Objective To establish an effective prognostic nomogram for patients with gastric cancer after radical gastrectomy.
Methods In the present study, we retrospectively analyzed the records of 1635 patients with gastric cancer treated at Changhai Hospital from January 2014 to December 2019. Final independent prognostic factors of overall survival to build a nomogram were screened using univariate and multivariate analyses. The predictive accuracy and discriminating ability of the nomogram were determined using the concordance index (C-index) and calibration curves. The results were validated using bootstrap resampling of 701 patients. The survival curves for low-, middle-, and high-risk patients were used to evaluate the accuracy of the nomogram.
Results Multivariate analysis of the primary cohort revealed independent factors for survival as follows: age, sex, body mass index (BMI), depth of tumor invasion (T), lymph node metastasis (N), perineural invasion (PNI), tumor deposits, Alpha fetoprotein (AFP), and lactate dehydrogenase (LDH). Combining these predictors in the nomogram achieved powerful prognostic ability in the training and test sets. The C-index of the training set was 0.772 (95% CI, 0.748 to 0.796) compared with the C-index values of the AJCC 8th (0.718), T (0.683), N (0.726) and that of the testing set (0.726 95% CI, 0.684 to 0.767), which were significant higher than the C-index values of the TNM system. The calibration curve for probability of survival showed good consistency between predictions made using nomogram compared with actual observations. The survival curves showed significant differences between- high, middle-, and low-risk groups.
Conclusion We constructed a more accurate and comprehensive nomogram to predict the prognosis of gastric cancer after surgery, which may serve as a potential tool to guide personalized treatment.
DownLoad: