Heart Protection Study: Cholesterol Lowering with Simvastatin
Abstract & Commentary
The hypothesis for this large and important study is that patients with increased vascular risk would benefit from LDL-cholesterol (LDL-C) lowering with a statin, irrespective of baseline LDL-C levels. An important conceptual viewpoint of the investigators is that LDL-C alone may not be an accurate guide in all individuals with a risk of cardiovascular disease, and that benefits of LDL-C lowering would occur in individuals at particularly high risk, such as those with noncoronary vascular disease, the elderly, or those with diabetes. Heart Protection Study sought to assess the long-term effects of cholesterol lowering on vascular and nonvascular mortality and major morbidity in a wide range of patients. This study was performed by the Oxford Clinical Trial Group in the United Kingdom and enrolled 20,536 individuals (age, 40-80) from 69 centers. Randomization occurred between 1994 and 1997; follow-up visits took place until study closeout in late 2001. Physicians were encouraged to use nonstudy statin when appropriate. Thus, there was a substantial "drop-in" use of statins in the placebo group end of the 5-year study duration. The investigators also asked whether an antioxidant cocktail would reduce vascular risk. The study was of 2 × 2 factorial design, which allowed for a separate assessment of statins and antioxidants as well as the combination. Simvastatin 40 mg per day and a combonation of vitamin E, vitamin C, and beta-carotene or matching placebos were provided. The primary end point was total mortality and coronary heart disease (CHD) death. Secondary end points included noncoronary death (eg, cancer); major coronary events, (nonfatal MI or CHD death); and major vascular events, including revascularization procedures and ischemic strokes. Potentially eligible individuals underwent an open label run-in with a fixed dose of 40 mg of simvastatin to assess "LDL lowering responsiveness of each individual." Multiple laboratory tests were carried out during this period. Compliant subjects who did not have a serious event during the run-in period were enrolled. The study cohort consisted of 15,454 men and 5082 women, of whom 5058 were at least 70 years of age at study entry. Also, 65% had documented CHD whereas 35% had no history of CHD. Of the latter 7150 subjects, 1820 had cerebral vascular disease, 2700 had peripheral vascular disease, and 4000 had diabetes. Some subjects had more than one of these conditions, as was true for the CHD cohort. Baseline lipids were as follows: total cholesterol 232 mg/dL, LDL-C 135 mg/dL, HDL-C 40 mg/dL, and triglycerides 80 mg/dL. Follow-up averaged 5 years. Medication compliance was documented with an analysis of individuals who went off study medication and those who went on a nonstudy statin. Thus, by the end of the trial, 82% of the simvastatin (S) cohort were taking their tablets or a nonstudy statin, with average statin use in this group of 85%. Placebo patients (P), had a significant drop-in statin use, such that 32% were taking nonstudy statins by the end of the fifth year.
Results:
All-cause of mortality was reduced in S vs. P by 17%, P < 0.0003, most of which was driven by vascular deaths. Coronary death rate was reduced by 18% P < 0.0001, and other vascular deaths were reduced by 16%, P = 0.07. There was no difference between P and S for nonvascular deaths. Initial nonfatal MI was reduced by 38%, 3.5% S vs. 5.6% P, (P £ 0.0001), and there was a 27% reduction in nonfatal MI or CAD deaths, 8.7% vs. 11.8%, P £ 0.0001. Hospital admissions for unstable angina were reduced, 8.6% vs. 10%; P = 0.0003. Ischemic stroke was decreased by 25%. Revascularization procedures were decreased by 24%, 9.1% vs. 11.7%; P < 0.0001. All major vascular events were 24% less in the S patients, 19.8% S vs. 25.2% P, P = 0.0001. Comparable reductions in risk were found in individuals with no history of CAD, as well as those with cerebral or peripheral vascular disease. Pretreatment cholesterol or triglyceride levels were not closely related to the degree of risk reduction. Many patients had relatively low LDL-C at baseline, and these achieved comparable risk reduction. A total of 6800 subjects had a baseline LDL < 116 mg/dL, and 3400 with a baseline LDL below 100 mg/dL. Event rates in these cohorts were 17.6% S vs. 22.2% P, and 16.4% S and 21.0% P, P < 0.0001 and P = 0.0006, respectively. Similar results were seen in those individuals with baseline total cholesterol level < 193 mg/dL. Of note, prerandomization LDL-C responsiveness to statin did not influence clinical event rate reduction. Overall proportionate reduction in event rates was comparable in all groups and was approximately 25%, irrespective of age and sex. Indeed, in the 1263 individuals older than age 75 at entry the event rate and protection with S was substantial, 23.1% S vs. 32.3% P, P = 0.0002. There was no interaction with other drug therapies.
There was no difference in cancer incidence, liver, or muscle enzyme elevations, or muscle symptoms, between P and S. Elevated CK was rare. There was a slight S excess of overt myopathy and rhabdomyolysis. There were no differences in neuropsychiatric function or cognitive disorders; respiratory disease; or fractures, including those particularly related to osteoporosis. The study concludes that HPS indicates that lowering of LDL-C "produces a substantial reduction on the incidence of major vascular events among a much wider range of high-risk individuals than had previously been shown." These included diabetics and those with cerebral or peripheral vascular disease. They emphasize that the HPS stroke data are the most robust in the literature because of the large numbers in the trial. In addition to decreases in ischemic stroke, there were also less transient ischemic attacks and need for carotid surgery.
They suggest that current guidelines may result in undertreatment of high-risk patients who have a baseline LDL-C close to published target goals, such as 100 mg/dL in the NCEP ATP III Guidelines. The HPS data broadly expand the indications for statins to individuals who are not only at high risk by Framingham Score but who may be at risk for coronary or cerebral vascular events related to individual risk profiles. The proportionate reduction in risk at all levels of baseline LDL-C also potentially extends the use of statins. They emphasize the importance of compliance, and believe that their study would have been even more positive without one sixth of the S cohort going off statin, and another one sixth of P subjects going on a statin, diluting the magnitude of LDL lowering and the clinical benefits. They conclude that statins "will be worthwhile for many types of high risk patients who are not currently being treated, particularly since it has been shown to be so well tolerated and safe." The results of the antioxident component of HPS, not formally published, were completely neutral (MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals. Lancet. 2002;360:7-22.)
Comment by Jonathan Abrams, MD
The HPS data have been "on the street" for many months and numerous overviews have discussed the results. The Lancet report does not substantially add to the information available on the website and other sources, but does allow a more careful and detailed assessment of the data. To be sure, it is most impressive. In essence, HPS suggests that individuals with vascular risk, even if they do not formally meet the NCEP ATP 2% per year projected event level, should be considered for a statin. Perhaps the two most striking findings in HPS are: 1) the reduction in clinical events in the individuals with a normal or low LDL-C at baseline, as well as the comparable event reduction seen in diabetics with no overt clinical vascular disease at entry. The debate regarding how low is enough has been fueled by the CARE investigators, who have promulgated the concept that individuals with a baseline LDL-C of 125 mg/dL or lower do not benefit from a statin, and that lowering the LDL level beyond 125 mg/dL is not necessary. HPS essentially obliterates that argument, which has never made sense to many experts in the field. The concept that a diabetic should be treated with a statin, irrespective of lipid levels and the presence of overt current vascular disease, is not new but has not been formally tested by any other trial. The results are positive, and they robustly support the view that adult onset diabetes is a target for statin therapy. Certainly, for the higher risk diabetics, ie, those with hypertension, smokers, elevated creatinine, obesity, or significant dyslipidemia, even without elevated LDL, a statin would appear to be important. One of the strengths of this trial is the large numbers of individuals in every category. It is no particular surprise that coronary events including coronary death and revascularizations are reduced by LDL-C lowering. Several other statin trials have also shown a reduction of stroke. HPS broadens and strengthens this database. They make too much of a point of the drop-in and dropout rates. Most RCT experience a loss of individuals on active therapy over the course of the trial, and it is not unusual for off-study medications to be provided, particularly as new data accumulate during the course of the trial. This is not a problem unique to HPS. These researchers are to be congratulated for their boldness in designing this study well before the Scandinavian 4S trial was reported, at a time when we did not fully understand or know the risks and benefits of the statins. The simplicity of providing a statin to a huge high-risk cohort is appealing. One of the most exciting aspects of HPS is the complete concordance of event reductions across so many strata of vascular risk, age, etc, confirming that the results are real. The NCP-ATP III guidelines may need to be modified in the near future. We await further reports on HPS from the Oxford group, who have already contributed so much to the treatment of coronary artery disease.
Dr. Abrams is Professor of Medicine, Division of Cardiology, University of New Mexico, Albuquerque.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.