Construction and validation of a dynamic nomogram prognostic model based on perineural invasion and lymphovascular tumor embolus for patients with gastric cancer after postoperative chemotherapy
10.3760/cma.j.cn113855-20240813-00529
- VernacularTitle:神经浸润和脉管癌栓对胃癌术后化疗患者预后的影响及预后预测模型的构建
- Author:
Buyun SONG
1
;
Wenbo LIU
;
Yong LI
;
Xiaohan ZHAO
;
Mingming ZHANG
;
Xinyu YUAN
;
Zhaoxing LI
;
Bingyu WANG
;
Jiaxiang CUI
;
Zaibo ZHANG
;
Bibo TAN
Author Information
1. 河北医科大学第四医院外三科,石家庄 050011
- Publication Type:Journal Article
- Keywords:
Stomach neoplasms;
Neoplasm invasiveness;
Prognosis;
Nomograms
- From:
Chinese Journal of General Surgery
2025;40(8):631-638
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To verify the prognostic value of perineural invasion and lymphovascular tumor embolus for patients with gastric cancer undergoing gastrectomy and postoperative chemotherapy, and establish a prognostic prediction nomogram model.Methods:According to 7∶3 radio, 781 gastric cancer patients were randomly divided into training cohort and internal validation cohort. One hundred fifty patients were utilized as the external validation cohort. Univariate and multivariate analysis were performed to evaluate the prognostic value of perineural invasion and lymphovascular tumor embolus, and construct the dynamic nomogram. The concordance index (C-index), net reclassification index and integrated discrimination improvement index, receiver operating characteristic curve, calibration curves and decision curve analysis were used to evaluate the nomogram.Results:Perineural invasion ( HR=1.486, 95% CI: 1.150-1.919, P<0.01) and lymphovascular tumor embolus ( HR=1.321, 95% CI: 1.030-1.693, P<0.05) were independent prognostic risk factors for patients with gastric cancer after gastrectomy and postoperative chemotherapy. C-index (training cohort: 0.734, internal validation cohort: 0.755, external validation cohort: 0.715), net reclassification index (training cohort: 0.228 for 3-year and 0.213 for 5-year OS prediction; internal validation cohort: 0.211 for 3-year and 0.279 for 5-year OS prediction; external validation cohort: 0.220 for 3-year and 0.440 for 5-year OS prediction) and integrated discrimination improvement index (training cohort: 0.051 for 3-year and 0.041 for 5-year OS prediction; internal validation cohort: 0.027 for 3-year and 0.036 for 5-year OS prediction; external validation cohort: 0.063 for 3-year and 0.153 for 5-year OS prediction) indicated that the nomogram performed better than the traditional TNM staging system ( P<0.05). Conclusions:Perineural invasion and lymphovascular tumor embolus are independent prognostic risk factors of gastric cancer patients after postoperative chemotherapy. The novel dynamic nomogram model based on perineural invasion and lymphovascular tumor embolus provides better assistance in evaluating prognosis of gastric cancer patients.