Lowering LDL--Not Only How Low, But How Long?

See allHide authors and affiliations

Science  24 Mar 2006:
Vol. 311, Issue 5768, pp. 1721-1723
DOI: 10.1126/science.1125884

The causal relation between plasma low-density lipoprotein (LDL) cholesterol (LDL-C) levels and coronary heart disease is well established. Compelling evidence from between-country comparisons shows that large and lifelong diet-related differences in LDL-C levels are associated with 10-fold differences in coronary mortality (1) (see the figure). Strong support comes from observations on genetic diseases such as heterozygous familial hypercholesterolemia, in which mutations in the LDL receptor gene double LDL-C levels throughout life and increase the risk of early heart attack by more than 10-fold (2). So, it has been somewhat disappointing that treatment with cholesterol-lowering statins for 5 years reduces the incidence of heart attacks by only 40%, even when LDL-C concentration is reduced by 80 mg/dl (3), a reduction that should give much more protection based on the population studies. A likely explanation is provided by Cohen, Hobbs, and their colleagues in this week's issue of the New England Journal of Medicine (4). In lowering LDL levels, the appropriate consideration may be not only how low, but also how long.

Cohen et al. studied middle-aged Americans with lifelong low LDL levels, owing to loss-of-function mutations in the gene encoding PCSK9, a secreted enzyme of the serine protease family. In a small number of subjects with severe nonsense mutations, the concentration of LDLC was reduced by 38 mg/dl, and the prevalence of coronary heart disease declined by a remarkable 88%. In a larger number of subjects with a less severe missense mutation, LDL-C concentration was reduced by only 21 mg/dl, yet coronary heart disease incidence declined by 47%.

What is the function of PCSK9, and how do mutations in the PCSK9 gene lower the concentration of LDL? Experiments in mice showed that overproduction of PCSK9 in liver and cultured hepatocytes severely reduces the number of LDL receptors (5, 6). The simplest hypothesis is that PCSK9 directly catalyzes the breakdown of LDL receptors, but this has not been demonstrated experimentally. Inasmuch as LDL receptors mediate high-efficiency removal of LDL from plasma, a reduction in the number of LDL receptors causes LDL to accumulate. Ablation of the PCSK9 gene in mice through gene-knockout technology increased the number of LDL receptors in liver and enhanced the clearance of LDL from the plasma (7). This striking finding indicates that PCSK9 functions tonically in mice to keep LDL receptor number lower and plasma LDL concentration higher than they would be otherwise.

PCSK9 appears to have the same effect on LDL in humans. A role for PCSK9 was first recognized in families with autosomal dominant hypercholesterolemia, owing to amino acid substitutions in PCSK9 that are postulated to increase its function (810). As expected, affected individuals suffered from premature heart attacks. Cohen et al. have now demonstrated the opposite effect—namely, that loss-of-function mutations in PCSK9 lower LDL levels and reduce the incidence of heart attacks.

In a previous study, Cohen et al. (11) sequenced PCSK9 in a population-based study designed to reflect the ethnic diversity of Dallas, Texas. Among 1802 individuals of African descent, 2% carried one of two loss-of-function nonsense mutations in PCSK9. Individuals harboring either of these mutations had plasma LDL levels averaging 40% lower than those without them. These mutations were rare (0.1%) in Americans of European ancestry (Caucasians) (11).

Cohen et al. (4) now demonstrate the cardio-protective effect of the LDL-lowering mutations in PCSK9. They analyzed data from a prospective study of 15,792 Caucasians and African-Americans from four U.S. communities that was initiated in 1987 (12). These randomly selected individuals averaged 53 years of age at entry, and have been followed for 15 years. Among the 3278 African-American individuals without a PCSK9 mutation, LDL-C levels averaged 138 mg/dl, and 319 of these individuals developed symptomatic coronary heart disease for an incidence of 9.7%. Among the 85 African-American individuals with a PCSK9 nonsense mutation, LDL-C concentration was reduced by 38 mg/dl (to 100 mg/dl). Remarkably, only one of these 85 people (1.2%) developed coronary heart disease, an 88% reduction. In these protected individuals, coronary heart disease was rare despite a high prevalence of hypertension (37%) and diabetes (13%). Although the number of subjects is small, the extremely low incidence of coronary heart disease in African-Americans with PCSK9 nonsense mutations is consistent with other studies (see the figure) that show an extremely low incidence of coronary heart disease in populations with lifelong low cholesterol levels (1, 13). Among Caucasians in the same study, 301 individuals (3.2%) had a missense mutation that lowered LDL-C levels by only 21 mg/dl, yet reduced coronary heart disease incidence by 47%.

Why does lowering of LDL-C concentration by 40 mg/dl by a PCSK9 mutation reduce coronary heart disease incidence by 88%, whereas a 40-mg/dl lowering with a statin reduces coronary heart disease prevalence by only 23% on average (3)? The most likely answer is duration. People with nonsense mutations in PCSK9 likely have maintained relatively low LDL levels throughout their lives. People in statin trials have had their LDL levels lowered for only 5 years. Atherosclerosis is a chronic disease that begins in the teenage years (14). In a statin trial, an individual destined to have a heart attack within the 5-year observation period must have had advanced atherosclerosis when entering the trial. Indeed, the degree of protection in statin trials increases with duration (3, 15).

Where high cholesterol and coronary death meet.

Beginning in 1952, Keys and colleagues measured the levels of total plasma cholesterol in 12,763 men aged 40 to 59 from selected population groups from seven countries, with wide variation in fat intake and plasma cholesterol levels (1). The men were followed for 10 years, and the number of fatal coronary events was measured. Inasmuch as the study did not measure plasma lipoprotein concentrations, we estimated the LDL-C levels based on the data for total cholesterol levels. For each country, we averaged the data from the populations that were studied. Coronary mortality between countries varied by 10-fold.


The lesson of PCSK9 is clear. If we are to attain an 88% reduction in the incidence of coronary heart disease, we must lower LDL levels well before atherosclerosis has become advanced. If we start early enough, it may be sufficient to lower LDL-C concentration only to 100 mg/dl, a goal that should be attainable for most people. These individuals must be prepared for lifetime treatment. Early intervention is designed to prevent a heart attack that might not occur for many years.

The physiological means to lower LDL concentration is through a stringent diet that is low in cholesterol and saturated fat. If this fails, drugs can be used. These include statins, cholesterol-absorption inhibitors, and bile acid-binding resins, all of which function by depleting the liver of cholesterol and increasing the number of hepatic LDL receptors. Statins inhibit 3-hydroxy-3-methylglutaryl coenzyme A reductase, the rate-controlling enzyme in cholesterol synthesis (16). This action depletes liver cholesterol and activates a transcription factor called SREBP (sterol regulatory element-binding protein), which increases the expression of mRNA encoding LDL receptor (17). The increased numbers of LDL receptors produce a selective fall in LDL concentration (18). Statins have been in widespread use for 20 years, and placebo-controlled studies in 90,056 patients have shown a very low incidence of side effects, primarily rare muscle necrosis and occasional increases in circulating liver enzymes (3).

The use of cholesterol-lowering drugs has been restricted to individuals suspected to be at high risk for myocardial infarction. Treatment is usually initiated at ages in which the atherosclerotic process is likely to have already advanced. One objection to earlier use has been cost. This objection may be overcome by the availability of low-cost generic statins. Generic statins are an option for relatively young individuals with LDL-C levels that are “normal” for the U.S. population, but are above the levels that offer protection from heart attacks. Selection of individuals for preventive treatment would improve if we had reliable noninvasive methods to diagnose early atherosclerosis.

Current data justify initiating more aggressive drug therapy at the first sign of hypertension or diabetes, even when blood pressure and glucose levels can be controlled. Current guidelines of the U.S. National Institutes of Health-sponsored Cholesterol Education Treatment Panel (19) recommend lowering plasma LDL-C concentration to 70 mg/dl in people at high risk for early heart attack. Concern over whether these recommended levels are too low should be tempered by the reality that the average level of plasma LDL-C in newborn infants throughout the world is only 50 to 70 mg/dl (20) and that LDL-C levels remain at or below 100 mg/dl on average throughout life in populations that consume low-fat diets (13).

Although studies of PCSK9 are still in their infancy, they suggest a new approach to enhancing the effectiveness of statins and other drugs that deplete cholesterol in the liver and raise LDL receptor number. Mouse experiments indicate that SREBP, the transcription factor that increases the expression of LDL receptors, also increases the production of PCSK9 (21, 22). Depletion of liver cholesterol activates SREBPs, thereby increasing LDL receptor numbers (17), but also increasing PCSK9 levels (23). The PCSK9 destroys some of the LDL receptors, thereby partially negating the LDL-lowering effect. A PCSK9 inhibitor should synergize with cholesterol-depletion therapy in raising LDL receptor number and lowering plasma LDL concentration (7).

Admittedly, our knowledge of the pathogenesis of atherosclerosis is incomplete, and more research is needed. We do not know precisely how LDL particles cause the inflammatory and proliferative lesions of the atherosclerotic plaque. Although we measure LDL by its cholesterol content, the most toxic component may be its fatty acids or phospholipids. Also, we cannot be certain that the atherogenic effect of PCSK9 is due solely to its LDL-elevating action. It is possible that PCSK9 also exerts a direct toxic effect on the arterial wall and that loss-of-function mutations reduce atheroscle-rosis by avoiding this toxicity as well as by lowering LDL levels. Despite these unanswered questions, the data on PCSK9 are consistent with the extensive genetic, epidemiologic, experimental, and therapeutic data that justify an aggressive public health program aimed at lowering LDL levels before the atherosclerotic process has become advanced. Early intervention may well put an end to the epidemic of coronary heart disease that ravaged Western populations in the 20th century.

References and Notes

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
View Abstract

Navigate This Article