CO_(2)激光联合5-氨基酮戊酸-光动力疗法治疗尖锐湿疣复发风险列线图模型的构建  

Construction of a Nomogram Model for Recurrence Risks of Condyloma Acuminatum Treated Jointly with Carbon Dioxide Laser and Photodynamic Therapy 5-Aminolevulinic Acid Photodynamic Therapy

作　　者：刘婷[1] 倪通 李遇梅[2] LIU Ting;NI Tong;LI Yumei(Department of Dermatology,Nanjing Tongren Hospital,School of Medicine,South East University,Nanjing 211102,China;Department of Dermatology, Affiliated Hospital of Jiangsu University)

机构地区：[1]东南大学医学院附属南京同仁医院皮肤科,南京市211102 [2]江苏大学附属医院皮肤科

出　　处：《中国激光医学杂志》2023年第4期187-193,238,239,共9页Chinese Journal of Laser Medicine & Surgery

基　　金：江苏省卫生健康委2019年度医学科研重点项目(H201925)。

摘　　要：目的 分析CO_(2)激光联合5-氨基酮戊酸-光动力疗法(5-aminolevulinic acid photodynamic therapy,ALA-PDT)治疗尖锐湿疣的复发危险因素,构建可视化效果强的风险列线图模型并进行多维验证。方法 回顾性分析2018年1月至2021年3月,尖锐湿疣患者512例的临床资料,均采用CO_(2)激光联合ALA-PDT治疗,随访6个月共记录复发82例。比较复发与未复发患者的临床资料,采用LASSO回归法对危险因素进行降维处理,然后采用多因素logistic回归分析筛选独立危险因素,根据回归系数(β)绘制列线图预测模型。结果 与未复发患者相比,复发患者持续HIV感染率升高,低风险和高风险HPV共同感染率增多,免疫功能低下增多,疣分布于肛周、肛周+生殖器均增加,皮肤和黏膜均感染增多,皮损数目增多,局部除疣洗剂应用增多,每个疗程应用ALA剂量增加,差异均具有统计学意义(P<0.05)。LASSO回归法从上述7个指标中筛选出3个具有非零系数特征的变量,即持续HIV感染、免疫功能低下和皮损数目。logistic回归分析显示,持续HIV感染(OR=4.6,95%CI=2.7~6.9,P<0.001)、免疫功能低下(OR=1.3,95%CI=1.1~1.9,P=0.022)和皮损数目(OR=5.5,95%CI=4.0~8.2,P<0.001)是尖锐湿疣复发的独立危险因素。HosmerLemeshow检验显示模型总体具有统计学意义,Hosmer-Lemeshow检验值为0.765,提示模型拟合优度较高。列线图模型总分260分,计算一致性指数(concordance index,C-index)为0.844,提示模型准确性较高。受试者工作曲线(receiver operator characteristic,ROC)计算曲线下面积(area under the curve,AUC)为0.839,提示模型预测效能较好。校准曲线和决策曲线显示模型一致性和获益性尚可。结论 CO_(2)激光联合ALA-PDT治疗尖锐湿疣复发的独立危险因素有持续HIV感染、免疫功能低下和皮损数目,建立简化的列线图风险预测模型能够指导临床早期筛选复发高危人群,并进行早期干预,具有重要的临床应用价值。Objective To analyze the risk factors leading to the recurrence of condyloma acuminatum after the joint treatment of carbon dioxide laser and 5-aminolevulinic acid photodynamic therapy(ALA-PDT),construct a visual risk nomogram model and carry out multidimensional verification.Methods Totally 512 patients with condyloma acuminatum treated jointly with carbon dioxide laser and ALA-PDT in our hospital over the period from Jan.2018 to Mar.2021 were retrospectively summarized.The 6 month's follow-up reported the recurrence of 82patients in total.The clinical data of recurrent and non-recurrent patients were compared.The dimension of risk factors was reduced via LASSO regression,independent risk factors were screened out through multivariate logistic regression analysis,and a nomogram predictive model was developed with regression coefficient(β).Results Compared with the non-recurrent patients,the recurrent patients had a higher infection rate of concomitant HIV and lowand high-risk HPV,more serious immune deficiency,wart distribution of perianal,perianal and genitals and infection of skin and mucosa,greater number of skin lesions,higher application of topical wart removal lotion,and more ALA vials required for each treatment session(P<0.05).With LASSO regression method,three variables with non-zero coefficient characteristics were screened out from seven indexes,namely,concomitant HIV infection,immune deficiency and number of skin lesions.The results of a logistic regression analysis showed that concomitant HIV infection(OR=4.6,95% CI = 2.7~6.9,P<0.001),immune deficiency(OR=1.3,95% CI=1.1~1.9,P=0.022) and number of skin lesions(OR=5.5,95% CI=4.0~8.2,P<0.001) were the independent risk factors leading to condyloma acuminatum recurrence.It was proved through omnibus tests that the model was of statistical significance,and the Hosmer-lemeshow test value was 0.765,suggesting a high goodness of fit of the model.The total score of the nomogram model was 260,and the value of the consistency index(concordance index,C-index) w

关 键 词：5-氨基酮戊酸-光动力疗法 尖锐湿疣 复发 危险因素 列线图 

分 类 号：R454.2[医药卫生—治疗学] R752.53[医药卫生—临床医学]