[The relationship between peripheral blood mitochondrial DNA copy number and incident risk of liver cancer: a case-cohort study].

Objective: To investigate the association between peripheral blood mitochondrial DNA copy number (mtDNAcn) and incident risk of liver cancer. Methods: At the baseline of Dongfeng-Tongji (DFTJ) cohort, 27 009 retirees were recruited from Dongfeng Motor Corporation in 2008. After excluding people without baseline DNA, with current malignant tumor and loss of follow-up, 1 173 participants were randomly selected into a sub-cohort by age-and gender-stratified sampling method at a proportion of 5% among all retirees. A total of 154 incident liver cancer cases identified from the cohort before December 31, 2018 (4 cases had been selected into the sub-cohort) were selected to form the case cohort of liver cancer. For the above 1 323 participants, their baseline levels of mtDNAcn in peripheral blood cells were measured by using quantitative real-time PCR method. The restricted cubic spline analysis was used to fit the shape of the association between baseline mtDNAcn and incident risk of liver cancer. The weighted Cox proportional hazards model was used to estimate the hazard ratio (HR) and 95%CI. Results: In this case-cohort study, the median follow-up time was 10.3 years. The restricted cubic spline analysis indicated that the relationship between peripheral blood mtDNAcn and incident risk of liver cancer followed a U-shaped pattern (Pnon-linear<0.05). All case-cohort population were divided into four subgroups by sex-specific quartiles of mtDNAcn levels among sub-cohort participants, when compared to participants in the Q2 subgroup of mtDNAcn, those in the Q1 subgroup (HR=2.00,95%CI:1.08-3.70) and Q4 subgroup (HR=4.11,95%CI:2.32-7.26) both had a significantly elevated risk of liver cancer, while those in the Q3 subgroup (HR=1.05,95%CI:0.54-2.05) had not. There were no significant multiply interaction effects of aging, gender, tobacco smoking, alcohol drinking and history of chronic hepatitis on the above association (Pinteraction>0.05). Conclusion: Both extremely low and high baseline level of mtDNAcn in peripheral blood cells are associated with an increased risk of incident liver cancer, but the underlying mechanisms need to be further clarified.目的： 探讨外周血线粒体DNA拷贝数（mtDNAcn）与肝癌发生风险的关联。 方法： 以东风-同济队列2008年基线调查纳入的27 009名东风汽车公司退休职工为基础，剔除基线DNA缺失、现患恶性肿瘤、失访的人群后，以5%的比例按年龄、性别分层随机抽取子队列人群1 173名，同时选取随访至2018年12月31日队列中发现的全部新发肝癌患者（共154例，其中4例已入选子队列）共同组成肝癌的病例队列。采用实时定量PCR法检测病例队列人群基线外周血mtDNAcn水平。采用限制性立方样条模型和加权Cox比例风险模型评估mtDNAcn与肝癌发病风险的关联。 结果： 本研究最终纳入研究对象1 323例。研究对象的中位随访时间为10.3年。限制性立方样条显示，外周血mtDNAcn与肝癌发生风险之间呈“U”型关联（P非线性<0.001）。将研究对象按子队列人群mtDNAcn的四分位数进行分组，调整年龄、性别、受教育程度、吸烟、饮酒、锻炼、体重指数和慢性肝炎史发现，与Q2亚组人群相比，Q1亚组（HR=2.00，95%CI：1.08~3.70）和Q4亚组（HR=4.11，95%CI：2.32~7.26）人群肝癌的发生风险均显著增加，未发现Q3亚组人群肝癌发生风险显著增加（HR=1.05，95%CI：0.54~2.05）。年龄、性别、吸烟、饮酒、慢性肝炎史与mtDNAcn在影响肝癌的发病风险上无交互作用（P交互>0.05）。 结论： 基线外周血mtDNAcn水平过高或过低均可增加肝癌发生风险，其作用通路有待更多机制研究进一步阐明。.