Approximately 4.5 million adults in the United States are living with liver disease, and chronic liver disease was the ninth leading cause of death in 2021 [1]. Liver disease is increasingly common among women of reproductive age, and consideration of safe and effective contraceptive use is important for those who wish to avoid pregnancy [2]. Hormonal methods of contraception have effects on and are metabolized by the liver; thus, consideration of the severity of existing disease is important in determining the risks, benefits, and a woman’s medical eligibility for hormonal contraception (HC). Estrogen-containing drugs have direct receptor-mediated impact on hepatocytes, whereas progestins do not. However, some noncontraceptive formulation progestins at high doses may be metabolized into estrogenic compounds [3–6]; thus, there are theoretical concerns of impaired hepatic metabolism of exogenous hormones in women with liver dysfunction and that use of HC in the setting of liver disease may worsen liver function and affect disease progression.

Two benign lesions of the liver are hepatocellular adenomas (HCAs) and focal nodular hyperplasia (FNH). HCAs have a rare risk of hemorrhage or malignant transformation. Increased exposure to estrogen from combined hormonal contraception (CHC) and from obesity has been associated with development and growth of HCAs [7]. Unlike for HCA, the relationship between the development of FNH and estrogen exposure is not clearly established [8].

Both infectious and noninfectious agents can lead to hepatitis. These processes may eventually cause fibrosis, leading to cirrhosis, compromised liver function, and even death [9]. Globally, an estimated 354 million people worldwide are living with hepatitis B or C, the most common causes of chronic viral hepatitis [10]. Among them are approximately 65 million women of reproductive age with hepatitis B [11]. In the United States, among women of reproductive age, the annual rate of acute hepatitis B infections has been stable nationally but increased in several states during 2011–2017 [12]. Rates of hepatitis C are increasing in the United States among women of reproductive age. Confirmed hepatitis C among women aged 18–44 years doubled during 2006–2014 [13]. Prevention of unintended pregnancy, along with other strategies to improve screening and timely treatment, are important to minimize vertical transmission of hepatitis B and C [14].

The US Centers for Disease Control and Prevention publishes the US Medical Eligibility Criteria for Contraceptive Use (US MEC) [15], which provides evidence-based guidance on the safety of contraceptive methods for patients with certain characteristics or medical conditions, including liver disease. As part of a process to update the US MEC, the objective of this systematic review was to assess whether, among women with liver disease, there is an increased risk of adverse health effects among those using HC (by formulation and type) compared with those who are using non-HC or not using contraception. This review updates two previous systematic reviews on HC safety among those with liver disease [16,17].

This systematic review was conducted according to an a priori protocol, which has been registered in the International Prospective Register of Systematic Reviews under the registration number CRD42019127163 [18] and reported according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines [19].

We screened 5507 articles, from which 164 articles were identified for full-text review (including one article [25] identified from a reference list of an included article; Fig. 1). Among these, 151 were excluded (96 wrong study design, 37 wrong patient population, nine wrong intervention, five wrong outcomes, and four abstract only). A total of 13 articles met the inclusion criteria [25–37]; two of these described data from one study [32,33]. Two articles from the previous reviews were excluded: one for wrong patient population (subjects with previous but not necessarily chronic hepatitis) [38] and one for wrong intervention (unknown HC exposure during follow-up period) [39].

Eight studies (nine articles) reported on women with liver lesions, including three studies (four articles) on women with FNH [26,30,32,33] and five studies on women with HCA [25,27,31,34,36] (Table 1). Among the studies on women with FNH, two (three articles) were comparative cohort studies [26,32,33] and one was a case-control study [30]. Women used COC [26,32,33], POP [32,33], or oral contraception (OC; type not specified but presumed mostly COC) [30]. Outcomes reported were changes in lesion number and size (progression and regression). Among the studies on women with HCA, two were comparative cohort [34,36] and three were noncomparative cohort studies [25,27,31]. Women used COC [31,34], POC (LNG-IUD, implant, depot medroxyprogesterone acetate [DMPA], or POP) [27,34], or OC (type not specified but presumed mostly COC) [25,36]. In one of these studies, some women were using noncontraceptive hormone formulations (e.g., menopausal hormone therapy) or doses (e.g., high-dose progestins for abnormal bleeding), and results were not reported separately for contraceptive formulations or doses; however, approximately 75% of participants were using contraceptive formulations [34]. In two studies, some women used COC containing estrogen at higher doses (≥50 mcg ethinyl estradiol) than those currently used [31–33]. Outcomes reported were lesion changes (progression and regression) and malignant transformation. One of the studies on FNH [30] and all the studies on HCA were newly identified since the previous systematic review [17]. We did not identify studies of HC use among women with malignant liver lesions.

Four studies reported on women with hepatitis [28,29,35,37], including hepatitis B [35,37], hepatitis C [28], or type unspecified [29] (Table 2). One was a nonrandomized trial [37], two were comparative cohort studies [28,35], and one was a noncomparative cohort study [29]. All studies examined COC use, except one study that did not specify OC type, but we presumed it to be mostly COC [35]. In one study, COC contained estrogen at higher doses (50 mcg ethinyl estradiol) or different formulations (mestranol) than those currently used [29]. Outcomes reported were changes in liver enzymes, hepatic fibrosis, and clinical course. All studies were included in the previous systematic review. No evidence was identified on women with cirrhosis, consistent with the previous review [16].

The evidence indicates that FNH and HCA might respond differently to exogenous hormones. Changes in FNH lesions were generally independent of HC use, with lesions demonstrating increases, decreases, or no change during COC or POP use and after COC discontinuation. Studies found that HCA lesions were generally estrogen responsive. HCA lesion progression was more likely among COC users than nonusers, and lesion size generally decreased with discontinuation. HCA lesion changes among POC users were similar to nonusers. The evidence suggested that COC use does not increase progression or severity of disease among women with hepatitis B or C, while there was no evidence on cirrhosis.

The role, if any, of HC in the progression of liver disease to hepatocellular carcinoma remains unclear. As HCA may progress to hepatocellular carcinoma (generally 5% of cases) [41], hormonal impact on HCA size and volume may theoretically alter the risk of progression to cancer. Only one case of HCA transformation to hepatocellular carcinoma was documented in the four studies reporting on this outcome, and it is not clear whether this occurred in a current or former OC user [36].

The new evidence we reported was from studies conducted among women with either FNH or HCA liver lesions. FNH is a relatively common liver lesion that is more prevalent among women than men; however, the role of sex hormones in etiology or development is unclear [8]. One new study on FNH was consistent with two previously included studies in finding that FNH lesions appeared to stabilize or change regardless of HC use or discontinuation [26,30,32,33], which may suggest HC use does not play an important role. However, data from five newly identified studies of women with HCA demonstrated greater progression with estrogen exposure and greater regression with its discontinuation when compared with progestin-only methods or no hormonal use [25,27,31,34,36].

The body of evidence on the safety of HC in women with liver disease has several limitations, and we rated it as very low for all outcomes. Most studies had a high risk of bias. Several studies did not specify contraceptive formulation or examined older formulations. Older formulations of contraceptives, however, included higher estrogen doses; newer formulations with lower doses of ethinyl estradiol or newer estrogen formulations may have less impact on liver function. Generalizability is limited because most studies had small sample sizes of mostly homogenous populations lacking a variety of liver disease etiologies. In addition, our interpretation of several studies is limited by lack of a comparison group.

Drug metabolism may be impaired in patients with significant liver dysfunction [42], with HC use, leading to increased exposure to exogenous hormones. Furthermore, elevated levels of exogenous hormones could theoretically increase risk for hormone-related side effects and adverse events, such as thromboembolism. Estrogen-containing formulations are associated with increased risk of thromboembolism compared with nonuse [43–46]. In addition, DMPA is a relatively higher dose than other POCs, and its use has been associated with a higher risk for venous thrombosis compared with nonuse [47–49]. Future study of thrombosis risk in women with liver disease who use HC is needed.

In conclusion, a limited body of evidence suggested that changes in FNH lesions were independent of HC use. COC use was associated with HCA lesion progression, but generally remained stable with POC use. COC use did not increase progression or severity of disease among those with chronic hepatitis. This information was reviewed for the US MEC 2024 [15].

Appendix A. Supporting information

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.contraception.2025.111012.