LaCroix AZ, Chlebowski RT, Manson JAE, et al. (2011) Health outcomes after stopping conjugated equine estrogens among postmenopausal women with prior hysterectomy. JAMA, 305(13), 1305-1314.
I decided that it would be interesting to read this article on hormone therapy (HT) and health outcomes after finishing “The Female Brain”. Dr. Brizendine describes in her book some of the symptoms associated with the transition to menopause as well as post-menopause. Every woman is different, and Dr. Brizendine suggests that in some cases hormone therapy might be advisable. After working on breast cancer genetic epidemiology for almost 5 years, her statement seemed surprising.
Hormone replacement therapy is a well-established risk factor for breast cancer, and in my little world of breast cancer epidemiology the use of hormone replacement therapy seemed inconceivable. And then I started thinking … the effect size of hormone replacement therapy on breast cancer risk might not be large enough to justify depriving individual women from an easier transition to menopause (less depression, normal sexual drive, fewer fractures, lower risk of CHD, etc.) on a case-by-case basis.
When I looked at my own data, the risk for breast cancer in Latinas who are currently taking HRT was twice as large as that of women who were never under HT (in this case the therapy includes both Estrogen and Progesterone). So OK, maybe for this particular combo the risk is large enough to seriously advocate for no replacement. But does the increased breast cancer risk in the case of estrogen only therapy offset its benefits? The paper by LaCroix et al. recently published in JAMA seems to show that for a particular set of women (postmenopausal 50 to 59 years old with prior hysterectomy), the risk of breast cancer actually decreases if estrogen only therapy is used for a period of about 6 years (this trend was observed during the intervention period). Adding to this, other adverse effects of the therapy were not observed in the treated group over almost 10.7 years of follow up (e.g. risk of stroke, colorectal cancer and total mortality are the same in the treated and control groups).
Do we alter clinical practice based on an article that uses data from a single study, in this case the Women’s Health Initiative (WHI)? An editorial in the same JAMA issue (by Emily S Jungheim and Graham A Colditz) advises a cautious interpretation of the LaCroix results and careful consideration when advising patients. Jungheim and Colditz cite an article published in JNCI in 2011 reporting results of the Million Women Study, which shows that postmenopausal estrogen use negatively influences breast cancer risk. This effect on risk is significant only for those women who initiate therapy within 5 years of menopause, but not for those who start the therapy 5 years or more following menopause. In other words, the results are opposite to those found in the WHI study. Jungheim and Colditz point out that 68% of women enrolled in the WHI were older than age 60 years at randomization and therefore, they question the generalizability of the WHI population’s findings for younger women on HT.
In my opinion, this whole debate clearly reflects that scientific research rarely leads to unquestionable conclusions (OK, I give you tobacco and lung cancer). I think that drawing unequivocal conclusions is particularly difficult in the case of epidemiological studies, which involve samples from different source populations and vary in the particular covariates examined or the way they are measured. This lack of conclusivity may be familiar to researchers, but may not be as obvious to many clinicians who often rely on the “bottom line”, expert opinions and/or national guidelines in treating patients. Currently, many women are probably not being offered estrogen only therapy despite having severe post-menopausal symptoms due to the assumed breast cancer risk increase. But were those women informed about the different studies and their different results? In their review, Jungheim and Colditz conclude: “There may still be a role for short-term use of unopposed estrogen for treating some women with menopausal symptoms, but this role may be vanishing as existing and emerging data continue to be better understood in terms of application to patients. In the meantime, the symptoms of menopause can be significant and require thoughtful management. This would include careful consideration and discussion of the long-term risks and short-term benefits of HT as well as thorough discussion of other treatment strategies and optimization of lifestyle to ensure the best outcomes for women in the many years they should enjoy post-menopause”.
I have to agree with them, but “thoughtful management”, given the complexity of current epidemiologic evidence on the use of hormone therapy in post-menopausal management, may be a daunting task for providers and patients alike.