ABSTRACT
BACKGROUND: Physicians seek new therapies to reduce the risk of fractures in osteoporotic post-menopausal women without increasing the risk of cardiovascular events or cancers. Raloxifene (Evista), a selective estrogen receptor modulator, may be a new choice for therapy. Using original data from the Multiple Outcomes of Raloxifene Evaluation, the authors present a secondary analysis of this randomized trial to evaluate its effect on cardiovascular events.
POPULATION STUDIED: The researchers enrolled 7705 women who were at least 2 years post-menopausal, from outpatient and community settings at 180 sites in 25 countries. Average age was similar by treatment group (overall mean = 67 years). All patients had osteoporosis documented by either prior vertebral fracture or a bone mineral density T score of less than -2.5. Study women were predominantly white (95%). Baseline characteristics of women were similar for most cardiovascular risk factors and concomitant cardiovascular medications, although women receiving raloxifene were significantly more likely to have diabetes.
STUDY DESIGN AND VALIDITY: The study was a double-blind randomized, controlled trial with concealed allocation assignment. It originally was designed to determine the effect of raloxifene on bone mineral density and vertebral fractures. Women were randomized to receive placebo, or 60 mg or 120 mg of raloxifene per day. This study was a secondary analysis of the data for cardiovascular outcomes. Risk scores, based on evidence of established coronary heart disease or the presence of cardiovascular risk factors, were assigned to a subgroup of women with increased cardiovascular risk.
OUTCOMES MEASURED: The authors collected cardiovascular event outcome data by asking women at each visit whether they had experienced a myocardial infarction, coronary artery bypass graft, percutaneous coronary intervention, or stroke since the previous visit. Unsolicited reports of cardiovascular events were also recorded.
RESULTS: The follow-up data were 90.8% complete for women who took placebo and 89.6% for those who took raloxifene at 1 year. There was no significant difference between combined treatment and placebo groups in the number of women with cardiovascular events during the first year of the trial. Nor was there a difference in the high-risk subset. The serious loss to follow-up for the 4 years of the study (25% of placebo and 22% of raloxifene women) makes the analysis unreliable for longer than the first year of study. We can use intention-to-treat analysis to assess the potential effect of missing cardiovascular events in those lost to follow-up.1 The resulting relative risk ranges from 0.11, for the extreme assumption that all missing women taking placebo suffered a cardiovascular event while those on raloxifene did not, to 6.89, for the opposite extreme that all missing women taking raloxifene suffered a cardiovascular event while those on placebo did not. The true relative risk lies somewhere between these boundaries. With so much data missing, we are unable to assess raloxifene’s effect on cardiovascular events in postmenopausal osteoporotic women in the longer term.
After 1 year of therapy, raloxifene did not increase the risk of cardiovascular events in older postmenopausal women with osteoporosis. Its effect on cardiovascular risk has not been assessed in women taking it for more than 1 year. Absence of a detrimental cardiovascular effect is a benefit, compared with estrogen replacement therapy, although both approaches prevent osteoporotic fractures. However, both of these hormonal approaches carry the same risk for thromboembolism. Raloxifene may cause or worsen hot flushes, whereas estrogen prevents them. Long-term compliance with either therapy is not good. Given the cost and risks of the biphosphonates, the optimal approach to osteoporosis prevention and treatment is a difficult clinical decision.