宿主DNA甲基化预测子宫颈锥切术后病理转归及风险模型构建  

Host DNA methylation predicting pathological outcomes and risk model construction after cervical conization

作　　者：方秋悦 余志成 王晓丽 李敏[2] 张秀云 洪慧 陈峥峥 彭程[2] 周颖[2] 赵卫东 FANG Qiuyue;YU Zhicheng;WANG Xiaoli;LI Min;ZHANG Xiuyun;HONG Hui;CHEN Zhengzheng;PENG Cheng;ZHOU Ying;ZHAO Weidong(Graduate School of Bengbu Medical University,Bengbu 233000,China;不详)

机构地区：[1]蚌埠医科大学研究生院,蚌埠233000 [2]中国科学技术大学附属第一医院(安徽省立医院)妇产科,合肥230001 [3]安徽中医药大学第一附属医院妇产科,合肥230031

出　　处：《中国临床保健杂志》2025年第1期83-88,共6页Chinese Journal of Clinical Healthcare

基　　金：国家自然科学基金项目(82172774,82372637);安徽省临床医学研究转化专项(202427b10020072,202427b10020070)。

摘　　要：目的构建模型并验证宿主DNA六基因甲基化预测子宫颈锥切术后病理转归的能力,为子宫颈病变的诊疗提供可靠依据。方法回顾性分析2023年7月至2024年6月中国科学技术大学附属第一医院(安徽省立医院)收治的400例子宫颈病变患者的临床资料,这些患者在采取锥切干预前进行了六基因甲基化检测,收集患者的临床相关信息,将六基因甲基化结果结合每个标本在手术治疗前后的病理结果对比进行综合分析,筛选出独立影响因素后构建锥切术后病理转归风险预测模型并对其进行验证后获得校准曲线。结果甲基化结果(P<0.05,AUC=0.848,95%CI:0.813~0.884)相较于人乳头瘤病毒(P<0.05,AUC=0.540,95%CI:0.512~0.568)、液基细胞学(P<0.05,AUC=0.609,95%CI:0.566~0.653)能够有效预测子宫颈病变锥切术后病理转归。通过列线图建立子宫颈病变锥切前后病理变化的风险模型,C-index为0.905,可以有效预测子宫颈锥切术后病理变化风险,校准曲线平均绝对误差为0.018,展现出风险模型良好的预测能力。结论宿主DNA六基因甲基化检测对子宫颈病变锥切术后病理转归具有良好的预测价值,为早期评估子宫颈病变的病理变化提供诊疗思路和预测方法,指导临床诊疗策略。Objective To develop and validate a model that utilizes the six-gene methylation of host DNA to predict pathological outcomes after cervical conization,thereby providing a reliable basis for the diagnosis and treatment of cervical lesions.Methods The clinical data of 400 patients with cervical lesions admitted to the First Affiliated Hospital of University of Science and Technology of China(Anhui Provincial Hospital)from July 2023 to June 2024 was analyzed retrospectively.These patients underwent six gene methylation testing before undergoing conization intervention,Relevant clinical data were collected,and methylation results were compared with pathological findings from both pre-and post-surgical specimens.After screening out the independent influencing factors,the risk prediction model of pathological outcome after conization was constructed and validated,and the calibration curve was obtained.Results The six-gene methylation test(P<0.05,AUC=0.848,95%CI:0.813-0.884)was more effective than human papillomavirus testing(P<0.05,AUC=0.540,95%CI:0.512-0.568)and liquid-based cytology(P<0.05,AUC=0.609,95%CI:0.566-0.653)in predicting pathological outcomes after conization.A risk model for pathological changes before and after cervical conization was established using a nomogram,which demonstrated a C-index of 0.905.This model effectively predicted the risk of pathological changes post-conization,with a mean absolute error of 0.018 on the calibration curve,confirming its strong predictive ability.Conclusions The detection of six-gene methylation in host DNA provides significant predictive value for the pathological prognosis of cervical lesions after conization.This method offers new insights and prediction strategies for the early assessment of cervical lesion pathology,guiding clinical diagnosis and treatment decisions.

关 键 词：宫颈肿瘤 癌前状态 DNA甲基化 风险评估 

分 类 号：R73[医药卫生—肿瘤]