Development and validation of nomogram for predicting lymph node metastasis in early gastric cancer

Abstract

Abstract Purpose: To establish and verify a prediction model for lymph node metastasis (LNM) in early gastric cancer (EGC) and provide a reference for the selection of appropriate treatment for EGC patients. Methods: The clinicopathological data of 1584 patients with EGC admitted to Zhejiang Cancer Hospital from January 2010 to April 2019 were retrospectively analysed. Univariate and multivariate logistic regression analyses were used to explore the correlation between various clinicopathological factors and LNM in patients with EGC. Univariate K‒M and multivariate Cox regression analyses were used to explore the influence of multiple clinical factors on the prognosis of patients with EGC. The discrimination and calibration of the established prediction model, which is presented in the form of a nomogram, were also evaluated. Results: The incidence of LNM was 19.6%. Multivariate logistic regression analysis showed that tumour size, location, differentiation degree and pathological type were independent risk factors for LNM in EGC. Tumour pathological type and LNM were independent factors affecting the prognosis of patients with EGC. The area under the curve in the training and verification group was 0.750 (95% CI: 0.701 ~ 0.789) and 0.763 (95% CI: 0.687 ~ 0.838), respectively. The calibration curve showed good agreement between the predicted and actual probability, and decision curve analysis indicated strong clinical practicability. Conclusion: Tumour diameter ≥ 2 cm, poor differentiation degree, middle-lower tumour location and signet ring cell carcinoma were identified as independent risk factors for LNM in EGC. Among them, tumour pathological type and LNM were independent risk factors for prognosis in EGC. This clinical model for predicting LNM in EGC was used to construct a nomogram, which has high diagnostic value and can serve as a reference in clinical treatment selection.