Kronos Early Estrogen Prevention Study (KEEPS)
A protocol to watch
Enthusiasts for post-menopausal hormone replacement therapy (HRT) make much of the fact that starting it immediately after the menopause might be cardio-protective, even though most trials show that it is harmful started later in life, or after cardiac disease is established.
No pharmaceutical company will fund trials to test such a speculative hypothesis, but a Phoenix-based charity, the Aurora Foundation, has paid for an outfit called the Kronos Longevity Research Institute to do just that, the KEEPS trial.
Auroroa is a philanthropic vehicle for the billionaire John Spurling to spend money on research into ageing. Nothing wrong with that, except that charity-funded trials tend not to be regulated as tightly as commercial ones, so there is sometimes room for a bit of data dredging. I hope not, but it’s worth being alert. If KEEPS shows any positive results we can anticipate some “told you so” from pharma-funded researchers. One of the investigators Rogerio Lobo is on my name and shame list. The trial is registered here and the trial website here.
The study population is women age 42 to 58, and within three years of the menopause.
According to clinicaltrials.gov the sample size is 728, divided into four arms, 1. oral estrogen and progesterone, 2. placebo, 3. transdermal estrogen and progesterone, 4. placebo. It is unclear whether it will be analysed as two parallel trials, or as one trial with the oral and transdermal routes as planned subgroups. Worryingly the study website states there are three arms, namely the two active treatments and one shared placebo group. If so, it is odd that the sample size is not divisible by three.
The primary endpoint is the rate of change of carotid intimal medial thickness by ultrasound, measured at screening, 12, 24, 36, and 48 months. This is ambiguous and implies at least four primary endpoints. The rate of change between screening and 48 months would be reasonable, so long it is measured without breaking of the blinding.
The trouble is that up to four primary endpoints, and ambiguity over whether we have two parallel trials, a single trial with a planned subgroup analysis, or even a three armed trial leaves multiple chances to get a “significant” result.
There are also no less than seven secondary outcome measures:
- Change in coronary calcium score by X-ray tomography
- Plasma lipid profiles
- Blood clotting factors
- Serum inflammatory factors
- Hormone levels
- Cognitive and affective scores on standard psychometric tests
- Quality of life
All, apart possibly from the first, can be measured in many different ways. Each is also being measured four times – at baseline, 12 or 18, 36 and 48 months.
The trial aimed to complete in 2010. It is behind schedule but the latest clinicaltrials.gov information states that the primary endpoint data will be collected by May 2012.
It would be reassuring to see an analysis plan – if anyone knows of one please let me know. It’s not too late yet, but it will be when the codes are broken.
Jim Thornton
Ripe-tomato.org 
Kronos Early Estrogen Prevention Study (KEEPS)
KEEPS is a desperate attempt to salvage a living spark of some sort of HRT from the funeral pyre which followed the devastating WHI studies and MWS results. The idea of the hormone pushers is to give sequential oestrogens and progesterone at the exactly “right” age of the menopause and hey presto to produce, by careful orchestration, some sort of proof of the much vaunted but erroneous beliefs that HRT prevents heart attacks and senility although increasing the risk of strokes. The MWS has already proved that the risk of breast cancer is greatest when HRT is given nearest to the age of the menopause.1,2
In comparison of the two main large studies, KEEPS is very small with only 729 unusually healthy women carefully selected after a very comprehensive exclusion process with numerous tests. Previous hormone use had to be stopped at least 3 months before and any women who had abnormal test results were excluded. I think, but it is not clear, that 243 women were to be given sequential pills; 243 sequential patches; and the rest either placebo pills or patches. Sequential regimes are more likely to cause irregular bleeding, endometrial hyperplasia with increased risk of endometrial cancer, and therefore a considerable first year discontinuation rate, which will reduce the chance of longer use increasing venous thrombosis, strokes, heart attacks and cancers in this small trial.
In the 1960s the tests I carried out in the London Council for the Investigation of Fertility Control trials of the effects of combined, sequential or progestogen–only regimes, included cervical smears and endometrial biopsies. The biopsies revealed changes in endometrial blood vessels especially in women with vascular side effects, and also sharp changes in enzyme activities. 50 years later, the additional tests most likely to show what is going on in hormone takers are serum zinc and copper levels, red cell superoxide dismutase activity, glutathione peroxidase activity, toxic metal sensitivities and DNA adducts, but none of these important tests will be carried out in the KEEPS study.
1 Beral V, Reeves G, Bull D, Green J, for the Million Women Study Collaborators. Breast cancer risk in relation to the interval between menopause and starting hormone therapy. J Natl Cancer Inst. 2011;103:296–305.
2 Grant EC. Re: Breast cancer risk in relation to the interval between menopause and starting hormone therapy. J Natl Cancer Inst. 2011 Jul 6;103(13):1069; reply 1069-70. Epub 2011 Jun 23
Review lectures are in my website http://www.harmfromhormones.co.uk . My references are at http://www.pubmed.gov grant ec. and http://www.bmj.com Ellen CG Grant
You’re too harsh Ellen.
Mitch Harmon, the KEEPS chief investigator, has clarified the design. It’s a three arm, phase two trial. 1. oral, 2. patch, 3. placebo in the ratio 12/12/13. The primary outcome is the slope of deterioration in carotid intimal thickness. For the primary analysis the two drug arms will each be compared separately with the placebo group. The primary analysis will be by intention to treat and the carotid thickness has been measured and recorded before the blinding was broken.
I’m pestering Mitch about whether all possible ambiguities in the analysis plan have been pre-specified before the codes were broken, but that’s not a bad trial design. I’d be surprised if they see any significant difference. But even if they do, it’s just a mechanistic trial. Anyone who takes HRT on the basis of a positive result from KEEPS, deserves everything they get. The most that a positive result could justify would be a new phase 3 “daughter of WHI” trial in younger women. Personally, I doubt if any funder will rush to spend millions on that.
Save your fire for the paid doctors who misrepresent the evidence on HRT. Not those who collect it. I’ve got a few more of the former to post about soon! 🙂
The name “Early estrogen prevention study” suggests that those who are collecting this evidence expect taking estrogen and progesterone at the menopause to prevent things other that the usual steroid suppression of warning menopausal symptoms. It may have been justifiable to collect evidence decades ago but I can see little justification now. Also intention to treat is rather meaningless as it gives no indication of the effect any use or of longer use. I found the greatest progesterone-induced arteriolar development happened in the women who were most likely to stop within the first year because of headaches and other side effects. My other point is that as long as trials are continuing, it would be very useful if the best tests were used..
Well KEEPS didn’t find what it was looking for (less CIMT). Instead, to sweeten up the results and make them more palatable, they mentioned how HRT did wonders for menopausal symptoms in this trial and that no CVD harm was found, thus making it safe in “early” menopause. Yet, when you look at it, this trial only confirms what’s already known (symptom relief) and lends no support to the timing hypothesis. Moreover, there was a small but significant adverse effect of estradiol on ability to recall past events as well as self-reported ability to remember things. Also, estradiol may have led to more abnormal mammograms. This also pokes holes in the bioidentical argument, that because something is “natural” it must be BETTER!