and Nordette are the same, one might predict the 3-keto-desogestrel formulation (Marvelon) would be markedly less androgenic. This speculation is confirmed in the literature in studies that examine changes in lipids or SHBG.

In a direct comparison of Marvelon and Femodene, there were no differences in 19 parameters of lipid metabolism, gonadotropins, prolactin, ovarian and adrenal steroids, and SHBG (Jung-Hoffmann et al., 1988). Like Femodene and Marvelon, the Cilest formulation appears to have little impact on the overall lipid profile, and any reported changes are in a positive direction (e.g., high-density lipoprotein [HDL]-cholesterol increases slightly and there is an improved HDL/ low-density lipoprotein [LDL] ratio).

An oral contraceptive containing progestogen and melatonin is presently undergoing phase 1 and 2 clinical trials (with phase 3 trials scheduled to begin in early 1991) in the Netherlands. If contraceptive effectiveness can be demonstrated, large-scale use may provide appropriate observational data to test the hypothesis (Cohen et al., 1978) that melatonin protects against breast cancer in perimenopausal women. The combination of hypothalamic-releasing agents with low doses of progestogen has also been suggested as a possible contraceptive modality that would reduce the incidence not only of ovarian and uterine cancer but of breast cancer (Pike et al., 1989). There is insufficient clinical experience to confirm or refute these hypotheses.

IMPLICATIONS FOR RESEARCH

Most of the epidemiological studies to date have provided information exclusively on long-term use of older oral contraceptive formulations that had high total steroid doses. Limited information is available on long-term use of lower-dose formulations, particularly as related to breast cancer, and nothing is known about the triphasic formulations. The new progestin formulations will further complicate the study of the relationship of long-term use of the pill from an early age and breast cancer.

Three important considerations with respect to studies of the potential link of oral contraceptives (and specific oral contraceptive formulations) and breast cancer loom large for the 1990s. First, there is wide variation among individuals in blood levels of ethinyl estradiol and the progestin component of an oral contractive following oral administration (Goebelsmann, 1986; Kuhl et al., 1988, Jung-Hoffman et al., 1988; Goldzieher, 1990). Given this individual variation in absorption and metabolism for each component, the average estrogen/ progestin ratio will vary for each individual and may be different



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement