Oral contraceptives containing estrogen–progestin combinations or progestin-only products are widely used by women of childbearing age. A survey conducted by Roper Starch Worldwide in 1999 suggested that 23–40% of women in the United States aged 18–49 years used oral contraceptives1, whereas the Ortho Annual Birth Control Study conducted by Ortho-McNeil Pharmaceutical Inc. indicated that 23% of women in the United States used oral contraceptives in 19992. The United Nations Population Fund estimated that, given adequate access to services and supplies, approximately 76.6 million women worldwide will use oral contraceptives annually by the year 20053. According to the World Health Organization (WHO), approximately 50% (19.2 million) of all adults living with AIDS are women4. Oral contraceptives (OCs) are commonly used by HIV-infected women to prevent pregnancy, normally combined with condom use. However, approximately 4–5% of HIV-infected women still use oral contraceptives as a sole birth control method5,6. Because of the multitude of other medications used by this group to treat HIV and its complications, drug–drug interactions are likely and can result in an increase or decrease of plasma concentrations of oral contraceptives and other drugs.
OCs are eliminated partly through metabolism. The estrogen component of the OC, ethinyl estradiol (EE), is a substrate of the CYP3A4 enzyme. In addition, it undergoes glucuronide and sulphate conjugations and enterohepatic recycling. Most progestins have similar metabolic pathways as EE. Some progestins (e.g. norethindrone and levonorgestrel) bind to sex hormone binding globulin.
Among HIV-related drugs, non-nucleoside reverse transcriptase inhibitors, protease inhibitors, and some drugs used for OI are inducers or inhibitors of CYP3A4 and glucuronyl transferase. Therefore, when these drugs are administered concomitantly with OCs, changes in oral contraceptive plasma concentrations can be anticipated. Although the exposure–response relationships of OCs are not fully understood, a reduction in OC plasma concentrations can potentially result in unintended pregnancy or breakthrough bleeding. Therefore, alternative contraceptive methods should be considered during HIV therapy with agents that induce OC's metabolism. Table 1 summarizes the marketed HIV-related drugs that reduce OC plasma concentrations.
Although the clinical significance of increased OC plasma concentrations is unclear, the potential increased risk of side-effects associated with OCs should be considered. Higher doses of EE have been associated with an increased risk of cardiovascular adverse events and hepatic adenomas. Higher doses of progestin have been associated with decreased HDL or increased insulin resistance, particularly in diabetic women. Table 2 shows the geometric mean increase in OC plasma concentrations caused by the co-administration of marketed HIV-related drugs.
The effect of OCs on highly active antiretroviral therapy (HAART)/OI drugs has not been studied very often. EE is a weak CYP3A4 and 2B6 inhibitor. EE may also induce conjugation. Studies in rats indicated that EE is an inducer of P-glycoprotein. However, no data have shown whether EE is also an inducer of P-glycoprotein in humans. A drug interaction study between amprenavir and OCs showed that EE combined with norethindrone caused a 22 and 20% reduction in amprenavir area under the plasma concentration curve (AUC) and Cmin, respectively14. However, the mechanism is unknown. Gestodene (not available in the United States), a progestin, is also a CYP3A4 inhibitor16.
Whereas transporters have gained attention in recent years for their effects on drug pharmacokinetics and drug–drug interactions, it is not clear whether OCs are substrates of transporters (e.g. P-glycoprotein) or if they can affect a transporter's affinity for other drugs.
The study design for drug–drug interaction studies between OCs and other drugs is unique because physiological changes during menstrual cycles can result in pharmacokinetic and pharmacodynamic variability. When evaluating the pharmacokinetic/pharmacodynamic effect of antiretroviral drugs or drugs for OI on OCs, the following components of study design are recommended: (1) multiple doses of HAART/OI drugs and OCs are preferred; (2) the highest dose of HAART/OI drugs should be used; (3) a crossover design is ideal; (4) two menstrual cycles may be sufficient to evaluate the pharmacokinetic effect (one cycle for OCs only, one cycle for OCs plus HAART/OI drugs); (5) three menstrual cycles may be needed if the pharmacodynamic effects are evaluated for a potential metabolism inducer to allow sufficient time for induction (one cycle for OCs and two cycles for OCs plus HAART/OI drugs); (6) monitoring follicle stimulating hormone, luteinizing hormone, progesterone, and sex hormone binding globulin levels is recommended to gain insight into a further understanding of the drug interaction; (7) women who enter the study should already be stabilized on the similar/same combination of OCs.
The use of single-dose OCs in interaction studies is also acceptable. However, if there are changes in OC concentrations, then a multiple dose study is recommended to investigate further the consequences (e.g. breakthrough bleeding, pharmacodynamic endpoints) and the magnitude of the change. Concerns have been raised regarding whether one cycle of HAART/OI drugs combined with OCs is long enough to see the pharmacodynamic effects of a potential CYP3A4 or conjugation inducer on oral contraceptives.
The effect of OCs on HAART/OI drugs can be evaluated in a separate study or combined in the study that evaluates the effect of HAART/OI drugs on OCs. A sufficient washout between the period of HAART/OI drugs alone and the period of HAART/OI drugs combined with OCs is needed to avoid the carryover effect.
In summary, OCs are widely used by HIV-infected women to prevent pregnancy. Because of the multitude of drugs they are taking, potential drug–drug interactions between OCs and other drugs (e.g. antiretroviral drugs or drugs used for OI) may affect the efficacy/safety of OCs and HAART/OI drugs. In order to understand the clinical significance of the OC plasma concentration changes, exposure–response relationships of OCs need to be better evaluated. Drug–drug interaction studies need to be well designed and conducted. Drug–drug interaction information should be included in drug product prescription information so the information can be conveyed to the healthcare provider and patients.
1. Roper Starch Worldwide. Most pill users would prefer monthly method. New York Reuters Health; May 1999.
2. Ortho Annual Birth Control Study, 1999. Raritan, NJ: Ortho Pharmaceutical Inc; 2000.
3. United Nations Population Fund. Contraceptive Requirements and Demand for Contraceptive Commodities in Developing Countries in the 1990s
. New York United Nations; 1994.
4. AIDS Epidemic Update, December 2002. Geneva: World Health Organization; 2002.
5. Belzer M, Rogers AS, Camarca M, et al
, and the Adolescent Medicine HIV/AIDS Research Network. Contraceptive choices in HIV infected and HIV at-risk adolescent females. J Adolesc Health 2001; 29(3 Suppl.):93–100.
6. Magalhaes J, Amaral E, Giraldo PC, Simoes JA. HIV infection in women: impact on contraception. Contraception 2002; 66:87–91.
prescription information. Ridgefield: Boehringer Ingelheim Pharmaceuticals, Inc; 2002.
prescription information. Kalamazoo: Pharmacia and Upjohn Company; 2002.
prescription information. San Diego: Agouron Pharmaceutics, Inc.; 2002.
prescription information. North Chicago: Abbott Laboratories; 2001.
prescription information. North Chicago: Abbott Laboratories; 2003.
prescription information. Kansas City: Aventis Pharmaceuticals; 2000.
prescription information. New York Bristol-Myers Squibb Company; 2002.
prescription information. Whitehouse Station: Merck and Co., Inc.; 2003.
prescription information. New York Pfizer; 1998.
16. Guengerich FP. Mechanism-based inactivation of human liver microsomal cytochrome P-450 IIIA4 by gestodene. Chem Res Toxicol 1990; 3:363–371.