The emerging role of statins in the prevention of coronary heart disease

Statins are effective but we need better ways of assessing risk

The Standing Medical Advisory Committee guidelines for the use of statins have ignited considerable debate in Britain, and similar discussions about the use of statins are, or soon will be, occurring in other countries. The crux of the controversy lies, on the one hand, in earnest efforts to prevent coronary heart disease by any effective means available and, on the other, in fiscal realism. Such controversy has been foreseeable since publication of three clinical trials showing that statins are highly effective in preventing heart disease. The British guidelines conservatively recommend statin treatment only for individuals with at least a 3% annual risk of coronary heart disease events—a threshold selected to minimise costs and focus on patients at highest risk. Ironically, the subsequent BMJ editorial criticised the guidelines for being fiscally irresponsible,1 whereas other issued guidelines and much of the journal correspondence2 call for the treatment of more, not fewer, individuals.3

Three interrelated challenges confront efforts to determine the optimal use of statins: (a) sufficient definition of the benefits, costs, and risks of treatment based on data from randomised clinical trials; (b) innovative strategies to minimise treatment costs; and (c) improved methods for estimating individual patients' risk. It is now clear that treatment of hypercholesterolaemic patients with statins reduces the incidence of fatal and non-fatal myocardial infarctions by 30-35%.4 These benefits are accompanied by fewer coronary revascularisation procedures and fewer strokes.5 Additional potential benefits that have not yet been clearly shown include reductions in the incidence of anginal symptoms, congestive heart failure, disability, and unemployment and improved quality of life. Moreover, the cost effectiveness of statins in secondary prevention is comparable to, or better than, other standard preventive interventions.6

Nevertheless, several important gaps in our knowledge remain. The quoted reduction in coronary heart disease produced by statins in patients with heart disease was achieved in studies in which many of the participants were smokers and were not receiving aspirin or ß blockers. Most of the benefit may have occurred in smokers receiving lipid therapy alone; conversely, statins may have produced additive or even synergistic effects in combination with non-lipid treatments like smoking cessation, aspirin, or ß blockers. The incremental benefit from statin therapy in patients already receiving other standard interventions needs to be quantified.

Importantly, the cost effectiveness of statins in the treatment of individuals without clinical signs of cardiovascular disease has not been defined. Primary prevention is clearly less cost effective than secondary prevention because of the lower absolute risk of coronary heart disease, especially in the short term, in overtly healthy individuals. The paper by Caro et al in this issue (p 1577) is the third published analysis on the cost effectiveness of primary prevention with statins, and each comes to different conclusions.7 8 9 Such analyses are necessarily complex and employ various models and assumptions at the discretion of the authors.10 Also, some cost effectiveness research is supported directly by pharmaceutical companies, which have much to gain from increasing statin markets. 6 7 Therefore consensus on this issue will be difficult.

Cholesterol lowering treatment in the elderly, for whom little clinical trial data exist, presents additional questions. Older people are at high risk of coronary heart disease because of age, yet elderly people with low cholesterol values do not necessarily have less heart disease or survive longer than their hypercholesterolaemic counterparts.11 12 Finally, with respect to safety, deaths from certain non-coronary causes, including cancer, accidents, and suicides, have been increased in some trials, and potential mechanisms are still being investigated.13 14 15 Also, non-coronary morbidity outcomes have not been fully reported, and long term follow up has not yet accrued.

The purchase price of statins is the single largest determinant of the cost of lipid lowering therapy. Therefore, when healthcare resources are fixed, drug price very directly affects our ability to prevent coronary heart disease. If purchasing leverage is used to drive competitive bidding for formulary contracts, considerable costs savings may be achieved (p 1616).2 For example, the Veterans Health Administration has obtained a price reduction of nearly half 16 17 that of the typical purchase price of simvastatin in America (which is similar to that in Britain).

Clinical decisions about which patients to treat essentially entail estimating the risk of coronary heart disease. That risk is quite high in patients with clinically apparent coronary artery obstruction, and three quarters of such individuals will die from ischaemic heart disease. We agree with the guidelines regarding hypercholesterolaemic patients with atherosclerotic disease outside the coronary arteries (aortic, peripheral, or carotid). Their risk of coronary heart disease is also high and proportional to the severity of their atherosclerosis.18 19 20 21

Difficulties and differences of opinion arise over the treatment of apparently healthy people, and for them we can and must improve ways of assessing risk. The more accurately coronary heart disease is predicted, the more efficiently our treatment can be targeted, and the more coronary events we will prevent at the least cost and risk. This greater precision can be achieved by systematic and quantitative consideration of multiple risk factors. Total cholesterol is not the best lipid parameter—we recommend the ratio of total cholesterol to high density lipoprotein (HDL) because this single number captures most of the value of the full lipid profile in most patients.22 Moreover, change in the total cholesterol:HDL ratio with treatment correlates with benefit in coronary heart disease better than other lipid measures.

Commendably, the Sheffield tables do quantify risk and, as well as cholesterol values, take into account age, gender, hypertension, current smoking, diabetes, and left ventricular hypertrophy on electrocardiography. However, we also know that early coronary heart disease in a first degree relative (before age 50 in men or 60 in women) increases an individual's risk and also helps to identify those with familial hyperlipidaemias. Other blood borne factors and genetic markers may soon be added to standard risk assessments.22 23 Ultimately, the disease we are trying to predict causes morphological changes which may be imaged or otherwise measured, and this should lead us from indirect risk assessment to non-invasive measurement of the disease itself, atherosclerosis. Here, accumulating evidence is establishing the utility of detecting subclinical atherosclerosis by the ankle-brachial blood pressure index, carotid artery sonography, and detection of coronary calcifications.24 25 26 27 The challenge is to offer doctors improved methods of risk stratification to use in their daily practice.

Irrespective of these deliberations, prescriptions for statins have jumped in Britain, and during 1997 in America annual sales of cholesterol lowering drugs increased by a remarkable 29% to $3.7bn (£2.3bn) (International Marketing Services). The great popularity of statins is largely a result of their efficacy and tolerability. However, we should also be cautious with these powerful drugs, for the reasons discussed above. Moreover, the comparable ease of reducing cholesterol concentrations by pharmacological means should not substitute for multiple risk factor interventions, including non-pharmacological approaches to preventing coronary heart disease.