FDA-funded Study: Combined Hormonal Contraceptives (CHCs) and the Risk of Thromboembolism and Other Cardiovascular Events

Several contraceptive products approved in the early 2000’s quickly became very popular forms of contraception, particularly among young women, and were considered relatively safe because of their lower estrogen content compared to older hormonal products. Safety concerns for serious thrombotic and thromboembolic adverse events as well as death, however, became a public concern soon after their market introduction. This concern stimulated adverse event reporting, which made appropriate review and interpretation of the reports received in FDA’s Adverse Event Reporting System (AERS) challenging. Despite a low incidence of venous thromboembolic events in this population, an increase in risk for these adverse events among users could put many young women at risk of a major life-altering event. FDA was concerned about its ability to interpret the information available in the AERS database.
Although several epidemiologic studies were initiated by the manufacturers of these new contraceptive products at the request of U.S. and European regulatory authorities around the time of their approvals, the studies were designed mainly to evaluate one specific product compared to one or a group of other contraceptive products. In addition, FDA had concerns about the ability of some of the epidemiologic studies to identify all deaths (including sudden deaths) that could be related to these products.
For these reasons, FDA sponsored an epidemiologic study involving insurance claims and medical record data. The objective of the FDA-funded study was to assess cardiovascular risks, including the risk of thrombotic and thromboembolic events and death, for several newly approved (early 2000s) products that had sufficient numbers of users to allow for an evaluation of these risks compared to those associated with use of older, more frequently prescribed contraceptives at the sites selected. If an increased risk was observed, the FDA-funded study would subsequently attempt to assess user characteristics and prescribing patterns that might help explain the increased risk. It was recognized at the time that a more in-depth assessment of potential reasons for increased risk would not be possible using only claims and electronic medical records and would require physician and patient contact, something that could be conducted later if needed.
The FDA-funded study was conducted at two HMO sites and two state Medicaid programs, each associated with an academic institution. In addition to having access to data from a large group of young reproductive age women, reasons for selecting study sites included: 1) the ability of the investigators to validate study outcomes with medical records; 2) the ability of the sites to link to state vital status files to identify deaths quickly; and 3) the ability of the sites to facilitate physician and patient contact if and when needed. Diversity of populations was favored over similarity to maximize capture of possible reasons for an increased risk if observed.
The FDA-funded study was designed as a retrospective cohort study of women age 10 to 55 years who were current users of the study contraceptives between January 1, 2001 and December 31, 2007. Two exposure cohorts, one of current users and the other of new users, were created for evaluation. For each study contraceptive in the primary analysis, the comparison group included a composite of frequently prescribed products that contained the progestins levonorgestrel, norethindrone, or norgestimate combined with 0.02 mg to 0.035 mg of ethinyl estradiol. As a secondary analysis following recent published studies, comparisons of the risk for serious thrombotic and thromboembolic events were also made for each study contraceptive with the levonorgestrel products containing 0.03 mg estrogen.
Final study results are under review and will be discussed at a joint meeting of the Reproductive Health Drugs Advisory Committee and the Drug Safety and Risk Management Advisory Committee on December 8, 2011.