AUTHOR=Ding Baicheng , Luo Panquan , Yong Jiahui TITLE=Model based on preoperative clinical characteristics to predict lymph node metastasis in patients with gastric cancer JOURNAL=Frontiers in Surgery VOLUME=9 YEAR=2022 URL=https://www.frontiersin.org/journals/surgery/articles/10.3389/fsurg.2022.976743 DOI=10.3389/fsurg.2022.976743 ISSN=2296-875X ABSTRACT=Background

The risk factors of lymph node metastasis (LNM) in gastric cancer (GC) remain controversial. We aimed to identify risk factors of LNM in GC and construct a predictive model.

Methods

A total of 1,337 resectable GC patients who underwent radical D2 lymphadenectomy at the first affiliated Hospital of Anhui Medical University from January 2011 to January 2014 were retrospectively analyzed and randomly divided into training and validation cohorts (n = 1,003 and n = 334, respectively) in a 3:1 ratio. Collecting indicators include age, gender, body mass index (BMI), tumor location, pathology, histological grade, tumor size, preoperative neutrophils to lymphocytes ratio (NLR), platelets to lymphocytes ratio (PLR), fibrinogen to albumin ratio (FAR), carcinoembryonic antigen (CEA), cancer antigen19-9 (CA19-9) and lymph nodes status. Significant risk factors were identified through univariate and multivariate logistic regression analysis, which were then included and presented as a nomogram. The performance of the model was assessed with receiver operating characteristic curves (ROC curves), calibration plots, and Decision curve analysis (DCA), and the risk groups were divided into low-and high-risk groups according to the cutoff value which was determined by the ROC curve.

Results

BMI, histological grade, tumor size, CEA, and CA19-9 were enrolled in the model as independent risk factors of LNM. The model showed good resolution, with a C-index of 0.716 and 0.727 in the training and validation cohort, respectively, and good calibration. The cutoff value for predicted probability is 0.594, the proportion of patients with LNM in the high-risk group was significantly higher than that in the low-risk group. Decision curve analysis also indicated that the model had a good positive net gain.

Conclusions

The nomogram-based prediction model developed in this study is stable with good resolution, reliability, and net gain. It can be used by clinicians to assess preoperative lymph node metastasis and risk stratification to develop individualized treatment plans.