Standardization of Apolipoprotein B, LDL‐Cholesterol, and Non‐HDL‐Cholesterol

ABSTRACT

Concern continues about whether the measurement of apolipoprotein B (apoB) is adequately standardized, and therefore, whether apoB should be applied widely in clinical care. This concern is misplaced. Our objective is to explain why and what the term “standardization” means. To produce clinically valid results, a test must accurately, precisely, and selectively measure the marker of interest. That is, it must be standardized. Accuracy refers to how closely the result obtained with 1 method corresponds to the result obtained with the standard method, precision to how reproducible the result is on repeated testing, and selectivity to how susceptible the method is to error by inclusion of other classes of lipoprotein particles. Multiple expert groups have determined that the measurement of apoB is adequately standardized for clinical care, and that apoB can be measured inexpensively, using widely available automated methods, more accurately, precisely, and selectively than low‐density lipoprotein cholesterol or non‐high‐density lipoprotein cholesterol. ApoB is a standard superior to low‐density lipoprotein cholesterol and high‐density lipoprotein cholesterol because it is a defined molecule, whereas the cholesterol markers are the mass of cholesterol within lipoprotein particles defined by their density, not by their molecular structure. Nevertheless, the standardization of apoB is being further improved by the application of mass spectrophotometric methods, whereas the limitations in the standardization and, therefore, the accurate, precise, and selective measurement of low‐density lipoprotein cholesterol and high‐density lipoprotein cholesterol are unlikely to be overcome. We submit that greater accuracy, precision, and selectivity in measurement is a decisive advantage for apoB in the modern era of intensive lipid‐lowering therapies.

Nonstandard Abbreviations and Acronyms

CDC

Centers for Disease Control

LSP

Lipids Standardization Program

non‐HDL‐C

non‐high‐density lipoprotein cholesterol

Atherosclerotic coronary artery disease has become preventable, and this has transformed cardiovascular care. Trapping of apolipoprotein B (apoB)‐containing lipoprotein particles with the cholesterol they contain drives the evolution of atherosclerosis within the arterial wall, from an innocent collection of lipids to a complex lesion that can trigger a clinical event. Most of the cholesterol in plasma is present within low‐density lipoprotein (LDL) and lipoprotein(a) particles, and this accounts for the strong relation of LDL‐C to cardiovascular risk and to the benefit from therapies that lower LDL‐C. Accordingly, LDL‐C has become the premier marker of the lipoprotein‐associated risk of cardiovascular disease, the standard of care promoted by all the prevention guidelines, and the measure accepted by physicians and patients worldwide. Indeed, the elucidation of the role of LDL‐C in atherosclerosis is one of the signal triumphs of modern medicine.

But does LDL‐C deserve to remain the preeminent lipid marker for clinical care? In 2019 the European Society of Cardiology and the European Atherosclerosis Society, based on the results of prospective observational studies, Mendelian randomizations analyses, and meta‐analyses of statin trials, concluded that apoB is a more accurate marker than LDL‐C or non‐high‐density lipoprotein cholesterol (non‐HDL‐C) of cardiovascular risk and the adequacy of therapy to reduce cardiovascular risk. ¹ Since then, the evidence has continued to mount in favor of apoB. To be sure, Helgadottir et al have recently reported a Mendelian randomization analysis, which showed non‐HDL‐C was a more accurate marker of cardiovascular risk than apoB. They also concluded cholesterol‐enriched apoB particles were more atherogenic than cholesterol‐depleted apoB particles. ² Their results must be considered, but they conflict with a large body of work to the contrary.

The most important additions to the evidence base have been the 4 clinical reports published since 2021, all demonstrating the superiority of apoB over LDL‐C and non‐HDL‐C to identify cardiovascular risk. The first is a large, prospective discordance analysis of patients on statin therapy, ³ then 3 randomized clinical trials: the Improved Reduction of Outcomes: Vytorin Efficacy Interventional Trial (IMPROVE‐IT) study, ⁴ a test of statin‐ezetimibe therapy versus statin therapy, and Futher Cardiovascular Outomes Research with PCSK9 inhibition in Subjects with Elevated RIsk (FOURIER) ⁴ and Evaluation of Cardiovascular Outcomes After an Acute Coronary Syndrome During Treatment With Alirocumab (ODYSSEY OUTCOMES) ⁵ the 2 trials of statin‐proprotein convertase subtilisin/kexin type 9–inhibitor therapy versus statin therapy. Taken together with the previous body of evidence, ⁶ the superiority of apoB over the cholesterol markers to evaluate cardiovascular risk and direct therapy has been firmly established. Notwithstanding all this evidence, it is still stated that failure to standardize the measurement of apoB means immediate clinical application is inappropriate. This caveat, which we will demonstrate is not scientifically valid, appears in the ODYSSEY OUTCOMES article itself ⁵ and in the Commentary on this article. ⁷

Accordingly, this review will focus on the issues of measurement and standardization so that clinicians and other readers can more fully understand these critical concepts in clinical care. What does standardization entail and is the statement that apoB is not standardized correct? Are the measurements of LDL‐C and non‐HDL‐C standardized? To answer these questions, we will first briefly review the fundamental criteria to judge the accuracy, precision, and selectivity of the measurements of lipids and apoB.

What Is Standardization?

To be standardized, a medical test must conform to the requirements of an international endorsed Reference Measurement System in order to assure consistency of test results in different laboratories at different times. A sequence of steps is involved, beginning with the selection of the standard and ending with the performance of the test in the clinical laboratory. Any error in any step will carry over as an error in the next and so the total error in measurement is the sum of all the errors in all of the steps in measurement. Laboratory errors can be divided into those that are preanalytical (errors that relate to biological variations or procedures to obtain and preserve a sample) and analytical (errors in the precision, accuracy, or selectivity of the measurement). These relate to the definition of the measurand, kit design, instrument application, and inherent instrument performance.

Standardization ensures that the results obtained by 1 in vitro diagnostics manufacturer or laboratory will conform to the results obtained by another test manufacturer or laboratory. But standardization goes much further. Standardization demands that measurement of the concentration of a sample be calibrated based on an agreed method to measure the mass of a purified and stable reference sample of the marker. To be standardized, the test result must be explicitly traceable to a higher‐order standard and this relationship must be stable and predictable over the range of concentrations that are measured for the marker. Put more formally, standardization can only be accomplished using an internationally endorsed Reference Measurement System with a calibration hierarchy according to ISO 17511. ⁸

Standardization is about precision, accuracy, and selectivity. By precision is meant reproducibility: how closely will repeated measures agree with each other? Imprecision is estimated by the SD and coefficient of variation, which is the SD of repeated results divided by the average concentration. Method validation typically includes assessment of within‐day precision (repeatability) and total imprecision (within‐laboratory imprecision). Accuracy refers to systematic differences (bias) between the result obtained with 1 method and the result obtained by the accepted reference method. Unlike imprecision, which can only be minimized, but not avoided, bias, in theory, can be eliminated by standardization.

Errors due to imprecision and bias are independent of each other. Thus, a method can be unbiased, but imprecise, and vice versa. Only if medical tests are standardized will results, within allowable measurement uncertainty, obtained from 1 laboratory using 1 method, be comparable to results obtained from another laboratory using a different method to measure the same marker. Even if the same method by the same manufacturer is used, a robust standardization program will ensure that new lots of critical reagents and calibrators are validated so that no bias occurs over time. In the absence of effective ongoing standardization, there is no assurance that test results at 1 moment in time will be exactly comparable to measurements at another.

There is yet another aspect of the measurement of lipoprotein particles and lipids: selectivity. However, the signal importance of selectivity as a criterion to assess the validity of cholesterol assays in particular has not been widely recognized. Therefore, we will briefly explain its significance. A marker must be measurable, accurately and precisely, over the full range of concentrations, irrespective of pathological variations encountered clinically. Pathological variation in this case refers to the different combinations of apoB‐containing lipoproteins that characterize the different dyslipoproteinemias. Thus, the triglyceride‐rich apoB lipoprotein particles include intact chylomicrons, normal chylomicron remnants, cholesterol‐enriched chylomicron, and very‐low‐density lipoprotein remnants characteristic of type III hyperlipoproteinemia, as well as the full range of different size very‐low‐density lipoprotein particles. Each of these particles contains 1 molecule of apoB but different amounts of cholesterol and triglycerides.

Total apoB can be accurately and precisely measured over the full range of concentrations encountered in whatever combination of these lipoprotein particles is present. However, this is not the case for LDL‐C or non‐HDL‐C, particularly at low concentrations, whatever method, direct or indirect, is used to measure or calculate their concentrations. ⁹ , ¹⁰ , ¹¹ LDL‐C and high‐density lipoprotein cholesterol (HDL‐C) are estimates of the cholesterol within density‐defined lipoprotein macromolecular complexes. These density ranges, particularly in dyslipidemic samples, may contain other lipoprotein classes, which also carry cholesterol. If so, the measurement is nonselective and is no longer a valid estimate of LDL‐C or HDL‐C, respectively. Unfortunately, nonselectivity (or more bluntly, contamination of 1 lipoprotein class with others) is a major problem, particularly at lower levels of LDL‐C and non‐HDL‐C. By contrast, apoB corresponds to an exact molecular structure and, therefore, total apoB is selectively measured.

Other arguments against the use of apoB that have been made in the past include lack of availability of measurements, lack of support in clinical epidemiologic studies and cholesterol‐lowering therapy trials, lack of reliability and reproducibility of assays, lack of support from guidelines, and expense. ¹² These concerns have been overcome. The equipment and expertise necessary to measure apoB are present in virtually all clinical laboratories and there are many Food and Drug Administration and European Union cleared assays available from all the major diagnostic companies. As discussed above, there is compelling evidence from epidemiological studies and randomized clinical trials that support apoB testing. The use of apoB has been approved by major guidelines and consensus groups and the cost of apo B testing would only marginally affect the total cost of care. ¹³ Moreover, cost will decrease with increased utilization.

ApoB measurement has been acknowledged to be important in those conditions recognized to be associated with abnormal LDL composition, such as obesity, hypertriglyceridemia, metabolic syndrome, and diabetes. But the reality is that cholesterol‐depleted apoB particles can commonly be present at any level of plasma triglyceride. ¹⁴ Finally, because statins decrease cholesterol content more than LDL particle number, ¹⁵ apoB measurement improves the assessment of residual risk in primary and secondary cardiovascular disease prevention.

These issues were not critical clinically until recently. However, with the development of potent cholesterol‐ and apoB‐lowering agents that substantially reduce cardiovascular risk, this is no longer the case. Selection of subjects for primary prevention with lipid‐lowering agents and decisions regarding adequacy of therapy for all patients being treated are based, to an important extent, on the measured concentration of a lipid marker, be it LDL‐C, non‐HDL‐C, or apoB. Measurement errors, whether due to inaccuracy, imprecision, or nonselectivity, diminish the quality of care.

Standardization of APOB

In 1994, the International Federation of Clinical Chemistry and the World Health Organization (WHO/IFCC) endorsed the standardization of the measurement of apoB, as reported by Dr. Santica Marcovina, Dr. John Albers, and their colleagues. ¹⁶ Because stable, lipid‐free preparations of apoB could not be produced, the standard reference materials were native human LDL particles isolated by ultracentrifugation with a “narrow cut” density 1.030 to 1.050 g/L that excludes intermediate density lipoprotein and HDL. The protein mass of this material, measured by Lowry assay, became the secondary reference standard for apoB. This compromise is a limitation. Ideally, as for example with cholesterol, the reference material would be isolated in pure crystalline form, allowing accurate gravimetric measurement. This could not be achieved with apoB. Nevertheless, a stable, commutable IFCC‐WHO reference standard was produced, and at a practical level, this has proven adequate to ensure acceptable standardization of apoB measurements, with commutability being the closeness of agreement between results for a reference material and clinical samples when measured by 2 or more measurement procedures.

Manufacturers that opt for Centers for Disease Control (CDC) certification of their lipid and lipoprotein tests are listed at the CDC website. However, participation of in vitro diagnostics manufacturers in the WHO/IFCC apoB standardization program has been voluntary. While the procedure developed by Dr. Marcovina and colleagues has ensured that the methods met the analytical performance requirements of the standardization process, given the importance of apoB in patient care, further work on refinement of the standard itself is justified. Accordingly, efforts by IFCC to use direct measurement of apoB by mass spectrometry in a primary Reference Measurement System are well under way and will allow the true SI‐traceable standardization of apoB, as defined by ISO 17511:2020. ¹⁷

Creating a compulsory monitoring process is an equally important objective to detect lot‐to‐lot drift or shift that may occur over time. Nevertheless, while improvement is possible, perfect should not be the enemy of good. As will be detailed later in this review, multiple consensus groups, External Quality Assessment (EQA) organizers, CDC Lipids Standardization Program, including those expert in clinical chemistry, have declared the standardization program for apoB, developed by the WHO/IFCC, meets the requirements for application of apoB in clinical care.

Are the Measurements of LDL‐C and NON‐HDL‐C Standardized?

Yes and no. Yes, because certification programs run by the CDC Cholesterol Reference Method Laboratory Network and the CDC Lipids Standardization Program have created adequate standards for the measurement or calculation of LDL‐C and non‐HDL‐C. To be sure, compromises have been made: most significantly, indirect reference measurement based on triglycerides and HDL‐C. Nevertheless, Lipids Standardization Program monitors trueness and precision of measurements performed in research studies and routine clinical laboratories by providing blinded samples traceable to the CDC Reference Laboratory for the measurement of total cholesterol, glycerides (triglycerides), HDL‐C, apoA‐I, and apoB.

The Cholesterol Reference Method Laboratory Network uses a network of endorsed calibration laboratories to help in vitro diagnostics manufacturers standardize their tests with CDC reference measurement procedures for total cholesterol, triglycerides, HDL‐C, and LDL‐C. The reference method for LDL‐C, commonly called beta quantification or “BQ,” involves ultracentrifugation at a density of 1.006 g/L leaving intermediate density lipoprotein, LDL, lipoprotein(a), and HDL in the infranatant. The apoB‐containing lipoproteins are removed from the infranatant by chemical precipitation, leaving HDL in the supernatant, and LDL‐C is calculated as infranatant cholesterol (intermediate density lipoprotein, LDL, lipoprotein(a), HDL) minus supernatant cholesterol (HDL). Cholesterol in these fractions is measured by the Abell Kendall cholesterol reference method. Therefore, by definition, LDL‐C includes intermediate density lipoprotein‐C and lipoprotein(a)‐C. Currently, genetically determined lipoprotein(a) has undergone a renaissance and selective lipoprotein(a) tests are now considered a first step towards precision cardiovascular diagnostics.

Notwithstanding all this effort, a critical limitation remains. Only normotriglyceridemic samples have been used to validate these assays, and high triglycerides >200 mg/dL (2.3 mmol/L), diabetes, and other disorders associated with abnormal lipid and lipoprotein profiles cause unacceptable between‐method variability and bias with most of the direct LDL‐C and HDL‐C assays commercially available, particularly at low concentrations. As well, multiple methods to calculate LDL‐C have been developed, ¹⁸ but there is no way for the clinician to account for differences in results from different laboratories, nor is it possible for any calculation method to take entirely into account the extreme differences in composition of the triglyceride‐rich lipoproteins that occur in the rare, but clinically significant dyslipoproteinemias, such as type III hyperlipoproteinemia.

Direct assays based on different principles may select different subclasses of LDL or HDL or show nonspecific reactivity to other lipoprotein particles, depending on the reagents used to selectively quantify the cholesterol in the particles intended to be measured. ⁹ , ¹⁸ , ¹⁹ Importantly, no evidence has been published demonstrating the clinical superiority of directly measured LDL‐C over calculated LDL‐C. van Deventer and colleagues, for example, report that most direct LDL cholesterol methods failed to improve cardiovascular disease risk classification over calculated LDL cholesterol estimations. ²⁰ Yet direct measurement of LDL‐C and its associated costs have been accepted without evidence of benefit.

HDL‐C is currently measured in clinical care by a variety of direct methods. Although well “standardized,” none are selective and significant variance among them remains. Errors due to imprecision and nonselectivity in the measurement of HDL‐C can be substantial and therefore can be an important source of error in the calculation of LDL‐C and non‐HDL‐C. ²¹ As noted above, the impact of this error will be larger at lower concentrations of LDL‐C and non‐HDL‐C, concentrations that in the modern era of potent lipid‐lowering agents are now common.

Thus, while LDL‐C and non‐HDL‐C assays have been harmonized, they have not actually been standardized.

Why Is Measurement of APOB Thought Not to be Adequately Standardized?

In support of their concern about the measurement of apoB, the authors of the ODYSSEY OUTCOMES study ⁵ cited a recent consensus report on laboratory measurement of lipids and apoB by the National Lipid Association. ²² The issue is covered in a single sentence: “Unfortunately, apoB has not been formally standardized.” Formally is not defined. The WHO/IFCC‐approved publication of the method to standardize apoB ¹⁶ is not acknowledged; nor is the decades‐long work of the Northwest Research Laboratory, CEQAL, and others to monitor the standardization of apoB. The statements by the American Association for Clinical Chemistry ¹⁰ , ²³ the joint statement by the European Federation of Laboratory Medicine/European Atherosclerosis Society, ¹¹ the 2019 European Society of Cardiology/European Atherosclerosis Society Guidelines ¹ that declare that the measurement of apoB is adequately standardized for routine clinical care, are not cited. The National Lipid Association statement stands apart from the multiple other consensus reports, which have reviewed the laboratory performance of apoB, LDL‐C, and non‐HDL‐C in detail.

The commentary by Elshazy and Quispe ⁷ cites the study by Delatour et al, ²⁴ which demonstrates significant differences among multiple methods using radically different methodologies to measure apoB. Delatour and colleagues note the need to standardize these different methods against the current reference method and the need to further refine the present reference standard by moving to a SI‐traceable mass spectrometric‐based reference system approach, a goal for all clinically important apolipoproteins. ²⁴ These are valid points, but they do not negate the clinical validity of results of the ODYSSEY OUTCOMES study ⁵ as well as the other recent studies ³ , ⁴ demonstrating the clinical utility of the standardized immunological measurements of apoB using the WHO/IFCC process. ¹⁶ Finally, although not cited by either the authors of the ODYSSEY OUTCOME study ⁵ or the authors of the accompanying commentary, ⁷ there is the report by Cao et al ²⁵ that documented significant difference among 3 immunological methods to measure apoB. Unfortunately, this report does not state whether the manufacturers of all 3 methods participated in the apoB standardization program, a critical point obviously, and further evidence that in the future, participation in such programs should be mandatory, not voluntary. This is illustrated in the national quality assessment program of clinical laboratories in The Netherlands, in which 28% of participating laboratories exceeded the recommended bias for apoB, whereas 68% exceeded the recommendation for LDL‐C compared with a Cholesterol Reference Method Laboratory Network reference measurement. ²⁶

Conclusions

This essay reviews the concept and advantages of standardization of laboratory markers. We submit that the measurement of apoB has been adequately standardized for clinical use. For this, we owe an extraordinary, unpayable debt to Dr. Santica Marcovina and her colleagues for all they have accomplished. But the work to improve the assay of apoB even further goes on. The efforts to develop a new Primary Reference System with a mass spectroscopy–based primary reference method will make apoB the first truly standardized lipoprotein marker, a feat unlikely ever to be accomplished for LDL‐C, HDL‐C, or non‐HDL‐C. We have focused on standardization, but it is the overall clinical performance of an assay that should be determinative. Multiple lines of evidence have demonstrated that apoB is a more accurate marker of cardiovascular risk and the adequacy of lipid‐lowering therapy to lower cardiovascular risk than LDL‐C and non‐HDL‐C. Moreover, the American Association of Clinical Chemistry, ¹⁰ , ²³ the European Federation of Laboratory Medicine, the European Atherosclerosis Society ¹¹ and the European Society of Cardiology ¹ have all concluded that apoB can be measured more accurately, particularly at low concentrations, than with LDL‐C or non‐HDL‐C.

Equitable and rational application of advanced potent but costly therapies in clinical care demands accurate measurement of the decisional marker. Trapping of apoB particles within the arterial wall is the fundamental cause of atherosclerosis. ⁶ , ²⁷ Variation in the mass of cholesterol within apoB particles is substantial and is the consequence of multiple metabolic processes. ⁶ Accordingly, no improvement in the measure of LDL‐C or non‐HDL‐C will overcome the biological pathophysiological superiority of apoB over LDL‐C and non‐HDL‐C as a marker of cardiovascular risk. The fact that apoB can be measured more accurately, precisely, and selectively than LDL‐C and non‐HDL‐C using widely available, inexpensive, standardized methods argues strongly for its broad application in clinical care.

Sources of Funding

None.

Disclosures

None.