Progestins and Venous Thromboembolism
Abstracts & Commentary
Synopsis: Two case-control studies suggest that the therapeutic use of progestins is associated with an increased risk of VTE, but this conclusion is weak and should be questioned.
Sources: Poulter NR, et al. Lancet 1999;354:1610. Vasilakis C, et al. Lancet 1999;354:1610-1611.
Two case-control studies, one using data from the WHO Collaborative Study and one using data from the U.K. general practices research database, assessed the risk of idiopathic venous thrombosis in users of progestins alone. The WHO Study focused on the use of progestins for therapeutic purposes and reported the following results (see Table 1):
| Table 1-Results of the WHO Study | ||
| Cases/Exposed Controls | Adjusted OR | |
| Stroke | 2/6 | 1.13 (0.18-7.13) |
| MI | 2/3 | 0.85 (0.05-15.3) |
| VTE | 3/3 | 5.92 (1.16-30.1) |
On the basis of seven cases and 12 controls who had used therapeutic progestins, Poulter and colleagues conclude that therapeutic progestins alone may be associated with an increased risk of venous thromboembolism (VTE).
Vasilakis and colleagues from Boston University, using the U.K. general practice database, reported these results (See Table 2):
| Table 2-Results Using the U.K. General Practice Database | |||
| Cases | Controls | Adjusted RR for VTE | |
| Progestin-only for contraception | 2 | 26 | 1.3 (0.3-6.8) |
| Progestin-only for therapy | 5 | 20 | 5.3 (1.5-18.7) |
Vasilakis et al conclude that progestin-only used for contraception had no effect on the risk of VTE, but the therapeutic use of progestin-only had a substantial association with an increased risk of VTE. Once again (as with combined oral contraceptives), smoking had no effect on the risk of VTE.
COMMENT BY LEON SPEROFF, MD
These two articles are excellent examples of weak epidemiologic conclusions with a potentially powerful clinical effect. The use of progestins-alone is a standard of gynecologic and contraceptive therapy, and these two reports challenge this standard (unfairly, in my view). The authors of both papers suggest a new caution for the use of progestins for therapeutic purposes in women at increased risk of VTE. Thus, the use of monthly progestins to regulate menstrual bleeding and prevent endometrial cancer in overweight women and the use of depot medroxyprogesterone for contraception in older women with hypertension would be questioned. Let’s take a closer look at the strength of these two case-control studies.
First, the WHO study. Note that the confidence interval (CI), while indicating statistical significance for the risk of VTE, has a wide spread (1.16-30.1). This is because of the small numbers that are the basis for the calculation, only three cases and three controls. Thus, the conclusion is imprecise and weak. Poulter et al bolster their conclusion by referring to a previous WHO report. I want to provide the exact quote: "More likely they (the results) suggest, along with previous observations on progestin-only contraceptives (POPs), that progestagens-alone may be associated with increased risk of VTE." The previous observations referred to are in the 1998 WHO report on cardiovascular disease and use of oral and injectable progestin-only contraceptives.1 This case-control analysis could not find a statistically significant increased risk of stroke or myocardial infarction, and for VTE, the results were as follows (See Table 3):
| Table 3-Results of 1998 WHO Report on Cardiovascular Disease and Use of Oral and Injectable Progestin-Only Contraceptives | |||
| Cases | Controls | Adjusted OR for VTE | |
| Oral progestins | 21 | 64 | 1.74(0.73-3.99) |
| Injectable progestins | 11 | 34 | 2.19 (0.66-7.26) |
Note that both of these conclusions are not statistically significant, a fact not mentioned by Poulter et al in their report.
The report based upon the U.K. general practice database suffers from the same problem, small numbers and a wide CI. Vasilakis et al appropriately point out that when progestins are used therapeutically, the population probably had some differing characteristics and many of these women may be treated with relatively high doses of progestins. The only adjustment for differing characteristics in these reports was for body mass index, a recognized risk factor for VTE.
These epidemiologic conclusions are based on extremely small numbers, have wide CIs, and are weak. Patients who receive progestin-only for therapeutic reasons are probably older and are more likely to have family histories of cardiovascular disease. In my view, a problem of preferential prescribing is probably present in that clinicians are more likely to promote the use of progestin-only for women they perceive to be at greater risk of VTE. Thus, it is likely that the case groups represent a higher risk group than the control groups in these reports.
I am disturbed that these reports might successfully challenge the important use of progestins for therapeutic purposes in gynecologic practice. I believe the weakness of the conclusions and the criticisms I have noted allow clinicians to conclude that while these epidemiologic reports achieved statistical significance, they are not clinically meaningful. These reports should not change current clinical practice.
The following statements are true of progestins and the risk of cardiovascular diseases except:
a. Although limited by small numbers, the available evidence indicates that progestins used alone for contraception are not associated with an increased risk of stroke or myocardial infarction.
b. Although limited by small numbers, the available evidence indicates that progestins used along for therapeutic purposes are not associated with an increased risk of stroke or myocardial infarction.
c. The evidence linking the use of progestins alone and the risk of venous thromboembolism is statistically significant in all studies and should be seriously considered.
d. Current clinical practice should not be altered on the bases of the published results.
Reference
1. WHO Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Contraception 1998;57:315-324.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content