GENDER-BASED MORTALITY FOLLOW-UP FROM THE PROGRAM ON THE SURGICAL CONTROL OF THE HYPERLIPIDEMIAS (POSCH) AND METAANALYSIS OF LIPID INTERVENTION TRIALS - WOMEN IN POSCH AND OTHER LIPID TRIALS

Objective The authors assessed the clinical results of lipid-lowering therapy in women. Summary Background Data The Program on the Surgical Control oi the Hyperlipidemias (POSCH) has demonstrated that effective lowering of total cholesterol and low-density lipoprotein cholesterol in a postmyocardial infarction population significantly reduces ather osclerotic coronary heart disease (ACHD) mortality, ACHD mortality com bined with a new confirmed nonfatal myocardial infarction, and the num ber of coronary artery bypass grafting and angioplasty procedures perf ormed. Methods A review and meta-analysis were performed of the seven primary or secondary lipid/atherosclerosis intervention triais-includi ng POSCH-published in the English-language literature that included wo men and published results in women separate from the results in men or in the entire trial population. The main outcome measure analyzed was overall mortality. Results The Scottish Physicians Clofibrate Study, the Newcastle upon Tyne Clofibrate Study, and the Pravastatin Limitati on of Atherosclerosis in the Coronary Arteries (PLAC I) Trial may have demonstrated a possible benefit in ACHD prognosis from effective lipi d intervention in women. The other four available trials did not. The Minnesota Coronary Survey reported a 15.6% increase in overall mortali ty rate and a 30.6% increase in a combined cardiovascular endpoint rat e in the lipid-intervention group. The Upjohn Colestipol Study demonst rated statistically significant reductions in overall and ACHD mortali ty in the men, but not in the women. The Scandinavian Simvastatin Surv ival Study (4S) showed highly statistically significant reductions in overall and ACHD mortality in the total trial population, but not in t he 827 women in this study. For the 78 women in POSCH, there was no ev idence of clinical benefit in the lipid-intervention group. Subjecting these seven studies to meta-analysis for overall mortality, a statist ically significant reduction in this clinical endpoint was observed in the men, but not in the women. Assuming independent binomial distribu tions, the probability of obtaining statistical significance, with a t wo-sided alpha of 0.05, in a study of 7066 women (the combined number of women in the seven trials) would be >0.90 against an alternative of the magnitude observed in men. Conclusions These observations have at least two possible interpretations: either the mechanism of coronary obstruction is different in men than in women, or the mortality rate i n the women in the reviewed studies is too low for a statistically sig nificant result. Nonetheless, the available clinical trial data fail t o demonstrate any overall mortality or other convincing clinical benef its from effective lipid intervention in women.