Oral Contraceptives and Ovarian Cancer

Abstract & Commentary

Synopsis: Oral contraceptives protect against ovarian cancer, but it is not clear whether specific progestins are more effective than others.

Source: Schildkraut JM, et al. J Natl Cancer Inst. 2002;94:32-38.

Schildkraut and colleagues from the duke University Medical Center conducted a case-control study examining the association between the risk of ovarian epithelial cancer and the specific estrogen and progestin components of oral contraceptives. The 390 cases and 2865 controls were obtained from the Cancer and Steroid Hormone Study (CASH). The cases in this multicenter study were diagnosed in a 2-year period, from December 1980 to December 1982. The results confirmed older studies finding that combined estrogen-progestin oral contraceptives (OCs) protect against ovarian cancer, but Schildkraut et al further concluded that formulations with a high-progestin potency produce a greater reduction in risk compared with low-progestin potency products. The results were expressed by using high-progestin potency as the referent; thus low-progestin potency products had a 2.2 odds ratio (CI = 1.3-3.9), a greater risk of ovarian cancer (based on 22 cases with high potency, and 82 cases with low potency). Analysis by duration of use indicated increasing protection with increasing duration of use, with a slightly better effect associated with high-potency products. Schildkraut et al concluded that OCs with a high potency progestin provide greater protection against ovarian cancer, and suggest that this is due to a direct biologic effect other than inhibition of ovulation.

Comment by Leon Speroff, MD

Schildkraut et al were partly motivated to perform this study because they had previously published the results of a study in monkeys that the progestin component of OCs produces a direct apoptotic effect on the ovarian epithelium.1 They were pleased with the results of their case-control study; therefore, because they supported their contention that progestin exerts a beneficial effect beyond inhibition of ovulation. However, the results are not so clear cut.

The conclusion depends upon the validity of assigning progestin potency to the different formulations. It is difficult to make this exercise clinically meaningful because potency varies depending upon the target organ and end point being studied. The 2 methods chosen were "delay of menses" and "subnuclear vacuolization in human endometrium." Differences in these assays may have little meaning in a different target tissue—in this case the ovary. Furthermore, animal responses differ compared with human responses. Finally, potency differences in various progestins have been accounted for clinically by appropriate adjustments in the administered doses.

Schildkraut et al acknowledge that there exists in the literature another case-control study of comparable size (actually twice as many cases) and design with a different conclusion (these are the 2 best studies on the subject). Ness and colleagues reported in 2000 that they could not detect a difference in the reduced risk of ovarian cancer comparing women who used higher dose pills before 1972 with women who used lower dose pills after 1980.2 In this study, both high progestin products and low progestin products had the same 50% reduction in ovarian cancer risk, and the case numbers were 49 and 140, respectively (compared with 22 and 82 in the current report). Both studies used the same methods to assess potency and the same technique to ascertain which products were used.

Schildkraut et al in this current report state in their discussion that "newer formulations have had lower potencies and, therefore, are likely to have a reduced effect on the risk of ovarian cancer." I strongly believe it is inappropriate to make such an important clinical conclusion based on these data. There is no way to know if the results of this most recent study are real or reflect the case numbers and/or the categorization and labeling of products into high and low potency. Furthermore, a case-control study of even larger size and similar design fails to reach the same conclusion. I’m sure the authors were delighted to have support for the biologic effects of progestins demonstrated in their monkey study, and this does lend biologic plausibility to their case-control results, but these results are not sufficient to influence clinicians as to which OCs should be prescribed.

The literature on OCs and the risk of ovarian cancer supports the following important conclusions:

  • OCs reduce the risk of ovarian epithelial cancer, about 40% overall;
  • The degree of reduction increases with increasing duration of use;
  • The protective effect is maintained for as long as 30 years after discontinuing use;
  • The protective effect is independent of age at initiation, young or old;
  • The protective effect is independent of the dose of estrogen.


1. Rodriguez GC, et al. J Soc Gynecol Investig. 1998;5: 271-276.

2. Ness RB, et al. Am J Epidemiol. 2000;152:233-241.

Dr. Speroff is Professor of Obstetrics and Gynecology, Oregon Health Sciences University, Portland.