Three-year follow-up of the Oxford Cholesterol Study: assessment of the efficacy and safety of simvastatin in preparation for a large mortality study

We report the results of a randomized single-centre study designedto assess the effects of simvastatin on blood lipids, bloodbiochemistry, haematology and other measures of safety and tolerabilityin preparation for a large-scale multicentre mortality study.Six hundred and twenty-one individuals considered to be at increasedrisk of coronary heart disease were randomized, following a2-month placebo ‘run-in’ period, to receive 40 mgdaily simvastatin, 20 mg daily simvastatin or matching placebo.Their mean age was 63 years, 85% were male, 62% had a historyof prior myocardial infarction (MI), and the mean baseline totalcholesterol was 7·0 mmol. 1^–1. Median follow-upin the present report is 3·4 years. Eight weeks after randomization, 40 mg daily simvastatin hadreduced non-fasting total cholesterol by 29·2% ±1·1 (2·03 ± 0·08 mmol. 1^–1)and20 mg daily simvastatin had reduced it by 26·8% ±1·0 (1·87 ± 0·07 mmol. 1^–1).Almost all of tile difference in total cholesterol at 8 weekswas due to the reduction in LDL cholesterol (40·8±1·6and 38·2%±1·4 among patients allocated40mg and 20mg of simvastatin daily respectively), but simvastatinalso reduced triglycerides substantially (19·0% and 17·3%)and produced a small increase in HDL cholesterol (6·4%and 4·8%). These effects were largely sustained overthe next 3 years, with 40 mg daily simvastatin producing a slightlygreater reduction in total cholesterol at 3 years (25·7%±1·9reduction) than did 20 mg daily simvastatin (22·2%±1·8). There were no differences between the treatment groups in thenumbers of reports of ‘possible adverse effects’of treatment or of a range of different symptoms or conditions(including those related to sleep or mood) recorded at regularclinic follow-up. Mean levels of alanine aminotransferase, aspartateaminotransferase and creatine kinase were slightly increasedby treatment, but there were no significant difference betweenthe treatment groups in the numbers of patients with significantlyelevated levels. A slightly lower platelet count in the simvasatingroup was the only haematological difference from placebo, withno difference in the numbers of patients with low platelet counts. In summary, the simvastatin regimens studied produced largesustained reductions in total cholesterol, LDL cholesterol andtriglyceride and small increases in HDL cholesterol. They werewell tolerated, with no evidence of serious side-effects duringthe first 3 years of this study. Consequently, simvastatin providesan opportunity to conduct large randomized studies that assessthe effects of cholesterol lowering on total mortality and oncause-spec mortality. Even with such an effective cholesterollowering treatment, however such mortality studies will stillneed to be prolonged and to involve some tens of thousands ofpatients at substantial risk of coronary heart disease if clearevidence is to emerge.