Introduction: Apolipoproteins and guidelines for prevention of cardiovascular disease

As the use of drugs to lower plasma cholesterol has increased, it has become increasingly important to identify those persons most likely, and least likely, to benefit from them. The algorithms for calculating risk of cardiovascular disease used in Europe [1], as well as in the United States [2], do this reasonably well for persons at very low- or very high-risk, but they do not perform very well for persons in the intermediate range of risk [3]. This is an important problem because in countries wealthy enough to finance wide-scale therapy with drugs such as the statins, most people are at intermediate risk. There are currently two approaches to solving it. One is some kind of imaging of the coronary or other arteries to see how atherosclerotic they might be. The European and US guidelines both recognize that techniques such as measurements of coronary calcium by computer tomography or ultrasonic measurement of the thickness of the carotid intima-media can be employed in sophisticated risk assessment of individual patients, but the guidelines have not incorporated imaging formally into their algorithms [1, 2]. Scott Grundy has suggested a way to do that [4], however, and the scientific support for imaging is getting stronger. A recent study suggested, for example, that measurements of coronary calcium differentiated quite well between persons destined or not destined to have atherosclerotic cardiovascular disease events, even in the intermediate range of risk as defined by the Framingham risk function [3]. Whether future guideliness committees will recommend imaging for routine risk assessment should depend, therefore, primarily on answers to questions about safety, availability and costs. The other approach is refinement of biochemical measurements of lipoproteins or other plasma substances such as thrombogenic factors, markers of inflammation, homocystein, etc. To be discussed, in this issue of Journal of Internal Medicine is one of the options for refining the measurement of lipoproteins. The idea is to replace measurements of the cholesterol content of low density and high density lipoproteins by measurements of the major protein components of these lipoproteins, apolipoprotein B (apo B) and apolipoprotein A1 (apo A1), respectively, and to combine the measurements in the ratio of apo B to apo A1. The higher the ratio, the higher the risk. The ratio makes good sense, at least intuitively, as it combines an apparently harmful with an apparently protective lipoprotein variable. Concentrations of apo B are a true measure of concentrations of atherogenic lipoproteins such as LDL and the smaller varieties of VLDL, and apo A1 is the major apolipoprotein of HDL, low concentrations of which are associated with high risk of disease. The idea is not new. In its original form, it was the ratio of total cholesterol (or LDL cholesterol) to HDL cholesterol, and it was well supported by observational epidemiology [5, 6]. That ratio has therefore been widely used to assess risk, and, indeed, it is included in the current European guidelines [1]. Nevertheless, the arguments for the version of the ratio based on protein measurements have been further buttressed by results of more recent epidemiological studies [7, 8], better understanding of lipoprotein metabolism as well as by improvements in laboratory technology and standardization, and there is more support in the articles that follow [9-14]. There are problems, of course. A particular variable in a set of clinical guidelines, blood pressure or cholesterol, for example, does not necessarily have to indicate goals of therapy as well as be a good marker of risk. Low social class increases risk of coronary heart disease, but ordinarily no one expects a physician to be able to elevate a particular patient to higher social class. A high ratio of apo B to apo A1 could therefore be used only as a marker of increased risk and not necessarily also as an index of how well the physician is doing in lowering his patient's risk. Moreover, if one wants to define a particular value of the ratio as a target of drug therapy to affect plasma lipoproteins, a problem is immediately apparent: does it make a difference whether apo B/apo A1 is lowered by decreasing the numerator or by increasing the denominator? The evidence to support the lowering of the numerator, apo B, is very strong, inter alia because lowering the apo B lipoproteins is what the statins do [15]. In large clinical trials of statins to lower LDL, the association of high concentrations of apo B with cardiovascular events in both the placebo and treatment groups is stronger than that of LDL [16]. In some of these trials, the apo B/apo A1 ratio does even better [17]. Nevertheless, it is tricky business translating these observations into lowering ratios, be they apo B/apo A1 or total cholesterol/HDL cholesterol, as new targets of therapy. These ratios could be lowered with unchanged concentrations of LDL if only concentrations of HDL are increased, but the proposition that patients will benefit from drug therapy to increase HDL, or apo A1, is supported mainly by in vitro and animal experimentation [18], and the evidence from clinical trials to test the hypothesis explicitly is very limited [19]. We also do not know whether we should increase HDL or whether, irrespective of HDL concentrations, we should try to increase flux of cholesterol through the various routes of ‘reverse cholesterol transport’ from arteries, arterial macrophages in particular, and other organs to the liver. The two are not necessarily proportional, just as the amount of water (concentration of HDL) in a bath tub is not a measure of how much water is flowing into and out of the tub (flux of cholesterol). One of the methods by which to increase concentrations of HDL (and apo A1) is by inhibiting the transfer of cholesterol from HDL to VLDL by cholesteryl ester transfer protein, and CETP inhibition therefore almost certainly decreases reverse cholesterol transport. The merits of this kind of drug therapy are currently being tested in clinical trials [20], and at this time we still do not have a fully satisfactory answer to the question of whether to use the ratio of apo B to apo A1 (or total cholesterol to HDL cholesterol) as a target of therapy as well as a measure of risk. Another question for guidelines committees is how easily a new measure of risk or goal of therapy can be expected to be accepted by physicians in general practice and in cardiology, diabetology, neurology etc. In a debate with Alan Sniderman [21], who argues the case for using apo B in clinical guidelines, Margo A. Denke, who coauthored the latest round of the US National Cholesterol Education Program guidelines [2], advances as her single most important argument against apo B that ‘the guidelines are named for cholesterol,’ and that confusion would result from replacing cholesterol with apo B [22]. Her argument might seem less than serious to Europeans, who happen not to have included the word ‘cholesterol’ in the title of their guidelines [1], but she is certainly right in drawing attention to continuity of concepts as guidelines develop over time. How well, then, does the apo B/apo A1 ratio fit into the continuity of physician concepts of risk assessment and targets of therapy? An attractive feature of the numerator, already mentioned, is the intuitively understandable, and correct, concept that concentrations of apo B are a measure of the number of atherogenic lipoproteins in a volume of plasma. Physicians could therefore easily accept the argument for using apo B as a measure of risk and target of therapy. They might also accept the idea of a ratio quite readily, especially because the total cholesterol/HDL cholesterol ratio has been in use for a long time. Everyone likes dichotomies. Good and bad, yin and yang. Indeed, the problem with the ratio is not pedagogical. As argued above, it is biological and clinical. The most important problem of pedagogy is probably the public's. Cholesterol is a household word, apo B is not, and it is naive to contemplate educating the world's public in apolipoproteinology. The wisest course might be to maintain cholesterol as the focus for patients and to use the apo B to apo A1 ratio as the physician's fine tuning of risk assessment and, perhaps, targets of therapy. Politically correct or not, physicians sometimes must know more than their patients do. No conflict of interest was declared.