摘要目的 探讨胆囊癌患者淋巴结阳性数与阴性数比值对数(LODDS)对胆囊癌患者预后的预测意义,并基于LODDS建立胆囊癌预测模型.方法 收集2010 ～2014年1 321例SEER数据库胆囊癌患者行淋巴结清扫术的临床资料.随机分为实验组(n =793)和验证组(n =528).采用COX多变量分析其独立预后因素,并建立列线图模型.用C指数评估列线图的准确性.采用ROC曲线评价列线图和TNM分期预测患者生存率的准确性.结果 COX分析显示年龄、肿瘤大小、分级、TNM分期和LODDS是独立影响因素,绘制列线图.实验组C指数为0.744(95％ CI,0.707-0.781),列线图预测3年和5年生存率的ROC曲线的AUC值分别为0.815,0.823.实验组TNM分期的C指数预测值为0.699(95％ CI,0.662-0.736),TNM分期预测3、5年生存率的ROC曲线的AUC值分别为0.779,0.790.结论 LODDS分级是胆囊癌患者独立的预后因素.以LODDS为基础的预测模型比TNM分期更为准确.

abstractsObjective To explore the predictive significance of logarithm of lymph node positive number to lymph node negative number (LODDS) in patients with gallbladder cancer,and to establish a predictive model of gallbladder cancer based on LODDS.Methods The clinical data of 1321 gallbladder cancer patients who underwent lymphadenectomy in SEER database from 2010 to 2014 were collected.They were randomly divided into test group (n =793) and verification group (n =528).COX multivariate analysis was used to determine independent prognostic factor,and a Nomogram prediction model was established.C index was used to evaluate the accuracy of patient Nomogram and TNM.ROC curve was used to evaluate the accuracy of the contour map and TNM staging in predicting survival rates.Results Independent risk factors for COX analysis were selected including age,tumor size,grade,TNM stage and LODDS analysis,and the contour map was drawn.In the experimental group,the C index was 0.744 (95％ CI,0.707-0.781),and the AUC values of ROC curves predicting 3-year and 5-year survival were 0.815 and 0.823,respectively.The predictive value of C index for TNM staging in the test group was 0.699 (95 ％ CI,0.662-0.736),and the AUC value of ROC curve for TNM staging predicting 3-year and 5-year survival was 0.779 and 0.790,respectively.Conclusion LODDS is an independent prognostic factor for gallbladder cancer.The prediction model based on LODDS is more accurate than TNM staging.