KEEPS: The Question It Asked But Didn’t Answer

08 Oct KEEPS: The Question It Asked But Didn’t Answer

In 2005, three  years after the famous (and famously dubious) Women’s Health Initiative study of the risks and benefits of hormone replacement therapy, clinical researchers at nine centers around the U.S. began a smaller but more focused study known as KEEPS. It concentrated on a question the WHI left unanswered: Does HRT slow the progression of cardiovascular disease in post-menopausal women?

The results were announced last week. They yielded some interesting and potentially valuable information. But what KEEPS didn’t do was answer that central question. And that’s a very big disappointment, given the buildup of expectations that surrounded its release at the annual meeting of the North American Menopause Society in Orlando, Florida.

UPDATE: A Danish study published by the British Medical Journal a few days after the KEEPS report found HRT cut heart disease by 50 percent without an increased cancer risk.

KEEPS stands for Kronos Early Estrogen Prevention Study. It was conducted over four years by the non-profit Kronos Longevity Research Institute. Here’s how the institute’s director, Dr. S. Mitchell Harman, summed up the key cardiovascular results: “[S]ome measures showed slight evidence that hormone therapy might be cardio-protective in this age group, although results were not definitive and would require additional study.”

I was at the meeting, and I can’t say I was especially surprised as I listened to the researchers describe their methods and results. Though I did find them somewhat ironic. The problem with the WHI study was that its 16,000 subjects were on average too old and too unhealthy to provide meaningful answers to women considering hormone replacement as they enter  menopause. The problem with KEEPS was the opposite: its subjects were on the whole too young and too healthy, especially for a four-year study. The researchers should have either used a broader cross-section of subjects or made the study much longer to measure how hormone replacement affects measures of atherosclerosis.

The study’s subjects were 727 women whose mean age was 52, who were within three years of the start of menopause, and who met fairly stringent standards of health. They were divided into three treatment groups. The first was given estrogen in the form of daily tablets of Premarin and Prometrium for the first 12 days of the month. The second was treated with Climara estradiol patches and cyclical Prometrium. The third group was given either a a placebo patch or pill and placebo Prometrium.

The researchers used two established methods of  measuring atherosclerosis, the accumulation of plaque that thickens the walls of arteries and increases the risk of heart attack and stroke. One scans for coronary artery calcium (CAC); the other, known as CIMT, uses ultrasound to measure thickness of the inner two layers of the carotid artery wall. After four years, there was no difference between the two treatment groups and the placebo group, good or bad, on these measures of the progression of atherosclerosis.

The only finding in either direction was a “non-significant trend toward less accumulation of coronary artery calcium.”  One reason it was “non-significant” might have been that only a small percentage of the women admitted to the study had any coronary calcium to begin with. Some 85 percent of the participants had measurements of zero. For the 15 percent who had some degree of coronary calcium, both the estrogen and estradiol treatments showed a trend toward a reduction. It might have been a hint of something significant. To find out, the trial needed to be bigger and longer.

It’s not that KEEPS was worthless. The news out of the announcement focused on the positives: That hormone replacement safely improves menopausal symptoms including hot flashes and night sweats, depression, diminished libido, and bone density. That’s reassuring to women, and should help continue to reverse the decade-long misinterpretation of WHI data that led many physicians to advise against HRT. But it’s hardly new information. Plenty of previous studies have established these benefits.

Yes, KEEPS adds to our base of knowledge of hormone replacement and cardiovascular disease. But it doesn’t advance the science. It doesn’t begin to answer the question that ultimately matters, the one it set out to resolve. That’s a pretty dismal failure, given the number of research centers involved and the expectations they raised.  And my concern is that the agnostic results will discourage support for the kind of longer, bigger and better study that would prove (or disprove) the relationship between hormone replacement and cardiovascular disease. What researchers really need to do—what KEEPS should have done—is a study on the order of 5,000 subjects, with a broader range of age and baseline cardiovascular health, who are followed for 10 years. The KEEPS investigators are loudly calling for it, but given the scale and investment required, it seems unlikely. In that light, KEEPS might be more than disappointing. It might have done some real damage.

For women of menopause age, this leaves open the question of whether to take hormone replacement in the absence of symptoms. Will it protect their arteries and reduce their risk of cardiovascular disease? And for how long should they take them? As with most questions related to age management, it comes down to what I call the N of 1 concept. In clinical research lingo, N signifies the number of subjects in a study. KEEPS, for instance, had an N of 727. (It would be written as N=727 in a scientific publication). But the only evidence that really matters to any one person is the kind that actually applies to them–which may or may not be the same that applies to those 727 subjects. Even a well-done study with persuasive evidence published by a respected journal might not necessarily apply to you if you are not like the average of the people in that study. Or it might apply to you, but not precisely in the way reported. Thus the N of 1  approach: An individual, evidence-based judgment, based on one’s own biomarkers.

In the case of HRT and cardiovascular and cognitive health, women should have annual CIMT and computerized cognitive testing to measure the effect of what they are doing, whether they are taking hormone replacements or a wait-and-watch approach.

All things considered, the risk-to- benefit ratio favors hormone replacement for symptoms around menopause, probably for as long as you have them. Despite its disappointing findings on cardiovascular benefit, even the KEEPS study confirmed once again that the risk of adverse events from HRT are very small, far outweighed in almost all cases by the many proven quality-of-life benefits.