| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
Mike Rosenbloom Laboratory for Cardiovascular Disease, Royal Victoria Hospital, 687 Pine Ave. West, Montreal, Quebec, H3A 1A1. Fax 514-982-0686; e-mail asniderm{at}is.rvh.mcgill.ca
 |
Abstract |
|---|
 Top Abstract IntroductionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionReferences |
|---|
I am a cardiologist, and coronary artery disease dominates my discipline. The initial clinical study by my colleagues and I involving apo B was published in 1980 (2). We reported, as did Avogaro et al. before us (3), that many patients with coronary disease had apo B concentrations that were disproportionately increased compared with the serum concentrations of total or LDL-cholesterol. But attention then—and since—as McNamara et al. make clear, has centered almost exclusively on total and LDL-cholesterol, not on apo B. Unfortunately, the fact of the matter is that most patients with coronary disease do not have markedly high cholesterol concentrations. Indeed, for the most part, their concentrations of total and LDL-cholesterol are indistinguishable from those in individuals who do not have coronary disease (4)(5)(6)(7). In my opinion, that is the reason why the "normal" limits for cholesterol have decreased successively, first from the 95th to the 90th percentile (8), thence to the 75th [9], and after that all the way to the 50th percentile in my country, Canada (10). Some very reputable people say that virtually everyone in Western Society is hypercholesterolemic in comparison with societies in which vascular disease is rare—and, of course, in an important way, they are right. But where does that leave those of us who try to diagnose and treat individuals from Western societies?
By definition, hypercholesterolemia used to be the exception in coronary patients, but by changing the definition, it became the rule—the difficulty being, of course, that as sensitivity increased, specificity plummeted. Changing the definition created two true, but contradictory, realities. The first is that for large groups, risk undoubtedly increases as cholesterol concentrations mount (11). By contrast, for individuals, because of the enormous overlap of values, unless total or LDL-cholesterol is markedly increased, little prognostic information of value is obtained. The clinical difficulty that this situation creates should have meant that new tests developed to yield additional information as to risk would be greeted with enthusiasm. That was what I thought would be the case with apo B. I was wrong.
Our editorial (1) dealt with the merits of measuring apo B. The above paper by McNamara et al. deals with whether the measurements of plasma triglyceride, HDL-cholesterol, and LDL-cholesterol are standardized. Although I recognize and respect the considerable efforts made by many to improve the accuracy and precision of the measurements of the plasma and lipoprotein lipids, I differ from McNamara et al. on the meaning of "standardization." For me, as a clinician, it means that the test measures accurately what it purports to measure and that if any laboratory uses any approved manufacturer's reagents in an approved fashion, that laboratory will get the same answer as any other laboratory that does the same. By that definition, the determinations of both cholesterol and apo B have been "standardized."
But nothing in life or medicine is that simple. In the case of cholesterol, a Definitive Method exists to measure it; apo B values, on the other hand, are assigned by comparison with a primary preparation whose mass has been determined by amino acid composition. In the hierarchy of analytical precision, the measurement of cholesterol would seem ahead of apo B. Indeed, it is, but not as far as one might initially assume, given that the cholesterol Reference Methods against which the manufacturer's products are calibrated have a bias when compared with the Definitive Method. That is to say, the everyday "reference methods" for neither cholesterol nor apo B are Definitive Methods. Perhaps another word would help us here: "harmonization," by which is meant the process that ensures that different manufacturers' products give the same answers in the everyday world; measurements of both cholesterol and apo B meet that standard. But harmonization has a weak sound to it; standardization has a stronger, more scientific ring to it, even though its definition in the Concise Oxford Dictionary, "obtain by analysis specific value of (solution etc.) for purposes of comparison," comes very close to harmonization. Whatever word we choose—standardization or harmonization—must not obscure the fact that both cholesterol and apo B can be accurately and precisely measured in clinical laboratories.
However, that is not yet the case for HDL- and LDL-cholesterol and triglyceride. To be sure, primary Reference Methods (not Definitive Methods) for HDL-cholesterol and triglyceride have been developed by the CDC, but they have not yet been fully implemented, even by the CRMLN laboratories. In my view, considering the measurements of HDL-cholesterol and triglyceride as being in the same category as cholesterol and apo B measurements would demand an elasticity in the word "standardization" that the Mad Hatter might approve of, but I do not. Moreover, just how many laboratories in the world are approved by use of this standard of comparison, and how are we clinicians to know which is which as we review their reports? My understanding is that even in North America only a relatively small number of the total are "approved." And even for those that are, compliance in the US, for example, is certified only at 6-month intervals.
Indeed, it was because the measurements of apoproteins had been truly standardized (12)(13)(14)(15) that we felt the studies of Contois et al. (16)(17) were particularly important. Unfortunately, recognition of this achievement is not as widespread as it should be, with the view persisting even amongst some eminent authorities that reliable measurement of apoproteins can still be achieved only in research laboratories. Even McNamara et al., while acknowledging—indeed, even applauding—the standardization of apoprotein measurement at the beginning of their article, seem to question it at the end, evidence of just how deeply old doubts are lodged and just how difficult they are to dispel. Meanwhile, while I continue to disagree with McNamara et al. on the issue of standardization, my views are similar to theirs on the too frequent inaccuracies involved in determination of LDL-cholesterol. To overcome this problem, they appear to suggest that direct measurement of LDL-cholesterol should rapidly be adopted. On this score, I believe that although the methods proposed to date are promising, they have not been fully evaluated and, in any case, they are far from being uniformly standardized.
Now let me use this opportunity to address the even more important issue: Why is measurement of apo B not routine? That was the major thrust of our editorial [1]: Measurement of apo B provides critical information, complementary to that obtained from measuring the lipoprotein lipids, information that is essential for accurate clinical diagnosis and the right choice of therapy. The normolipidemic patient may have an increased number of LDL particles but if apo B is not measured, that will be missed. In normolipidemic individuals with an increased apo B, the risk of cardiovascular disease is increased to the same extent as in persons with type II hyperlipoproteinemia (18). The same is true for the hypertriglyceridemic patient with an increased apo B (18). Moreover, how can therapy be intelligently selected for the hypertriglyceridemic patient if apo B is not known (19)? Indeed, is any therapy required for hypertriglyceridemic patients if apo B is not increased (19)? Packard and colleagues (20) reported that the commonest dyslipoproteinemia associated with coronary disease is mild hypertriglyceridemia, low HDL-cholesterol, and increased numbers of small, dense LDL particles. Genest et al. showed that familial hypoalphalipoproteinemia can also be associated with an increased apo B concentration (21). Surely these diagnostic advantages should not be restricted to research laboratories, now that apo B assays have been standardized.
Assume for a moment that measurement of apo B does add important information to the characterization of vascular risk and the choice of therapy. Why then is it available in so few laboratories? I suggest it is because decision-making in medicine has changed radically in the past few years. There is little difficulty generating a lobby for a test where payment is assured, but it is virtually impossible to obtain support for a test for which payment is not already in place. I am not arguing that anyone can, or should, do tests for nothing. But what criteria are being used to judge whether a test will be paid for and who are applying the criteria?
Much of our decision-making has been turned over to consensus conferences, typically small groups of individuals who consider questions of the moment and then issue guidelines on how these should be managed. The consensus or guidelines are promulgated as the formal views of whatever group chose the members of the panel. The acronym of the meeting becomes the author and the authority of the report, eclipsing the identities of the individuals who actually constructed it. Much good has occurred by this route, but the limitations of the process have evaded scrutiny.
In a word—dare I say it—the process is unstandardized. Indeed, no approach I am aware of has been accepted so broadly with so little analysis of its contents and methods. From 1992 to January 1997, the Medline Database and the Health Star database list 913 consensus conferences, 141 from the NIH alone. As for guidelines, 3286 are indexed as publication types. I have not reviewed each citation, but unquestionably a number will be retrieved both as a guideline and as a consensus statement. As well, some have been published more than once. But even halving the numbers still leaves a remarkable growth industry. Moreover, the conclusions of these conferences can carry enormous consequence, both for the practice of medicine and for the economics of the practice of medicine. In some areas, many millions of lives may be affected and the allocation of many millions of dollars determined by a process not subject to peer review.
We, the readers, are usually told little or nothing about the total base of information considered, how it was analyzed, how often the members met, and for how long. The strength of evidence is often weighted, but how were these decisions reached? In general, no minority reports are presented. What then does the word "consensus" mean? How often is what comes out determined by who went in? How often does the need to achieve "consensus" mean that such gatherings become meetings of the like-minded?
Regarding lipid analyses, I believe it is fair to say that the NCEP guidelines (9)(22) drive the process around the world. Each cycle has begun with an NCEP report, after which other groups and countries respond. The reports differ, in some ways importantly, but overall they are also very similar. With respect to the last NCEP conference (22), I would be the first to point out that, at that time, important information about apo B was missing. Most critically, the measurement of apoproteins had not been standardized; moreover, clear-cut results from a large prospective study favoring apo B determinations were not available. Both of those requirements have now been met (12)(13)(14)(15)(23). But does that mean that nothing will change until this group meets again? And what if they say apo B measurements are not clinically useful? Given the authority the process has acquired, does this mean any change is put on hold indefinitely?
Delay is not neutral. The strategies that lower LDL-cholesterol concentrations lower apo B. The strategies that lower LDL-cholesterol and apo B save lives and reduce the need for bypass surgery and angioplasty (24)(25)(26)(27). Because small, dense LDL particles contain less cholesterol than normal LDL particles do, apo B is a better marker of LDL particle number than is LDL-cholesterol (28). Using apo B, therefore, does not change the LDL argument; it merely extends it. Bluntly put, by not measuring apo B, we are not treating a large number of people who could otherwise have been helped.
What about cost? No one can deny the thrust of governments everywhere to control medical costs. Who believes governments' first priority is now quality of care rather than cost of care? In the US, the situation is even more complex with, for the first time, healthcare being organized into large economic units for which the ultimate goal is profit. Given this reality, how likely is it that new tests and therapies will be introduced as early as possible? In this case, however, without developing the argument here, I believe measurement of apo B will save money as well as lives. As already noted, the key problem with the cholesterol algorithm is that unless values are markedly increased, cholesterol is a poor marker of individual risk. Many patients must be treated when only a few will benefit. Apo B now, and other markers soon (29)(30), will markedly improve our ability to recognize the truly high-risk individuals so that fewer need be better treated and overall cost will be less.
In this Counterpoint, I have ranged far from the narrow issue of standardization that prompted it. I have addressed issues, not individuals, but I am aware of the risks that criticism holds for the critic. Misinterpretation of motive can divert attention from the message to the messenger. Even so, I must express concern about the inherent risks in what seems to me to be a powerful, unseen, and dangerous trend to excessive concentration of decision-making in modern medicine. I am also concerned about the dangers of excessive commercialization of the practice of medicine. I am concerned that we are at risk of losing sight of who we are and why we do what we do. If that occurs, we will lose the essence of what we are, a profession committed to the best care of each life entrusted to us.
|
References |
|---|
|
TopAbstractIntroductionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  I. Shai, E. B. Rimm, S. E. Hankinson, G. Curhan, J. E. Manson, N. Rifai, M. J. Stampfer, and J. Ma Multivariate Assessment of Lipid Parameters as Predictors of Coronary Heart Disease Among Postmenopausal Women: Potential Implications for Clinical Guidelines Circulation, November 2, 2004; 110(18): 2824 - 2830. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Jiang, M. B. Schulze, T. Li, N. Rifai, M. J. Stampfer, E. B. Rimm, and F. B. Hu Non-HDL Cholesterol and Apolipoprotein B Predict Cardiovascular Disease Events Among Men With Type 2 Diabetes Diabetes Care, August 1, 2004; 27(8): 1991 - 1997. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Grundy Low-Density Lipoprotein, Non-High-Density Lipoprotein, and Apolipoprotein B as Targets of Lipid-Lowering Therapy Circulation, November 12, 2002; 106(20): 2526 - 2529. [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Grundy Non-High-Density Lipoprotein Cholesterol Level as Potential Risk Predictor and Therapy Target Arch Intern Med, June 11, 2001; 161(11): 1379 - 1380. [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Crouse III and C. D. Furberg Treatment of Dyslipidemia : Room for Improvement? Arterioscler. Thromb. Vasc. Biol., November 1, 2000; 20(11): 2333 - 2335. [Full Text] [PDF]
|
|
|
|
 |
|
 I. Jungner, S. M. Marcovina, G. Walldius, I. Holme, W. Kolar, and E. Steiner Apolipoprotein B and A-I values in 147 576 Swedish males and females, standardized according to the World Health Organization–International Federation of Clinical Chemistry First International Reference Materials Clin. Chem., August 1, 1998; 44(8): 1641 - 1649. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
