Abstract & Commentary
Synopsis: HPS provides definitive evidence that . . . statin therapy can produce substantial reductions in the risk of heart attacks, strokes, and revascularization in diabetics, even if there is diagnosed coronary or other vascular disease.
Source: Collins R, et al. For the Heart Protection Study Collaborative Group. Lancet. 2003;361: 2005-2016.
Over a year ago the Heart Protection Study (HPS) investigating the efficacy of simvastatin (simva) and antioxidant vitamins was published.1 An across-the-board, robust reduction in vascular events was seen in the simva cohort in this huge trial of 21,000 individuals who had vascular disease or were at high risk by nature of having diabetes or treated hypertension. The HPS results were particularly tantalizing in that they demonstrated comparable relative risk reduction in individuals who had baseline normal and below normal total and LDL cholesterol values, as well as in diabetics of whom 50% had no clinical vascular disease. The present report is a detailed analysis of the approximate 6000 diabetics in HPS; of this cohort, 33% had a history of coronary artery disease (CAD), 18% had "other occlusive arterial disease," and 49% had no history of arterial disease. Ninty percent were type II diabetics, and 10% were type I. Compared to nondiabetics, the diabetic cohort was slightly younger and had somewhat less men (who represented over 70% of the group); there were no significant differences in lipid levels between groups. Subjects from general practice clinics who met entry criteria were randomized to simva 40 mg daily or placebo. Mean follow-up was 4.8 years for diabetics and 5.0 years for all remaining subjects. Diabetics represented approximately 29% of the entire HPS cohort.
The results in the diabetic cohort were essentially identical to the larger group. There was an approximate 27% reduction in the primary end point of coronary death or nonfatal myocardial infarction (P < .0001), as well as a comparable reduction in all major coronary events. Diabetic CAD mortality was 8.0% placebo vs 6.5% simva (P = .02) and for first nonfatal MI, 5.5% vs 3.5% (P = .0002). There was a 25% reduction in stroke in both the diabetic and nondiabetic cohorts. There was a 17% reduction in coronary revascularization in diabetic subjects compared to 24% in the entire HPS cohort, but this difference was not significant. Diabetics on simva had a slightly lower incidence of peripheral vascular disease complications, and renal function impairment was somewhat slowed by the statin. Of great interest, in the 50% of diabetics who did not have a diagnosis of vascular disease at entry, there was a 33% reduction in first major vascular event (9.3% vs 15.5%; P = .003), with an overall 25% reduction for the entire diabetic cohort, regardless of age or gender.
Individuals with a variety of baseline lipid abnormalities had no difference in the relative risk reduction of 27%, including those whose initial LDL cholesterol was < 116 mg/dL. A total of 1343 diabetics had a pretreatment LDL cholesterol of < 116 mg/dL; these achieved a 30% reduction in first major vascular events, 8% vs 11% (P = .05). The placebo group in the entire HPS cohort had a 5% per year risk of a major vascular event over the 5-year follow-up; simvastatin reduced this risk by approximately 25%. As in the original publication, the investigators argue that the rates of significant drop-in statin use in the placebo group of 17%, as well as a 15% drop-out rate in the simva cohort, suggest that the actual risk reduction with simvastatin vs no lipid therapy would approach one-third rather than one-fourth. There was no difference in the onset of new diabetes, and there were no effects on diabetic control. The authors conclude that "HPS provides definitive evidence that . . . statin therapy can produce substantial reductions in the risk of heart attacks, strokes, and revascularization in diabetics, even if there is diagnosed coronary or other vascular disease."
The reduction of LDL cholesterol with simva was approximately 1.0 mmol/L (39 mg/dL) and was achieved in both diabetics and nondiabetics. In HPS diabetics who began with an LDL cholesterol of < 116 mg/dL, the LDL levels were lowered to < 77 mg/dL; this was accompanied by a 25% reduction of macrovascular disease, comparable to the relative risk reduction in those individuals with higher baseline cholesterol. HPS investigators stress that vascular risk should drive statin therapy as determined by a diagnosis of occlusive arterial disease or diabetes, rather than baseline lipid levels. In fact, LDL cholesterol in HPS was relatively unremarkable at baseline. Diabetics who had vascular disease at entry had a 3-fold greater likelihood of having a major vascular event, although the relative risk reduction was similar. Of interest, not only was the first event reduced by statin therapy, but subsequent events during the 5-year trial also were decreased. They concluded, "HPS has shown that the benefits of cholesterol-lowering statin therapy are additional to those of other cardioprotective treatments. In particular, statin therapy should now be considered routinely for all diabetic patients at sufficiently high risk of such major vascular events, irrespective of their initial cholesterol concentration."
Comment by Jonathan Abrams, MD
The results in the diabetic cohort in HPS speaks for itself. This is an extremely large cohort, far larger than all of the diabetic patients combined in prior statin trials. While the NCEP-ATPIII and American Diabetes Association have already established diabetes as a coronary risk equivalent, with a target LDL cholesterol of 100 mg/dL, it is noteworthy that many to most type II diabetics do not have a significant LDL abnormality and actually have comparable LDL levels to the general population. Statins are less effective for the high-triglyceride/low-HDL abnormality of diabetics.
Of interest, another study of statin therapy in diabetics from the United Kingdom and Ireland (CARDS) has just been prematurely terminated because of a highly significant reduction in the primary end point of multiple major vascular events (P = .005). In CARDS, 2800 patients were randomized to placebo or atorvastatin 10 mg. These were type II diabetics with at least one major risk factor for CAD, but who had no overt vascular disease at entry and would not otherwise qualify for lipid-lowering therapy according to UK guidelines. Specific data is not yet available. On the other hand, recent satin results from ALLHAT and ASCOT were somewhat disappointing in the diabetic cohorts. A narrow LDL differential between active statin therapy and placebo during these trials rendered the results less conclusive than HPS, particularly in diabetics.
In conclusion, it seems incontrovertible that a statin should be prescribed for diabetics who are at high risk. These include those with one or more major CAD risk factors, older individuals, and all diabetics with overt vascular disease, be it peripheral, cerebrovascular, or coronary. Half of the diabetics in HPS had no vascular disease at entry and an unknown frequency of the major risk factors; the mean age in HPS in the diabetics was 62. These data suggest that any middle-aged type II diabetic should receive a statin, even if the LDL cholesterol is close to 100 mg/dL or lower. The premature cessation of CARDS and the HPS data suggest that it does not matter which statin is used, so long as a clinically effective dose is chosen. Diabetics without overt vascular disease have a roughly comparable risk to nondiabetics with documented CAD; one should be able to extrapolate HPS and CARDS data to any adult type II diabetic. HOPE supports the use of an ACE inhibitor in diabetics older than 55 with an additional risk factor, and other data, although less conclusive, indicates that aspirin is protective in these patients. Current guidelines support aggressive blood pressure control in diabetics, with a target of < 120/80 mm Hg. Thus, a preventive cocktail for the type II diabetic should include aspirin, an ACE inhibitor, a statin, and usually 2 or more drugs for hypertensive therapy in those with elevated blood pressure. Glycemic control is deemed to be most important, but data are lacking supporting a significant macrovascular event differential between low vs high HgbAIC. It may well be that greater vascular protection is provided by a generic prevention cocktail than optimal fine-tuning of blood sugar in type II subjects.
The HPS had a 4 × 4 factorial design and also examined an antioxidant combination compared to placebo. That part of the study was negative. In the same issue of Lancet2 as the diabetic HPS report, a meta-analysis of antioxidant vitamins for the prevention of cardiac disease concluded that there is no data supporting use of these agents.
Dr. Abrams, Professor of Medicine, Division of Cardiology, University of New Mexico, Albuquerque, is on the Editorial Board of Clinical Cardiology Alert.
1. Lancet. 2002;360:7-22.
2. Lancet. 2003;361:2017-2023.