By Leon Speroff, MD
In the weeks following the publication of the first results from the Women’s Health Initiative (WHI), I have talked by phone and e-mail with clinicians all over the country. It has been a gratifying experience for me to share your thinking and reactions. Sharing with you as you exercise your abilities and obligations as clinicians has made me proud to be your colleague. I have always believed that what sets clinicians apart is the special function called "medical judgment."
A clinician is unique in bringing a special personal relationship to the patient in the process of using his or her store of knowledge. It is not just the pure administration of knowledge. The process requires individualization, applying the knowledge in a modified form based on the clinician’s experience and the clinician’s familiarity and understanding of the individual patient—a process we call "medical judgment." Medical judgment is always based on a foundation of knowledge, the accumulated information and understanding acquired through experience, education, and appraisal of the medical literature. The final effect on a patient is never the result of a single, solitary fact or one scientific study. The entire process is the art and science of clinical medicine, the reason why we enjoy being clinicians and the reason why we are so valued by our patients.
My medical judgment tells me that clinical recommendations should not be made regarding postmenopausal hormone therapy based solely on the published results of the WHI.
The WHI did not answer the following important questions:
- Will postmenopausal hormone therapy begun at or near the time of the menopause and maintained for a relatively long duration of time provide protection against coronary heart disease?
- Are the small increases in cardiovascular events reported by the WHI real or are there other explanations for the reported results?
- Does postmenopausal hormone therapy initiate the growth of new breast cancers?
The design and the length of follow-up of the canceled arm of the WHI did not allow the study to answer the above questions. Women with significant menopausal symptoms were excluded from the study to avoid an exceedingly high dropout rate in the placebo group. For this reason, less than 10% of the subjects were close to their age of menopause (the number is probably even smaller). Therefore the study was not a primary prevention study of cardiovascular disease, but a study of older women who undoubtedly already had a significant degree of atherosclerosis.
The increases in cardiovascular events were not large, and therefore the results are vulnerable to changes if other explanations cause a shift in a small number of cases. Possible mechanisms that could produce such a shift include adjudication of final diagnoses and new statin and aspirin treatment. The published results depend on cardiovascular diagnoses made "in the field." Adjudication of the diagnoses was in process and not completed at the time of publication. Consider the possibility of diagnostic bias. A total of 40.5% of the treated group in contrast to 6.8% of the placebo group was unblinded because of vaginal bleeding. How many patients with presumed cardiac events told the clinician in charge of their care that they were in the WHI study and they were experiencing vaginal bleeding? What was the effect on the clinician’s final diagnosis?
The cardiac and stroke conclusions were dependent upon the events in the first and fifth years. Was there truly an increase in events in the treated group, or were the conclusions the result of new statin or aspirin treatment in the placebo group, producing a falsely high rate in the treated group (we know that statins stabilize atherosclerotic plaques within a few months)?
The apparent increase in breast cancer in the WHI was the result of events in years 4 and 5 of the study. This rapid appearance of breast cancers, together with the observed return to a hazard risk almost to 1.0 in year 6, and no difference in noninvasive breast cancer argue in favor of hormonal stimulation of pre-existing tumors. With longer follow-up of the patients, I anticipate that we will see a better survival rate in the treated group, consistent with what is now a large body of published reports indicating lower grade and stage disease in hormone users. The WHI did agree with convincing evidence in the literature that postmenopausal hormone therapy does not increase the risk of breast cancer beyond that already associated with recognized risk factors; in other words, a positive family history of breast cancer is not a contraindication for the use of postmenopausal hormone therapy.
Finally, the intention-to-treat analysis, favored as the best method of analysis for clinical trials, is handicapped by the high dropout rates in both the treated and placebo groups. The analysis was also affected by drop-ins (hormone treatment initiated by women in the study after the study began). The inability to maintain patients in the study also affects the statistical power of an as-treated analysis.
Multiple publications from the WHI are sure to follow the first report. As the various issues and problems in the study are identified and assessed, will the publicized conclusions change? Given the relatively small magnitude of the conclusions, shouldn’t a more in-depth analysis have been performed before publication (for example, adjudication of cardiac diagnoses before, not after, releasing results to the public)?
The problems associated with the WHI, together with the expense, make it unlikely that clinicians will ever have definitive answers from a properly designed and carried out single clinical trial involving postmenopausal hormone therapy. But when has a single epidemiologic study provided a clear-cut, unequivocal answer? Clinicians must continue to exercise medical judgment, drawing on the large body of knowledge accumulated over the last 20 years, making informed decisions to translate the accumulated knowledge into effective and appropriate clinical practice.
At this point in time, my medical judgment leads me to the following conclusions:
1. Individualized postmenopausal hormone therapy is appropriate for most women. Clinicians and patients together can make an annual decision based upon the individual goals and objectives prioritized by each patient and the assessment of new information without disregarding the large body of knowledge gained over the last 2 decades.
2. Choices from the multiple treatment options available are guided by the goals and objectives of the patient. One study with many problems does not constitute sufficient reason, in my view, to avoid any specific treatment choice.
3. The results of 3 secondary prevention trials and the WHI provide a reasonably solid basis not to recommend postmenopausal hormone therapy for women with existing atherosclerosis in the anticipation of preventing future cardiovascular events. Results from multiple studies indicate that postmenopausal hormone therapy increases the risk of venous thromboembolism, mostly in the first or second year of treatment, almost all (if not all) nonfatal cases, and a risk that is reduced in users of statins or aspirin.
4. Long-term postmenopausal hormone therapy is not precluded by the results reported by the WHI. There continues to be good reason to believe that there are benefits associated with long-term treatment, including improvement of quality of life beyond the relief of hot flushes, maximal protection against osteoporotic fractures, a reduction in colorectal cancers, maintenance of skin turgor and elasticity, and the continuing possibility of primary prevention of coronary heart disease and Alzheimer’s disease. Maximal benefit, however, requires early onset of treatment, near the time of menopause. Maximal benefits of postmenopausal hormone therapy require treatment of healthy target tissues to allow effective response to estrogen and maintenance of health.
In the last few weeks I have become aware of a groundswell force rising from clinicians and their patients in reaction to the WHI. I am convinced that a balance will be restored because of this groundswell coming from clinician-patient dialogues throughout our country. "Medical judgment" will prevail.