Akademik Veri Yönetim Sistemi
Asian Journal of Surgery, 2024 (SCI-Expanded)
Objectives: We aimed to develop a basic, easily applicable nomogram to improve the survival prediction of the patients with stage II/III gastric cancer (GC) and to select the best candidate for postoperative radiotherapy (RT). Methods: In this multicentric trial, we retrospectively evaluated the data of 1597 patients with stage II/III GC after curative gastrectomy followed by postoperative RT ± chemotherapy (CT). Patients were divided into a training set (n = 1307) and an external validation set (n = 290). Nomograms were created based on independent predictors identified by Cox regression analysis in the training set. The consistency index (C-index) and the calibration curve were used to evaluate the discriminative ability and accuracy of the nomogram. A nomogram was created based on the predictive model and the identified prognostic factors to predict 5-year cancer-specific survival (CSS) and progression-free survival (PFS). Results: The multivariate Cox model recognized lymph node (LN) involvement status, lymphatic dissection (LD) width, and metastatic LN ratio as covariates associated with CSS. Depth of invasion, LN involvement status, LD width, metastatic LN ratio, and lymphovascular invasion were the factors associated with PFS. Calibration of the nomogram predicted both CSS and PFS corresponding closely with the actual results. In our validation set, discrimination was good (C-index, 0.76), and the predicted survival was within a 10% margin of ideal nomogram. Conclusions: In our relatively large cohort, we created and validated both CSS and PFS nomograms that could be useful for underdeveloped or developing countries rather than Korea and Japan, where the D2 gastrectomy is routinely performed. This could serve as a true map for oncologists who must make decisions without an experienced surgeon and a multidisciplinary tumor board.