基于游离DNA的乙肝相关肝细胞癌高危人群风险筛查的临床应用  

Clinical application of risk screening for high-risk population of hepatitis B-related hepatocellular carcinoma based on cell-free DNA

作　　者：戚德涛 李娟[2] 刘天府[2] 孔银[2] 刘元元[2] 麻爱娣 裴志燕 张岭漪[2] QI De-tao;LI Juan;LIU Tian-fu;KONG Yin;LIU Yuan-yuan;MA Ai-di;PEI Zhi-yan;ZHANG Ling-yi(The Second Clinical medical College of Lanzhou University,Gansu,Lanzhou 730030,China;Department of Hepatology,Lanzhou University Second Hospital,Gansu,Lanzhou 730030,China)

机构地区：[1]兰州大学第二临床医学院,730000 [2]兰州大学第二医院肝病科,730000

出　　处：《现代消化及介入诊疗》2023年第12期1480-1488,1494,共10页Modern Interventional Diagnosis and Treatment in Gastroenterology

基　　金：甘肃省自然科学基金(23JRRA0974);甘肃省教育科技创新项目(2022B-047);兰州大学第二医院“萃英科技创新”计划(CY2021-QN-A18)。

摘　　要：目的 探究基于游离DNA(cell free DNA,cfDNA)检测的肝细胞癌(hepatocellular carcinoma,HCC)高/低风险组的相关人群特征,分析并构建预测HCC发生的多因素预测模型。方法 对2021年12月至2023年7月来自兰州大学第二医院门诊及住院部就诊的慢性乙型肝炎(CHB)患者中HCC高危、极高危人群进行回顾性分析。依照cfDNA检测的HIFI值分为:HCC高风险组(HIFI值≥1)和HCC低风险组(HIFI值<1);对上述患者进行随访至2023年12月,依照随访中是否临床诊断HCC分为:HCC组与非HCC组。分别回溯收集2组患者的临床资料进行组间差异性比较,并进一步行相关性及多因素分析并构建多因素预测模型。结果 共纳入140例CHB患者,其中高风险组60例、低风险组80例;随访中HCC组24例,非HCC组116例。高低风险组在经过1∶1倾向性匹配去除混杂因素后,两组人群在随访发生HCC的人数占比的差异有统计学意义(P=0.048)。对研究人群的随访结果(经1∶4倾向性匹配)中,与非HCC组患者数据相比,HCC组患者呈现甲胎蛋白(AFP)偏高(P<0.001),异常凝血酶原(PIVKA-Ⅱ)偏高(P<0.001)以及存在拷贝数变异(CNV)的人数占比偏高(P=0.018)的特征。Logistic回归分析提示异常凝血酶原、存在拷贝数变异为CHB患者发生HCC的独立危险因素。基于发生HCC的独立危险因素构建的多因素HCC预测模型B[AUC 0.831 (95%CI:0.734~0.928)]与AFP联合PIVKA-Ⅱ预测模型[AUC 0.818(95%CI:0.718-0.912)]对HCC发生均有较好预测价值。结论 预测HCC的发生应结合多因素考虑,高PIVKA-Ⅱ水平、存在拷贝数变异,同时合并高甲胎蛋白水平的CHB患者存在发生HCC的较高潜在风险,应予以临床的密切随访。Objective To characterize high and low-risk groups for hepatocellular carcinoma(HCC) using cell-free DNA(cfDNA) detection and to develop a multivariate prediction model for HCC occurrence.Methods A retrospective analysis was undertaken on high and very high-risk groups of hepatocellular carcinoma(HCC) within chronic hepatitis B(CHB) patients attending Lanzhou University Second Hospital's outpatient and inpatient departments from December 2021 to July 2023.These patients were categorized based on their HIFI value detected by cfDNA into HCC high-risk(HIFI value≥1) and HCC low-risk(HIFI value<1) groups.Follow-up assessments continued until December 2023,during which clinical diagnosis of HCC determined their assignment into HCC and non-HCC groups.Retrospectively collected clinical data from these groups facilitated comparative analysis,leading to further correlation and multifactorial analysis aimed at constructing a multifactor prediction model.Results Among the 140 chronic hepatitis B(CHB) patients,60 were classified as high-risk and 80 as low-risk individuals.Subsequently,24 cases were identified in the HCC group,with 116 cases in the non-HCC group during the follow-up period.There was a statistically significant difference in the incidence of HCC between the high and low risk groups after using 1∶1 propensity score matching to remove confounding factors(P=0.048).In the study's follow-up results(using a 1∶4 propensity score matching),patients in the HCC group exhibited elevated levels of alpha-fetoprotein(AFP)(P=0.001)and abnormal prothrombin(PIVKA-Ⅱ)(P<0.001),along with a higher detection rate of copy number variation(CNV)(13.0%,P=0.036).Logistic regression analysis identified PIVKA-Ⅱ and CNV as independent risk factors for HCC among CHB patients.The multivariate HCC prediction model B[ AUC 0.831(95%CI:0.734-0.928)] based on independent risk factors for HCC and the combined AFP with PIVKA-Ⅱ model[AUC 0.818(95%CI:0.718~0.912)]exhibited good predictive capabilities for HCC occurrence.Conclusion The pre

关 键 词：游离DNA 肝细胞癌 倾向性评分匹配 异常凝血酶原 拷贝数变异 

分 类 号：R575[医药卫生—消化系统]