The Raloxifene RUTH Trial Subgroup Analysis
The Raloxifene RUTH Trial Subgroup Analysis
Abstract & Commentary
By Leon Speroff, MD, Editor, Professor of Obstetrics and Gynecology, Oregon Health and Science University, Portland, is Editor for OB/GYN Clinical Alert.
Synopsis: In women younger than age 60 who were at increased risk of coronary events, raloxifene reduced the incidence of coronary events.
Source: Collins P, et al. Effects of the selective estrogen receptor modulator raloxifene on coronary outcomes in the Raloxifene Use for The Heart trial. Results of subgroup analyses by age and other factors. Circulation 2009;119:922-930.
The raloxifene use for the heart (RUTH) trial was an international randomized, placebo-controlled trial that enrolled 10,101 postmenopausal women in two groups, one at high risk for coronary heart disease, and one with established coronary heart disease. The women in each group were randomized to raloxifene, 60 mg/day, or placebo. In the initial report from the RUTH trial, no overall effect on coronary events was detected and there was an increase in strokes.1 Because the Women's Health Initiative (WHI) and Nurses' Health Study reported a reduction in coronary events associated with estrogen therapy administered to young postmenopausal women, the RUTH trial performed a post-hoc analysis of the impact of raloxifene according to age of the women at entry to the study as well as subgroups such as the use of medications, including statins.2 Overall, raloxifene did not increase or decrease coronary events in either of the treated groups. The only category demonstrating a significant difference, a 40% reduction in coronary events, consisted of women younger than 60 years of age.
Commentary
Raloxifene was expected to exert a preventive reduction in coronary heart disease because it lowered total cholesterol and LDL-cholesterol and induces coronary artery relaxation. In a 2-year randomized trial in monkeys, raloxifene exerted no protection against coronary artery atherosclerosis despite changes in circulating lipids similar to those achieved in women (increases in HDL-cholesterol and decreases in LDL-cholesterol).3
Despite the statistically significant reduction in coronary events in women younger than age 60, there was no relationship in any subgroup according to years since menopause, even in the group less than 10 years postmenopausal. The women who were younger than 60 years of age were an average of 9.9 years since menopause, compared with 19.4 years for the overall study population. The finding of a beneficial effect of raloxifene in the youngest postmenopausal women in the RUTH trial does not jibe with the failure to find a relationship with years since menopause.
The RUTH investigators pointed out that no evidence for coronary harm was found with raloxifene in the oldest age groups, contrary to the increase observed in the WHI when estrogen was administered to women older than age 70 (a similar age analysis of stroke in the RUTH trial would be of interest). They speculate that this may be because raloxifene does not stimulate a proinflammatory response as does estrogen, citing the lack of raloxifene effect on C-reactive protein (CRP) levels. However, CRP is a protein synthesized in the liver and atherosclerotic arteries. An estrogen-induced increase in CRP levels may be due to estrogen's well-known effect to stimulate the hepatic synthesis of proteins, especially because of the first-pass phenomenon with oral administration. Studies with multiple inflammatory markers report that oral estrogen therapy increases only CRP, the only marker synthesized in the liver. In fact, while increasing CRP, oral hormone therapy reduces the circulating levels of other inflammatory markers with inconsistent effects on interleukin-6. Most importantly, we don't know if the decrease in CRP levels with statins and the increase with estrogen (if this is indeed the case) are instrumental in clinical outcomes or reflect other effects. Thus, raising or lowering CRP levels will not necessarily increase and decrease the risk of clinical disease. A study from the Women's Health Initiative confirmed the correlation between baseline levels of CRP and an elevated risk of coronary heart disease, but the increase in CRP induced by oral hormone therapy did not further increase the risk.4
Two-thirds of the women in the RUTH trial were being treated with statins. It is possible that a harmful effect of raloxifene was masked by statin treatment. We should not rely on the RUTH trial to convince us that raloxifene treatment is safe for older postmenopausal women, free of an adverse effect on arterial thrombosis.
The well-written editorial that accompanies the RUTH report was written by Janet Wittes, PhD, a statistician.5 She addressed the likelihood that if enough subgroups are analyzed, some mathematical finding will emerge that appears to be fact. The first problem is the difficulty in achieving sufficient numbers in each subgroup. Of the 10,101 women, there were only 360-460 women (14%-18%) in each of the patient groups who were younger than age 60, and only 134 women younger than age 60 experienced a coronary event. This one subgroup demonstrated a statistically significantly different conclusion than the overall finding, of 51 analyzed subgroups.
So, which conclusion is more important and more reliable: that raloxifene has no overall effect on preventing coronary events, or that raloxifene reduces the risk of coronary events in the youngest postmenopausal women? The statisticians say that the overall conclusion, the primary objective of the designed trial, is the most reliable. But the beneficial finding in young postmenopausal women is consistent with the analyses of the WHI and the Nurses' Health Study, lending credibility to the age subgroup analysis. Wittes pointed out that in the age groups defined by the original study design, there were no differences by age. Only in this post-hoc analysis, stimulated by the desire to match the recent age reports from the WHI, did a single subgroup benefit appear. Furthermore, Wittes agrees with my point about the failure to demonstrate a difference according to years since menopause, making the single subgroup conclusion either a chance finding or an effect smaller than applied mathematics in the report suggest. Wittes argues that analyses of subgroups that consist of less than 30% of the study population are not reliable (in the RUTH trial, the younger than age 60 years subgroups were 14%-18% of the total number).
In conclusion, a decision to use raloxifene should not be influenced by its effects on the cardiovascular system. This is a bone and breast decision.
References
- Barrett-Connor E, et al; Raloxifene Use for The Heart (RUTH) Trial Investigators. Effects of raloxifene on cardiovascular events and breast cancer in postmenopausal women. N Engl J Med 2006;355:125-137.
- Collins P, et al. Effects of the selective estrogen receptor modulator raloxifene on coronary outcomes in the Raloxifene Use for The Heart trial: Results of subgroup analyses by age and other factors. Circulation 2009;119:922-930.
- Clarkson TB, et al. Lack of effect of raloxifene on coronary artery atherosclerosis of postmenopausal monkeys. J Clin Endocrinol Metab 1998;83:721-726.
- Pradhan AD, et al. Inflammatory biomarkers, hormone replacement therapy, and incident coronary heart disease: Prospective analysis from the Women's Health Initiative observational study. JAMA 2002;288:980-987.
- Wittes J. On looking at subgroups. Circulation 2009;119:912-915.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.