An Overview

Robert L. Barbieri

Department of Obstetrics, Gynecology and Reproductive Biology, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts 02115


    Introduction
 Top
 Introduction
 The quality of the...
 Benefit-risk assessments are...
 The leadership role of...
 References
 
IN THIS Clinical Controversy, Drs. Speroff, Walker, and McPherson review the possible association between the use of oral contraceptives containing the synthetic progestins gestodene or desogestrel and the risk of venous thromboembolism (VTE). Speroff interprets the available data as indicating that all oral contraceptives are associated with a small increase in the relative risk of VTE, and that there are no differences in the risk of VTE among various contraceptives containing second (norgestrel) or third (gestodene or desogestrel) generation progestins. In contrast, Walker and McPherson interpret the data as indicating that the use of oral contraceptives containing the third generation progestins are associated with a greater risk of DVT. Two major issues were discussed by all three experts: 1) how to assess the quality of the evidence; and 2) developing a meaningful framework for integrating an analysis of the benefits and risks of a contraceptive hormone. In this overview, I would like to discuss these two issues and then reflect on the leadership role the endocrinologist must take to better define the pharmaco-endocrinology of gestodene and desogestrel.


    The quality of the study design influences the quality of the data
 Top
 Introduction
 The quality of the...
 Benefit-risk assessments are...
 The leadership role of...
 References
 
Most clinical investigators believe that, for any experiment, the research design is the single, most critically important variable in determining the quality of the data generated and the power of the study to isolate cause-effect mechanisms. Most clinical investigators believe that the prospective, controlled clinical trial with randomization of subjects is the study design most likely to generate reliable, high quality data that can yield generalizable, cause-effect insights. Unfortunately, no prospective randomized studies have been performed that explore the relative risk of VTE in users of oral contraceptives that contain gestodene or desogestrel compared with those containing norgestrel or norethindrone. The absence of data from high quality clinical trials forces clinicians to rely on information from epidemiological studies to explore this association.

As reviewed by Walker and McPherson, case control studies have suggested an association between the use of oral contraceptives with gestodene or desogestrel (third generation progestins) and an increased risk of DVT compared with oral contraceptives that contain norgestrel (second generation progestin). Walker and McPherson believe that the reported case-control studies adequately control for both confounding and bias and that the data is reliable. In contrast, Speroff suggests that the case-control studies reported in 1995 and 1996 had design problems that limit the reliability of the data. Speroff believes that the initial case-control studies did not control for important clinical differences between the users of second and third generation oral contraceptives. Speroff cites subsequent case-control studies that controlled for duration of oral contraceptive use and/or that were focused on first time users. These studies found no differences between second and third generation oral contraceptives and the risk of VTE. For example, using the German MediPlus data base, Farmer and colleagues (1) reported that, among women on oral contraceptives who were treated with an anticoagulant for VTE, there was no increased use of the third generation progestins. Supporting these results are findings by Suissa and colleagues (2), which indicate that, in a case-control study limited to first time users and controlled for the duration of oral contraceptive use, there was no difference in the risk of VTE with second or third generation oral contraceptives. Interestingly, Suissa and colleagues (2) and Lidegaard and colleagues (3) both reported that the risk of VTE among oral contraceptive users was greatest in the first year of use and decreased markedly with extended use. These studies demonstrate the importance of controlling for duration of use in studies that explore the association between hormones and VTE. My assessment of the literature is that there is no convincing evidence of an increase of VTE in users of third generation versus second generation oral contraceptives.


    Benefit-risk assessments are extremely complex
 Top
 Introduction
 The quality of the...
 Benefit-risk assessments are...
 The leadership role of...
 References
 
A second major focus of the commentaries of Speroff, Walker, and McPherson is a discussion of the proper framework for analyzing the benefits and risks of contraceptive hormones. Benefit-risk analysis is inherently complex because it is difficult for the human mind to balance the total effect of a large number of benefits and a small number of risks. Many studies suggest that humans tend to overestimate the probability that a rare event (such as winning the lottery, or a major asteroid striking the earth) might actually occur to them. As noted by McPherson, the potential risk of VTE associated with the use of oral contraceptives containing gestodene or desogestrel is far less than the risks associated with unintended pregnancy or the risks associated with the use of tobacco products. The oral contraceptives that contain gestodene or desogestrel have many health benefits, including a reduced risk of pregnancy, ovarian cancer, and endometrial cancer. In one cohort study, long-term use of oral contraceptives reduced the risk of endometrial cancer by 90% and the risk of ovarian cancer by 60% (4). From my perspective, the benefits of oral contraceptives that contain gestodene or desogestrel clearly far outweigh the risks of the agents. Speroff discusses the importance of clinical judgment in weighing the benefits and risks of an oral contraceptive. For example, he notes that clinicians should be careful about prescribing oral contraceptives to women with a previous history of idiopathic VTE or a close family history (parent or sibling) of VTE.

It is important for all clinical investigators to minimize the possibility that a rare potential risk is overemphasized by the media, resulting in unintended adverse public health consequences. The British Pregnancy Advisory Service reported that media coverage of a potential link between the third generation progestins and VTE was followed by a 10% increase in the number of abortions performed (5). Public health would have been better served if the media had reported the harmful effects of tobacco and emphasized the beneficial effects of all oral contraceptives.


    The leadership role of the endocrinologist
 Top
 Introduction
 The quality of the...
 Benefit-risk assessments are...
 The leadership role of...
 References
 
During the recent debate on the effects of the second and third generation oral contraceptives, epidemiologists published far more reports than clinical endocrinologists. However, it is the clinical endocrinologist, not the epidemiologist, who has the depth and breadth of knowledge of steroid biochemistry, molecular and cellular endocrinology, and endocrine physiology to scientifically design the critical experiments in this arena. Clinical endocrinologists need to be leaders both in investigating the potential physiological effects of combination estrogen-progestin medications and in communicating the findings of their studies to the public. The liver is the largest endocrine target organ, and given the powerful research tools available to the endocrinologist, additional investigations of the effects of novel estrogen-progestin combinations on liver production of clotting factors, both in vitro and in vivo, is timely and necessary.


    Footnotes
 
Address correspondence and requests for reprints to: Robert L. Barbieri, Department of Obstetrics and Gynecology, ASB1-3, Brigham and Women’s Hospital, 75 Francis Street, Boston, Massachusetts 02115.

Accepted March 3, 1999.


    References
 Top
 Introduction
 The quality of the...
 Benefit-risk assessments are...
 The leadership role of...
 References
 

  1. Farmer RD, Todd JC, Lewis MA, MacRae KD, Williams TJ. 1998 The risks of venous thromboembolic disease among German women using oral contraceptives: a database study. Contraception. 57:67–70.[CrossRef][Medline]
  2. Suissa S, Blais L, Spitzer WO, Cusson J, Lewis M, Heinemann L. 1997 First-time use of newer oral contraceptives and the risk of venous thromboembolism. Contraception. 56:141–6.[CrossRef][Medline]
  3. Lidegaard O, Edstrom B, Kreiner S. 1998 Oral contraceptives and venous thromboembolism. A case control study. Contraception. 57:291–301.[CrossRef][Medline]
  4. Vessey MP, Painter R. 1995 Endometrial and ovarian cancer and oral contraceptives—findings in a large cohort study. Br J Cancer. 71:1340–2.[Medline]
  5. Dillner L. 1996 Pill scare linked to rise in abortions. BMJ. 312:996.[Free Full Text]