e8

Background: Personalized survival prediction is important in gastric cancer patients after gastrectomy based on large datasets with many variables including time‐varying factors in nutrition and body morphometry. One year after gastrectomy might be the optimal timing to predict long‐term survival because most patients experience significant nutritional change, muscle loss, and postoperative changes in the first year after gastrectomy. We aimed to develop a personalized prognostic artificial intelligence (AI) model to predict 5 year survival at 1 year after gastrectomy.Methods: From a prospectively built gastric surgery registry from a tertiary hospital, 4025 gastric cancer patients (mean age 56.1 ± 10.9, 36.2% females) treated gastrectomy and survived more than a year were selected. Eighty‐nine variables including clinical and derived time‐varying variables were used as input variables. We proposed a multi‐tree extreme gradient boosting (XGBoost) algorithm, an ensemble AI algorithm based on 100 datasets derived from repeated five‐fold cross‐validation. Internal validation was performed in split datasets (n = 1121) by comparing our proposed model and six other AI algorithms. External validation was performed in 590 patients from other hospitals (mean age 55.9 ± 11.2, 37.3% females). We performed a sensitivity analysis to analyse the effect of the nutritional and fat/muscle indices using a leave‐one‐out method.Results: In the internal validation, our proposed model showed AUROC of 0.8237, which outperformed the other AI algorithms (0.7988–0.8165), 80.00% sensitivity, 72.34% specificity, and 76.17% balanced accuracy. In the external validation, our model showed AUROC of 0.8903, 86.96% sensitivity, 74.60% specificity, and 80.78% balanced accuracy. Sensitivity analysis demonstrated that the nutritional and fat/muscle indices influenced the balanced accuracy by 0.31% and 6.29% in the internal and external validation set, respectively. Our developed AI model was published on a website for personalized survival prediction.Conclusions: Our proposed AI model provides substantially good performance in predicting 5 year survival at 1 year after gastric cancer surgery. The nutritional and fat/muscle indices contributed to increase the prediction performance of our AI model.

Background: Patients with gastric cancer have an increased nutritional risk and experience a significant skeletal muscle loss after surgery. We aimed to determine whether muscle loss during the first postoperative year and preoperative nutritional status are indicators for predicting prognosis. Methods: From a gastric cancer registry, a total of 958 patients who received curative gastrectomy followed by chemotherapy for stage 2 and 3 gastric cancer and survived longer than 1 year were investigated. Clinical and laboratory data were collected. Skeletal muscle index (SMI) was assessed based on the muscle area at the L3 level on abdominal computed tomography. Results: Preoperative nutritional risk index (NRI) and postoperative decrement of SMI (dSMI) were significantly associated with overall survival (hazards ratio: 0.976 [95% CI: 0.962–0.991] and 1.060 [95% CI: 1.035–1.085], respectively) in a multivariate Cox regression analysis. Recurrence, tumor stage, comorbidity index were also significant prognostic indicators. Kaplan-Meier analyses exhibited that patients with higher NRI had a significantly longer survival than those with lower NRI (5-year overall survival: 75.8% vs. 63.0%, P < 0.001). In addition, a significantly better prognosis was observed in a patient group with less decrease of SMI (5-year overall survival: 75.7% vs. 66.2%, P = 0.009). A logistic regression analysis demonstrated that the performance of preoperative NRI and dSMI in mortality prediction was quite significant (AUC: 0.63, P < 0.001) and the combination of clinical factors enhanced the predictive accuracy to the AUC of 0.90 (P < 0.001). This prognostic relevance of NRI and dSMI was maintained in patients experiencing tumor recurrence and highlighted in those with stage 3 gastric adenocarcinoma. Conclusions: Preoperative NRI is a predictor of overall survival in stage 2 or 3 gastric cancer patients and skeletal muscle loss during the first postoperative year was significantly associated with the prognosis regardless of relapse in stage 3 tumors. These factors could be valuable adjuncts for accurate prediction of prognosis in gastric cancer patients.

Jin Hee Noh, Jun Young Shin, Jeong Hoon Lee, Young Soo Park, In-Seob Lee, Ga Hee Kim, Hee Kyong Na, Ji Yong Ahn, Kee Wook Jung, Do Hoon Kim, Kee Don Choi, Ho June Song, Gin Hyug Lee, and Hwoon-Yong Jung. Gut and Liver 2023;17:69-77. https://doi.org/10.5009/gnl210593