早期胃癌淋巴结转移预测模型的建立及验证

摘要：目的?探讨早期胃癌(EGC)淋巴结转移危险因素并构建预测模型。方法?回顾分析南京医科大学第一附属医院2017年1月至2019年6月172例行胃癌切除的EGC患者临床资料，随机分为预测组和验证组，Logistic回归分析用于确定与淋巴结转移相关的变量，ROC曲线判断预测模型的准确度，并用验证组验证预测模型。结果?172例EGC患者淋巴结转移率为18.60%(32/172)。肿瘤大小、浸润深度、分化程度和脉管癌栓与淋巴结转移有关(P<0.05)，但是浸润深度(OR=8.975，95%CI：1.653～48.712，P=0.011)和脉管癌栓(OR=17.178，95%CI：2.336～126.304，P=0.005)是EGC淋巴结转移的独立危险因素。预测组预测淋巴结转移的ROC曲线下面积(AUCROC)是0.864。结论?建立和验证了预测EGC淋巴结转移的预测模型，该模型可以帮助临床预测术前是否发生淋巴结转移。

Establishment and validation of prediction model for lymph node metastasis in early gastric cancer

Objective The aim of the study was to investigate the risk factors of lymph node metastasis in early gastric cancer (EGC) and develop a prediction model for the presence of lymph node metastasis. Methods The clinical data of 172 patients with EGC who underwent gastric cancer resection in the First Affiliated Hospital of Nanjing Medical University from January 2017 to June 2019 were retrospectively analyzed. Then the patients were randomly divided into test set and validation set. Logistic regression analysis was used to determine the variables significantly related to lymph node metastasis. The accuracy of prediction model was judged by ROC curve, and the prediction model was verified by validation set. Results The lymph node metastasis rate of 172 EGC patients was 18.60% (32/172). Tumor size, depth of invasion, degree of differentiation and vascular tumor thrombus were associated with lymph node metastasis (P < 0.05), but invasion depth (OR = 8.975, 95% CI: 1.653-48.712, P = 0.011) and vascular cancer embolus (OR = 17.178, 95% CI: 2.336-126.304, P = 0.005) were independent risk factors for lymph node metastasis of EGC. The area under the curve (AUC) of the test set was 0.8641. The most common site of metastasis was group 2, followed by group 3. Conclusion We established and validated a predictive model for predicting lymph node metastasis of early gastric cancer, which can help clinical preoperative decision-making.