摘要
目的 探讨男性尖锐湿疣(CA)患者人乳头瘤病毒(HPV)感染的基因分型,并进行复发危险因素预测模型的构建.方法 收集2021年1月—2023年6月到同济大学附属东方医院胶州医院就诊的724例CA患者的疣体细胞,提取HPV核酸,检测并统计HPV基因分型;采用多因素Logistic回归分析CA患者物理治愈后6个月内复发的危险因素,构建CA患者复发预测模型;采用受试者工作特征曲线(ROC)评价预测模型对CA患者复发的预测价值.结果 724例CA患者HPV主要基因分型为HPV 6、11、16和42型;多因素Logistic回归分析提示,病程≥6个月、高危型HPV感染、HPV多重感染和疣体数目≥5为CA患者复发的独立危险因素(均P＜0.05),使用安全套为保护因素(P=0.010).多因素分析成功构建预测模型.ROC曲线评价结果显示,当logit(P)＞13.00时,敏感度为71.39％,特异度为83.01％,曲线下面积(AUC)为0.863.结论 男性CA患者HPV感染基因分型主要为HPV6、11、16、42.根据复发危险因素进行的模型构建对CA患者复发具有较高的预测价值.
Abstract
Objective To investigate the genotypes of human papillomavirus(HPV)infection in male patients with condyloma acuminatum(CA)and to develop a model for predicting recur-rence risk.Methods Wart samples were collected from 724 CA patients at Jiaozhou Branch of Shanghai East Hospital,Tongji University from January 2021 to June 2023.Human papillomavirus DNA was extracted and genomic DNA was genotyped.Multivariate logistic regression analysis was used to analyze the risk factors for recurrence within 6 months after being cured with physiothera-py,and a prediction model for recurrence risk was developed.Receiver operating characteristic(ROC)curves was used to verify the prediction value of the prediction model.Results Among 724 CA patients,the main genotypes were HPV 6,11,16,and 42.Multivariate logistic regres-sion analysis showed that a disease course of≥6 months,high-risk HPV infection,multiple HPV infections,and the number of warts ≥ 5 were independent risk factors for recurrence(P＜0.05),while using condom was a protective factor for recurrence in CA patients(P=0.010).Multiple factor analysis was used to construct recurrence predictive model.The validation results of the ROC curve showed that when logit(P)＞13.00,the sensitivity and specificity were 71.39％and 83.01％,respectively,and the area under the curve(AUC)was 0.863.Conclusions The main genotypes of HPV infection are HPV 6,11,16,and 42 in male CA patients.The prediction model for recurrence risk factors has high predictive value for predicting recurrence of CA.
国家科技期刊平台
NSTL
万方数据