Establishment and verification of a nomogram for predicting the risk of lymph node metastasis in early gastric cancer.

BACKGROUND

endoscopic submucosal dissection (ESD) has been widely recognized by patients and doctors due to its advantages in early gastric cancer (EGC). The accurate prediction of the risk of lymph node metastasis (LNM) in EGC is important to select suitable treatments with this procedure for patients. Unfortunately, the accuracy of endoscopic ultrasound and computed tomography in the diagnosis of EGC lymph node status is extremely limited. The purpose of the present study was to establish an LNM nomogram risk model of early gastric cancer patients based on clinical data, to guide treatment for clinicians.

METHODS

a retrospective examination of the records of EGC patients undergoing radical gastrectomy from August 2012 to August 2019 in the Gastrointestinal Center of Subei People's Hospital was performed. The clinicopathological data were classified into a training set and validation set according to the time. Univariate and multivariate analyses were performed to identify risk factors related to LNM. A risk model for predicting the occurrence of LNM in EGC was established and validated.

RESULTS

of the 503 EGC patients, 78 (15.5 %) had lymph node metastasis. Logistic stepwise regression analysis showed that the predictive factors included sex, tumor location, tumor diameter, differentiation, ulcer and lymphatic vascular invasion. The discrimination of the LNM prediction model was satisfactory with an AUC of 0.8033 (internal validation) and 0.7353 (external validation). The correction effect of the calibration was satisfactory and the DCA decision curve analysis showed a strong clinical practicability.

CONCLUSION

the nomogram risk prediction model of LNM has been established for EGC patients to assist in formulating personalized treatment plans.