Watling Cody Z, Sweetland Siân, Wojt Aika, Butera Gisela, Graubard Barry I, Floud Sarah, Matthews Charles E, Reeves Gillian K, McGlynn Katherine A
Metabolic Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rochville, MD, USA.
Cancer Epidemiology Unit, Nuffield Department of Population Health, University of Oxford, Oxford, UK.
Lancet Oncol. 2025 Aug;26(8):1031-1042. doi: 10.1016/S1470-2045(25)00222-0. Epub 2025 Jul 3.
Oral contraceptive use has been suggested to increase the risk of liver cancer. Although the International Agency for Research on Cancer concluded in 1999 that there was sufficient evidence of an association, this was based on case-control studies with few liver cancer cases. We aimed to provide more robust epidemiological evidence on this association by analysing data from two large prospective UK cohorts and additionally conducting a systematic review and meta-analysis of previous observational studies.
In our population-based study, the relationship between oral contraceptive use and liver cancer risk was examined using data from the Million Women Study (MWS) and the UK Biobank. We included women from both cohorts who did not have a prevalent cancer at baseline (except non-melanoma skin cancer) and had provided data on oral contraceptive use; incident liver cancer diagnoses were determined using linkage to National Health Service cancer registries. We compared risk in women who had ever used oral contraceptives with women who had never used oral contraceptives. Multivariable Cox proportional hazards regression was used to calculate hazard ratios (HRs) and 95% CIs. For the systematic review and meta-analysis, we searched PubMed, Embase, CINAHL Plus, Web of Science, and Scopus from database inception to June 28, 2024, for existing observational studies. Study-specific log odds ratios (ORs) or log HRs were pooled and we determined the relative risk (RR) between oral contraceptive use and liver cancer across all studies using a fixed-effects model (PROSPERO number CRD42024552518).
A total of 2765 (0·21%) of 1 305 024 participants developed liver cancer in the MWS cohort (median follow-up 21·4 years; IQR 18·4-22·4) and 191 (0·08%) of 253 408 participants developed liver cancer in the UK Biobank (median follow-up 12·6 years; IQR 11·8-13·4). No association was observed between ever versus never use of oral contraceptives and liver cancer risk in either the MWS (HR 1·05, 95% CI 0·97-1·13; p=0·27) or the UK Biobank (1·08, 0·76-1·55; p=0·66). The meta-analysis of 23 observational studies, which included 5422 individuals with liver cancer, found no evidence of an association between ever versus never oral contraceptive use and liver cancer (RR 1·04, 0·98-1·11; I 45·9%, p=0·0080). In the meta-analysis of duration of oral contraceptive use, there was a slightly increased risk of liver cancer per 5 years of use of oral contraceptives (RR 1·06, 1·02-1·10; I 63·9%, p<0·0001), with corresponding subtype-specific RRs of 1·07 (1·00-1·14) for hepatocellular carcinoma and 1·06 (1·01-1·11) for intrahepatic cholangiocarcinoma (p=0·82).
The totality of observational studies suggests there is no association between ever versus never use of oral contraceptive and liver cancer risk. When looking at associations by duration of oral contraceptive use, there was little or no association with all liver cancer or its two main subtypes. There might be a small increased risk of liver cancer with longer duration of use, but residual confounding cannot be ruled out.
Canadian Institutes of Health Research, National Institutes of Health Intramural Program, and Cancer Research UK.
有研究表明，口服避孕药会增加患肝癌的风险。尽管国际癌症研究机构在1999年得出结论，有充分证据证明二者存在关联，但这一结论是基于病例对照研究，其中肝癌病例较少。我们旨在通过分析来自英国两个大型前瞻性队列的数据，并对既往观察性研究进行系统评价和荟萃分析，为这一关联提供更有力的流行病学证据。
在我们基于人群的研究中，利用来自百万妇女研究（MWS）和英国生物银行的数据，研究口服避孕药的使用与患肝癌风险之间的关系。我们纳入了两个队列中在基线时没有患过癌症（非黑色素瘤皮肤癌除外）且提供了口服避孕药使用数据的女性；通过与国家医疗服务体系癌症登记处的数据链接确定新发肝癌诊断。我们比较了曾经使用过口服避孕药的女性和从未使用过口服避孕药的女性的风险。采用多变量Cox比例风险回归计算风险比（HRs）和95%置信区间（CIs）。对于系统评价和荟萃分析，我们检索了PubMed、Embase、CINAHL Plus、Web of Science和Scopus数据库，从建库至2024年6月28日，查找现有的观察性研究。汇总各研究的特定对数比值比（ORs）或对数风险比（HRs），并使用固定效应模型（PROSPERO注册号CRD42024552518）确定所有研究中口服避孕药使用与肝癌之间的相对风险（RR）。
在MWS队列的1305024名参与者中，共有2765名（0.21%）患肝癌（中位随访21.4年；四分位间距18.4 - 22.4年），在英国生物银行的253408名参与者中，有191名（0.08%）患肝癌（中位随访12.6年；四分位间距11.8 - 13.4年）。在MWS队列（HR 1.05，95% CI 0.97 - 1.13；p = 0.27）或英国生物银行队列（HR 1.08，0.76 - 1.55；p = 0.66）中，均未观察到曾经使用与从未使用口服避孕药与患肝癌风险之间存在关联。对23项观察性研究的荟萃分析，其中包括5422例肝癌患者，未发现曾经使用与从未使用口服避孕药与肝癌之间存在关联的证据（RR 1.04，0.98 - 1.11；I² 45.9%，p = 0.0080）。在口服避孕药使用时长的荟萃分析中，每使用口服避孕药5年，患肝癌的风险略有增加（RR 1.06，1.02 - 1.10；I² 63.9%，p < 0.0001），肝细胞癌的相应亚型特异性RR为1.07（1.00 - 1.14），肝内胆管癌为1.06（1.01 - 1.11）（p = 0.82）。
所有观察性研究表明，曾经使用与从未使用口服避孕药与患肝癌风险之间不存在关联。按口服避孕药使用时长来看，与所有肝癌或其两种主要亚型几乎没有关联。使用时长较长可能会使患肝癌风险略有增加，但不能排除残留混杂因素的影响。
加拿大卫生研究院、美国国立卫生研究院内部项目以及英国癌症研究中心。
