文章摘要
权莉,张子晚,李莉.Ⅰ~Ⅲ期结直肠癌病人神经 /脉管浸润风险因素分析并构建列线图预测模型[J].安徽医药,2024,28(6):1203-1207.
Ⅰ~Ⅲ期结直肠癌病人神经 /脉管浸润风险因素分析并构建列线图预测模型
Risk factor analysis and construction of a prediction nomogram for perineural or lymphovascular invasion in patients with stage Ⅰ-Ⅲ colorectal cancer
  
DOI:10.3969/j.issn.1009-6469.2024.06.031
中文关键词: 结直肠肿瘤  神经 /脉管浸润  危险因素  预测模型  列线图
英文关键词: Colorectal neoplasms  Perineural or lymphovascular invasion  Risk factor  Prediction model  Nomogram
基金项目:
作者单位E-mail
权莉 徐州医科大学第一临床医学院江苏徐州 221004  
张子晚 徐州医科大学第一临床医学院江苏徐州 221004  
李莉 徐州医科大学附属医院消化内科江苏徐州 221006 lily9711214@126.com 
摘要点击次数: 90
全文下载次数: 50
中文摘要:
      目的建立预测 Ⅰ~Ⅲ期结直肠癌病人神经 /脉管浸润( perineural or lymphovascular invasion,PNI/LVI)的列线图预测模型,验证和预测其效能。方法回顾性分析 2017年 1月至 2022年 9月徐州医科大学附属医院行结直肠癌根治术治疗的结直肠癌病人 684例的临床病理资料,利用 R软件采用随机数种子的方式将上述资料按 7∶3分成训练组( 479例)和验证组( 205例)。通过单因素相关分析,把 P<0.1的参数纳入到多因素 logistic回归分析,筛选影响结直肠恶性肿瘤 PNI/LVI的危险因素,建立列线图模型。通过受试者操作特征曲线( ROC曲线)评估风险模型的预测价值,然后通过一致性指数、校准曲线和临床决策曲线分析(DCA)评估预测模型识别、校准和临床实用性方面的有效性。结果糖尿病、肿瘤长径、组织学分级、癌结节、 T分期、 N分期、血小板计数 /淋巴细胞计数( PLR)、血小板计数 ×中性粒细胞计数 /淋巴细胞( SII)、纤维蛋白原( g/L)/前白蛋白( g/L)(FPR)、癌胚抗原( CEA)、糖类抗原( CA)199在单因素分析中 P<0.1。而多因素分析结果表明, CA199、T分期和 N分期与 Ⅰ~Ⅲ期结直肠癌病人 PNI/LVI的发生独立相关因素。训练组和验证组的一致性指数分别为 0.88[95%CI:(0.85,0.91)]和 0.84[95%CI:(0.78,0.89)]同样校准曲线和 DCA表明该模型具有良好的准确度和临床实用性。结论以 CA199、T分期和 N分期为基础建立的列线图模型,有助于预测 Ⅰ~Ⅲ期结直肠癌病人的 PNI/LVI风险,该风险评价模型在临床上具有较好的预测效果。
英文摘要:
      Objective To establish a nomogram prediction model for predicting perineural or lymphovascular invasion (PNI/LVI) in patients with stage Ⅰ-Ⅲ colorectal cancer, and to verify and predict its efficacy.Methods The clinicopathological data of 684 patients who underwent radical resection for colorectal cancer therapy at the Affiliated Hospital of Xuzhou Medical University from January 2017 to September 2022 were retrospectively analyzed, and the aforementioned patients were divided into a training group (479 patients) and a validation group (205 patients) at a ratio of 7∶3 by using R software with a random number seeding method. Through uni variate correlation analysis, parameters with P<0.1 were included in multivariate logistic regression analysis to screen the risk factorsaffecting the PNI/LVI of colorectal malignant tumors and establish a nomogram model. The predictive value of the risk model was estimated by receiver operating characteristic (ROC) curve analysis, and then the validity of the prediction model in terms of identification,calibration and clinical practicability was evaluated by consistency (C) indices, calibration curve and clinical decision curve analysis(DCA).Results Diabetes mellitus, tumor size, histological grade, cancer nodules, T stage, N stage, platelet count/lymphocyte count(PLR), platelet count × neutrophil count/lymphocyte (SII), fibrinogen (g/L)/prealbumin (g/L) (FPR), carcinoembryonic antigen (CEA),and glycoantigen (CA199) were P<0.1 according to univariate analysis. Multivariate analysis showed that CA199, T stage and N stagewere independently associated factors with the occurrence of PNI/LVI in patients with stage Ⅰ-Ⅲ colorectal cancer. The C indices of the training and validation groups were 0.88 [95% CI: (0.85,0.91)] and 0.84 [95% CI: (0.78,0.89)], respectively, and the same calibration curves and DCA indicated that the model had good accuracy and clinical utility. Conclusion The nomogram model based on CA199, T stage and N stage is helpful for predicting the risk of PNI/LVI in patients with stage Ⅰ-Ⅲ colorectal cancer, and this risk evaluation model has good predictive ability in the clinic.
查看全文   查看/发表评论  下载PDF阅读器
关闭

分享按钮