Abstract
The paper by Ramachandran et al. 1 in this edition reports a paradoxical decrease in serum high-density lipoprotein cholesterol (HDL-c) concentration with the use of simvastatin and atorvastatin; this is of interest and also concern. HDL-c is considered cardioprotective not only because of the reverse cholesterol transport system, which helps to remove cholesterol from the peripheral tissues to the liver, but also because of other mechanisms that include increased atherosclerotic plaque stability, protection of low-density lipoprotein (LDL) from oxidation, and maintaining the integrity of the vascular endothelium. 2 A meta-analysis showed that every 0.026 mmol/L increase in serum HDL-c concentration is associated with a significant coronary heart disease risk reduction of 3% in women and 2% in men. 3
Serum HDL-c can be increased by various lifestyle changes, such as smoking cessation, regular exercise and weight loss. The fibrate drugs or nicotinic acid are sometimes used if these measures fail. There are no reports to the author's knowledge of a paradoxical decrease in HDL-c with nicotinic acid use, which can increase concentrations by about 30%, although side-effects such as flushing and gastrointestinal disturbance can make use of this drug problematic. Some causes of a low serum HDL-c are depicted in Box 1. 4
Some causes of a low serum high-density lipoprotein (HDL) cholesterol 4
Familial hypoalphalipoproteinaemia
ApoA1 abnormalities (deficiency or A1 variants)
Familial deficiency of apoA1 and apoC3
Tangier's disease
Lecithin-cholesterol acyltransferase (LCAT) deficiency
Fish-eye disease (partial LCAT deficiency)
Tobacco smoking
Physical inactivity
Hypertriglyceridaemia
Obesity
Poorly controlled diabetes mellitus
Insulin resistance and metabolic syndrome
Chronic kidney disease
Dysglobulinaemia
Certain drugs, e.g. androgens, probucol, beta-blockers (without intrinsic sympathomimetic activity), progestogens, anabolic steroids, bexarotene, retinoids
HDL is synthesized in both the intestine and liver and secreted from them as small, nascent HDL particles rich in free cholesterol, phospholipids, apolipoprotein (apo) A1 and apoE. Some HDL particles also contain apoA2. This cholesterol acquisition is stimulated by adenosine triphosphate-binding cassette protein 1 (ABC1). If the serum concentration of very low-density lipoprotein (VLDL) or chylomicrons is low, apoC is also carried in HDL, but as the serum concentrations of these lipoproteins rise, these particles take up apoC from HDL. In addition, HDL can be formed from the surface coat of chylomicrons and VLDL.
HDL contains the enzyme lecithin-cholesterol acyltransferase (LCAT), which catalyses the esterification of free cholesterol and is activated by apoA1, the predominant apolipoprotein of HDL. Most of this esterified cholesterol is transferred to LDL, VLDL and chylomicron remnants, and thus ultimately reaches the liver while some cholesterol may be stored within the core of the HDL particle and taken directly to the liver. Cholesterol ester transfer protein is involved in these processes.
There are HDL subclasses, namely nascent HDL, HDL2 and HDL3. The HDL2, which is a precursor of smaller HDL3 particles, interconvert as a result of acquiring cholesterol by HDL3 through the actions of hepatic lipase and LCAT. HDL also contains other enzymes, including paroxanase, which may have an antioxidant role. Removal of HDL may occur by endocytosis, although there are specific receptors such as the murine class B type I scavenger receptor in liver and steroidogenic tissue, such as the adrenals and gonads. Thus, HDL-derived cholesterol can be transferred in the liver and secreted in bile or taken up and utilized for steroid synthesis. 4
Over the last twenty years or so, a series of papers has reported a paradoxical decrease in HDL-c by certain fibrates and also some thiazolidinediones; both groups of drugs usually increase serum HDL-c concentrations. 5–13 This phenomenon is seen with fenofibrate, ciprofibrate and bezafibrate, although interestingly has not been described with another fibrate, gemfibrozil. It has also been reported with rosiglitazone but not to date for pioglitazone. This effect is not necessarily reproducible in the same patient, may manifest after a couple of months of drug therapy and a drug dose dependency is sometimes observed, suggesting that there may be an interaction between the drug and some other factor. The mechanism may be a reduction in HDL synthesis or a decrease in its clearance or both.
The peroxisome proliferator-activated receptors (PPARs) are a family of nuclear receptors that are regulated by fatty acids. They are relevant to insulin resistance and dyslipidaemia and act via a number of lipid pathways, including increasing lipoprotein lipase activity, reducing apoC3 and VLDL synthesis and increasing apoA1 synthesis. The PPARs can be subdivided into alpha-PPARs, which are activated by the fibrate drugs, and gamma-PPARs, which are activated by thiazolidinedione drugs. Fibrates also regulate the Rev-erb alpha gene at the transcriptional level which in turn is mediated by PPARalpha; there is thus competition between the two signalling pathways and in some individuals, and under certain circumstances, the balance may result in HDL-c reduction. 14
As with the fibrates, the mechanism of how statins may paradoxically reduce HDL-c is also not known. The paper by Ramachandran et al. does not report HDL subclasses or clarify where this statin effect may be acting in the HDL metabolic pathways described above. Statins inhibit the rate limiting enzyme in cholesterol synthesis, namely 3-hydroxy-3-methyl-glutaryl-CoA reductase, and they may also increase apoA1 production and the formation of nascent HDL particles. 15 In both primary and secondary prevention trials with various statins, only small increases in serum HDL-c were observed, varying between 5 and 8%. 16 In the Voyager study, the HDL-c raising ability of rosuvastatin and simvastatin was similar, although both were superior to atorvastatin. Indeed, increases in HDL-c were positively related to dose with the former statins but inversely related to the dose of atorvastatin. 15 The increases in HDL-c were independent of LDL-cholesterol decreases, although baseline HDL-c and triglyceride concentration and also the presence of diabetes mellitus were predictive. 17
What implications do these observations have for those working in clinical biochemistry? First, we need to ensure assays for HDL-c are suitably precise and accurate. Given concerns over current performance, this is of particular relevance as HDL-c is used in many cardiovascular risk prediction calculators such as QRISK or Framingham-based tables. 18 Furthermore, the notion that a patient can be put on a lipid-lowering drug such as a fibrate or statin and then their lipid profile not subsequently monitored is questionable, as some individuals may show a paradoxical HDL-c decline, which in turn may have a detrimental effect upon their cardiovascular risk status. It is not known how often statins cause a paradoxical HDL-c reduction and the case presented in this journal is of only one patient. Nevertheless, the statins are among the most commonly prescribed drugs and thus even if this effect is rare, there could still be significant numbers of patients manifesting the decrease. Further research is warranted to ascertain how and why this iatrogenic HDL-c lowering occurs and how common it is. There could also be therapeutic implications if different statins expressed this effect or indeed were found to have better serum HDL-c raising ability, especially if this translated into better cardiovascular outcomes.
DECLARATIONS