Not long ago, researchers were confident that they understood the role of high-density lipoprotein (HDL) in atherosclerosis: While low-density lipoprotein (LDL) collects cholesterol and uses it to form artery-obstructing plaques, HDL polices blood vessels, scavenging cholesterol from places it doesn’t belong. This view was supported by study after study showing that higher levels of cholesterol associated with HDL (HDL-C) were linked to lower risk of cardiovascular disease (CVD).

In recent years, however, the research community has had to adjust its view of HDL. First, multiple studies of HDL-C-elevating drugs showed no protection against CVD. Second, observational studies suggested that very high levels of HDL-C are also associated with a higher risk of atherosclerosis.

In short, “we realized HDL is still an undiscovered country,” said Carlos Santos-Gallego, MD, a cardiologist and researcher at Mount Sinai School of Medicine in New York City. These findings have prompted clinical lab researchers to ask whether HDL is being measured in the right way—and how those measurements should be interpreted.

HDL-C: The Current Standard

At present, clinicians interested in patients’ HDL levels typically measure blood levels of cholesterol associated with ApoA1, the major protein component of HDL particles (HDL-P). Although HDL-C is only a surrogate for HDL, and the relationship between HDL-C and CVD risk is not as simple as once thought, most experts believe it is still a valid biomarker.

“Since measurements of HDL-C are relatively cheap, reproducible, and supported by large data sets, they are going to remain the measure of choice for now,” said Hagai Tavori, PhD, an assistant professor of medicine at Oregon Health Sciences University’s Center for Preventive Cardiology in Portland.

Nevertheless, HDL-C levels mask a tremendous amount of diversity in HDL-P. “HDL carries more than 200 different species of lipids and 80 different proteins,” said Alan Remaley, MD, PhD, a senior investigator and section chief of the lipoprotein metabolism laboratory at the National Heart Lung and Blood Institute in Bethesda, Maryland. “Two people can have identical HDL-C levels, but it doesn’t mean their particles are the same in size or function.”

The HDL-C assays currently in use also have some limitations. In the past 10 to 15 years, labs have moved from older assays that involve chemically precipitating HDL-C to more streamlined assays that directly measure HDL-C. However, research has shown that these direct assays can be problematic when used on individuals with dyslipidemia or CVD.

“The tests are not achieving the analytical performance goals that we’d like,” said Remaley, though he notes that manufacturers are tweaking the assays to minimize interference by other cholesterol fractions, a common source of problems.

HDL-P: A Better Metric?

“Many of us think that HDL-C is a terrible way to quantify these particles,” said W. Sean Davidson, PhD, professor of pathology and laboratory medicine and vice chair of research at the University of Cincinnati College of Medicine in Ohio. “HDL-P number is better, and some studies show marginally better predictions with HDL-P than HDL-C.”

One method of measuring HDL-P is already Food and Drug Administration-approved: nuclear magnetic resonance (NMR) spectroscopy of plasma or serum samples. This assay not only quantifies the number of HDL-P, it breaks down the distribution of subspecies by size: small, medium, and large.

At present, this test is rarely ordered and “we don’t really know what the large, medium, and small HDL particles mean,” said Remaley. However, current research suggests that smaller HDL-P are associated with CVD, and learning more about how particle size changes in response to disease could yield important insights into the role that HDL plays in atherosclerosis. Another benefit of NMR is that “at hardly any additional cost, you can get LDL levels and a lipid panel,” from the same sample, said Remaley.

NMR measurement of HDL-P has its drawbacks as well.

Jay Heinecke, MD, professor of medicine and Karasinski endowed chair in metabolic research at the University of Washington in Seattle, warned that this method has not been sufficiently validated as a metric for HDL-P in the peer-reviewed literature. He also noted that different companies are using their own instruments and algorithms, which makes it difficult to compare results. In addition, “NMR still groups particles by physical properties rather than their business end, the proteins,” said Davidson. Finally, few facilities currently perform this test, limiting the number of doctors who order it, said Santos-Gallego.

Other methods of measuring HDL-P and its size subspecies are also being explored, among them calibrated ion mobility analysis and 2-D gel electrophoresis. However, these methods are still far from being ready for clinical use.

The Next Frontier

In addition to measuring the number of HDL-P in a sample, interest is growing in measuring the functional capacity of the particles. “Cholesterol efflux capacity has been shown to predict both incident and prevalent CVD risk,” said Heinecke. Consequently, the scientific community is now leaning toward measuring efflux capacity, said Tavori.

This type of testing is currently confined to the research setting, however, and scaling it up for clinical use will be challenging. “The test is based on cultured cells, is difficult to do on large numbers of samples, and has not been validated as giving similar results in different labs,” said Heinecke.

In addition, the results yielded by different methods of measuring efflux capacity—such as using tritium-based cholesterol esters or fluorescence-labeled reagents—are only moderately correlated, said Santos-Gallego, and it is not yet clear which method is best.

This type of testing also has more limitations. For example, prior to testing, patient samples are depleted of other lipoproteins, but other proteins remain. “Some of these plasma proteins have been shown to have overlapping action with HDL,” said Tavori, which complicates the interpretation of test results. In addition, he warned that efflux capacity in a blood sample may not reflect the biological processes at work within vessel walls, where atherosclerosis occurs.

Heinecke also cautioned that relying on efflux capacity may lead to some of the same problems as relying on HDL-C levels, as drugs that increase HDL-C but fail to provide a clinical benefit also increase efflux capacity.

Matching HDL Particle Function With Outcomes

As novel HDL tests are refined and become more widely used, the research community will learn more about how these particles function in atherosclerosis.

To understand why past trials of HDL-C-increasing drugs have been so disappointing, “We need to know the key HDL players: both the particle sizes and protein compositions,” said Heinecke. “Both [cholesterylester transfer protein] inhibitors and niacin produce large HDL particles within the body, but there is strong evidence that small particles are the major pathway for promoting cholesterol efflux from macrophages, at least in vitro. These agents likely lower the concentration of small HDL particles.”

New tests that improve researchers’ knowledge of HDL’s function may result in even better tests. “Once we know the distinct HDL particles associated with disease protection, we can design clinical tests that look for those particles, based on specific protein content,” said Davidson.

That said, it likely would take years before a new test could be widely adopted clinically. Tavori noted that any new test would have to demonstrate superior value to HDL-C in terms of risk prediction, as well as demonstrate major improvements in feasibility, reproducibility, and cost relative to the HDL-C alternative assays available today.

Despite the work that lies ahead, Davidson remains optimistic. “I think the field needs to take a breath, stop, and do the hard work of unpacking what actually makes up HDL,” he observed. “But I think we are on the right track.”

Kristin Harper is a freelance writer in Seattle, Washington. +Email: [email protected]