中国实用外科杂志

中国实用外科杂志

• 论蓍 • 上一篇    下一篇

胃癌病人淋巴结转移nomogram图预测模型构建及评价

郑    晋a，桑玉尔b，施伟斌a   

  1. 上海交通大学医学院附属新华医院  a.普外科         b.消化内科，上海 200092
  • 出版日期:2021-05-01 发布日期:2021-05-17

  • Online:2021-05-01 Published:2021-05-17

摘要： 目的    探讨胃癌淋巴结转移的影响因素并构建nomogram图预测模型预测淋巴结转移概率。 方法    回顾性分析2013-01-01—2020-01-01上海交通大学医学院附属新华医院收治的879例未发生远隔转移及腹膜转移且经手术治疗的胃癌病人临床资料，利用单因素logistic回归、lasso回归筛选等统计学方法筛选变量，建立nomogram图并用ROC曲线进行评价模型。结果    CA242、肿瘤病灶大小、肿瘤浸润深度以及肿瘤分化程度被纳入构建nomogram图预测模型，内部验证评价模型AUC值为0.802（95%CI 0.766~0.838），外部验证评价AUC值0.791（95%CI 0.699~0.883）。 结论    构建胃癌淋巴结转移nomogram图预测模型，纳入变量在术前易于获得，模型具有较好的预测效果和较低的泛化误差，或可为临床医生在胃癌术前分期及治疗决策选择上提供参考。

关键词: 胃癌, 淋巴结转移, lasso回归, nomogram图

Abstract: Establishment and evaluation of a nomogram for prediction of lymph node metastasis in gastric cancer                        ZHENG Jin*, Sang Yu-er, SHI Wei-bin. *Department of General Surgery, Xinhua Hospital, Affiliated to Shanghai Jiaotong University School of Medicine, Shanghai 200092, China
Corresponding author: SHI Wei-bin，E-mail：shiweibin@xinhuamed.com.cn
Abstract    Objective    To explore the factors affecting lymph node metastasis of gastric cancer and to model with a nomogram to predict probability of that. Methods    The clinical data of 879 gastric cancer patients who underwent surgery without distant or peritoneal metastasis in Xinhua Hospital Affiliated to Shanghai Jiaotong University School of Medicine from January 1,2013 to January 1,2020 were analyzed retrospectively. The nomogram was built with variables selected by univariate logistical regression and lasso regression, and evaluated with internal and external validation ROC curve. Results    Four factors including carbohydrate antigen 242, tumor size, depth of tumor invasion and tumor differentiation were enrolled in the nomogram. The AUC value was 0.802 (95%CI 0.766—0.838） in internal validation and 0.791(95%CI 0.699—0.883) in external validation. Conclusion    A nomogram is developed with risk factors easily accessible before surgery in patients with gastric cancer, which can predict the probability of lymph node metastasis well and would be helpful for clinicians to make appropriate therapy strategies and preoperative staging of gastric cancer.

Key words: gastric cancer, lymph node metastasis, lasso regression, nomogram