Author + information
- ↵⁎Reprint requests and correspondence:
Dr. Daniel Steinberg, School of Medicine, Basic Science Building, Room 1080, University of California-San Diego, 9500 Gilman Drive, La Jolla, California 92093-0682
- coronary heart disease
- randomized clinical trials
- risk factors
- treatment guidelines
From time to time we need to remind ourselves that atherosclerosis begins in childhood as fatty streaks. These lesions are of course benign in that they are asymptomatic, do not obstruct blood flow, and do not predispose to thrombosis. However, they are actually anything but benign. They are the precursors of the advanced lesions that ultimately—decades later—will precipitate coronary thrombosis and myocardial infarction (1,2). Fatty streak lesions and even some fibrous plaques are already well established in young adulthood. Their anatomic locations in the arterial tree are very much the same as those of the later lesions (1,3). Moreover, the risk factors that correlate with the extent of such early lesions are the same risk factors that correlate with myocardial infarction later in life (2,4,5). In other words, the disease does not somehow morph into another form as we get older; the disease progresses, and the lesions get larger and become life-threatening. Based on these insights into the natural history of atherogenesis, pathologists began urging many years ago that preventive measures should be instituted earlier in life (6). However, the drugs available at the time were less than ideal, and there was no way to directly assess the value of early intervention and make a meaningful risk/benefit assessment, so the issue was moot. Today, thanks to advances in our understanding of the mechanisms regulating blood cholesterol levels and the genes involved in that regulation, it has become clear that earlier intervention could improve in a major way the impact of lowering blood cholesterol levels on coronary heart disease (CHD) risk.
The newer genetic evidence
The work by Cohen et al. (7) on the PCSK9 gene represented a breakthrough in this area. The PCSK9 gene plays a key role in the regulation of the low-density lipoprotein (LDL) receptor, suppressing the expression of the receptor and thus increasing LDL levels. Mutations causing overexpression of PCSK9 can increase LDL to levels as high as those found in familial hypercholesterolemia (8). Conversely, nonsense mutations in the PCSK9 gene allow increased expression of the receptor, thus lowering plasma LDL levels. What Cohen et al. discovered is that individuals with a particular nonsense mutation in PCSK9 had an LDL level 28% lower than that in the rest of the population under study. The astonishing finding was that the CHD risk in these individuals was reduced by >80%! By comparison, the same 28% decrease in LDL in the statin trials has only reduced CHD risk by 25% to 35%. Cohen et al. proposed that the much greater effect was due to the fact that LDL levels in people with PCSK9 loss of function mutations were lower right from birth, not just for the 5 or 6 years of a statin trial, a trial that usually started in middle age. Their findings have been amply confirmed (9,10).
In this issue of the Journal, Ference et al. (11) have provided a meta-analysis of published data that points up the magnitude of protection potentially offered by a lifetime lowering of LDL cholesterol (LDL-C). The authors conclude that lifetime exposure to an LDL level lowered by ∼40 mg/dl because of mutations in 1 or more of these alleles could reduce CHD risk by almost 55%! Ference et al. (11) used Mendelian randomization to examine the impact of gene polymorphisms affecting LDL levels and CHD risk. They included PCSK9 and data on 5 additional genes related to cholesterol or lipoprotein metabolism. The data on PCSK9 are the most striking, but the data on the other alleles studied, although less impressive, are consonant. If there are no hidden problems with the methodology; these data also suggest that the magnitude of the benefit conferred by the mutations studied relates directly to the extent that they lower plasma LDL no matter what the mechanism by which they do so. These alleles all have functions relating to lipid and lipoprotein metabolism, although the precise mechanisms leading to the low LDL are not known in all cases. At any rate, the mutations studied are unlikely to be influencing atherogenesis in ways unrelated to lipoprotein metabolism. Thus, lowering LDL earlier in life, using diet and/or drug approaches, could prevent not just 30% of events, as in the statin trials, but possibly more like 60%.
A growing body of literature has provided the rationale for earlier intervention (12–16), not only with respect to hypercholesterolemia but also with respect to other risk factors such as hypertension, obesity, and diabetes. Here we confine our discussion to hypercholesterolemia. There is good reason to predict a very favorable risk/benefit ratio for early intervention, and there is a growing consensus that treatment—whether dietary or pharmacological—should start much sooner than is the practice today (17–19). There are understandable concerns about treating younger people and treating them for a lifetime. However, the bulk of evidence suggests that the benefits will easily outweigh the risks (12,20–22). The meta-analysis presented by Ference et al. (11) adds strength to the case. Still, there is understandable hesitation to proceed without explicit randomized, controlled trial (RCT) evidence.
What are the options available?
First, we could, as proposed by Domanski et al. (15), undertake a 10-year RCT in a younger age group (30 to 50 years). To give a clear-cut result, a sizable investment of time and money would be required. At least 15 years would be needed overall for organization, implementation, and interpretation of the results. Because absolute risk at baseline would be low, risk reduction might turn out to be limited to those at highest risk at baseline, making generalization problematic. Nonetheless, a trial of this type, even if inconclusive with respect to major endpoints, could yield valuable information. For example, it could document the safety and tolerability of statins with long-term use. It could demonstrate the willingness of younger, asymptomatic individuals to adhere to statin therapy for many years. A positive result would, of course, represent a milestone, and a major revision of guidelines would follow. However, a negative result could have a chilling effect on further attempts to document the value of early intervention.
Alternatively, an even more expensive trial design but one more likely to yield a conclusive result could be considered. Study subjects would be younger (e.g., 30 to 35 years), have no existing CHD, have an LDL level <130 mg/dl, and no more than 2 additional risk factors. Detailed entrance criteria would have to be carefully chosen in relation to the study power wanted and the number of subjects that would yield enough hard endpoints in 20 to 30 years. The aim would be to select a study population typical for the age group but not at very high risk. Undoubtedly, this would cost more and last longer, but in view of the high stakes involved, it might be the wisest way to go.
A second option would be to take the position that the evidence already in hand is strong enough to justify changing the guidelines without waiting for an RCT (16). Heretical as that may sound, it is by no means unprecedented. For example, we already treat children with familial hypercholesterolemia with drugs as early as age 8 even though there are no explicit RCT data to justify that intervention (13). Also, we do not hesitate to recommend against smoking even though there are no RCT data to back up that recommendation and there never will be. The acceptability of this option depends on the strength of our conviction that the totality of evidence, discussed here and in other recent papers (23,24), justifies acting without waiting for the results of a RCT.
Even if one or another type of controlled intervention trial is undertaken, it will be many years before the results can become available. In the interim, there are other options to consider.
The first would be to replace the current use of the 10-year Framingham risk score with a lifetime risk score, as proposed by Lloyd-Jones et al. (13,14). The 10-year Framingham risk score heavily weights age as a factor. Intervention is not recommended in subjects in their 30s or 40s, even though their lifetime risk may be very high. In other words, they may get through the next 10 years without an event, but their risk factor profile predicts that the chance that they will eventually, perhaps in their 50s or 60s, have an event is nevertheless high. A shift to the use of lifetime risk would result in significantly earlier intervention.
A second approach would be to rethink our criteria for what represents an acceptable LDL level and set goals that should be reached at an early age. The 50th percentile for LDL-C in American adults older than 35 years of age is ∼130 mg/dl, and the 75th percentile is ∼160 mg/dl for men and somewhat lower in women. If the goal for LDL-C for all adults, not just those with clinical atherosclerotic disease or diabetes, were set at <100 mg/dl, the majority of people could achieve that goal by lifestyle alone. If not, they could opt to take a low dose of a cholesterol-lowering drug. For example, 10 mg of simvastatin will lower LDL-C levels on average by 30%, and if combined with lifestyle change, by >40%. This strategy would get at least three fourths of the population to an LDL-C level of <100 mg/dl. Those with higher baseline LDL-C might require a higher dose of statin, but drug treatment for anyone with a baseline LDL-C of >160 mg/dl is not difficult to justify.
A final option is to be conservative and stay with guidelines similar to those currently in place. National guidelines invariably tend to gravitate toward the conservative position, which for hypercholesterolemia is to intervene later in life. We reject this as an option. The evidence that earlier intervention will significantly reduce the toll of CHD is too strong to be ignored. The public and the profession need to become aware of the potential benefit of keeping cholesterol low, not just after age 50 or 60 but, even in those at intermediate risk, beginning at age 30 or 35. To the widely accepted dictum “the lower the better,” we should probably add “the earlier the better.”
Both authors have reported that they have no relationships relevant to the contents of this paper to disclose.
↵⁎ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
- American College of Cardiology Foundation
- Stary H.C.,
- Chandler A.B.,
- Glagov S.,
- et al.
- Ference B.A.,
- Yoo W.,
- Alesh I.,
- et al.
- Steinberg D.,
- Glass C.K.,
- Witztum J.L.
- Lloyd-Jones D.M.,
- Leip E.P.,
- Larson M.G.,
- et al.
- Steinberg D.
- Cholesterol Treatment Trialists' (CTT) Collaborators,
- Mihaylova B.,
- Emberson J.,
- Blackwell L.,
- et al.
- Avis H.J.,
- Hutten B.A.,
- Gagne C.,
- et al.
- Law M.R.,
- Wald N.J.,
- Thompson S.G.