Framingham Score Predicts Raloxifene's Stroke Risk


ORLANDO — The Framingham stroke risk score can predict a high-risk postmenopausal woman's likelihood of a future cerebrovascular event with raloxifene use, according to new analysis of data from the Raloxifene Use for the Heart (RUTH)study.

Investigators in 26 countries enrolled 10,101 women at risk for a major coronary event in RUTH. A total of 5,031 women had documented coronary heart disease and the remaining 5,070 had multiple coronary heart disease risk factors (Am. J. Cardiol. 2002;90:1204–10). Although overall stroke risk was not significantly different between women randomized to raloxifene versus placebo, a higher number of fatal stroke events occurred in the treatment group, 59, compared with 39 in the placebo group during a mean of 5.6 years follow-up (N. Engl. J. Med. 2006:355:125–37).

Investigators launched a second study to see how this increased risk associated with raloxifene (hazard ratio, 1.49; absolute risk increase, 0.7 per 1,000 woman-years) would apply to women stratified by baseline Framingham stroke scale score.

David Cox, Ph.D., and his associates retrospectively calculated 10-year cumulative risk. He presented findings at the annual meeting of the North American Menopause Society. Eli Lilly & Co. supported the study, and Dr. Cox is a clinical research scientist for the company.

Not surprisingly, risks appeared to congregate in the third and fourth highest quartiles of Framingham score risk. However, there were no significant differences between treatment groups in either all strokes or nonfatal strokes, regardless of baseline Framingham score.

Regarding fatal stroke, Dr. Cox said, “after 2 years, you start to see a split between placebo and raloxifene for risk of fatal stroke by Framingham stroke risk score in RUTH. We think the pattern and data would suggest most of the excess risk was among those in the upper half of risk.” Specifically, women who scored a 13 or greater on the Framingham tool at baseline (the median) were at increased risk of stroke death, compared with women who scored lower.

A meeting attendee asked at what age raloxifene should no longer be considered, offering 70 years as an example. “Even at age 70, she might not have enough risk factors,” Dr. Cox said. “But it depends on the risk factor.” For example, he said, “atrial fibrillation would put her over the 13 score.”

Dr. Cox and his colleagues also retrospectively assessed application of the Framingham risk score to 7,705 postmenopausal women with osteoporosis randomized to raloxifene or placebo in the Multiple Outcomes of Raloxifene Evaluation (MORE) study (JAMA 2002;287:847–57). In this study, raloxifene was not associated with an increase in rate of all strokes (hazard ratio, 0.71) or stroke death (HR, 0.57). He noted that the MORE population was generally lower risk than women in RUTH, with baseline Framingham scores of less than 13 in 80%.

Raloxifene's effect on the stroke death rate might differ by baseline stroke risk, Dr. Cox said. Assessing cumulative stroke risk may help target raloxifene treatment to postmenopausal women with the most favorable risk-benefit profile.

Next Article: