Is apolipoprotein B the best measure of CVD risk?

Amy Carpenter Aquino

November 2022—The evidence in favor of measuring apolipoprotein B routinely, with other lipid parameters, is now so overwhelming, says cardiologist Allan Sniderman, MD, that he believes it’s unreasonable to deny patients the advantage of apoB.

“If evidence is what counts,” he says, “then the care Americans receive should include apoB.” ApoB measurement would “simplify, unify, and clarify,” and without it patients and physicians have only a partial picture, he says.

Dr. Sniderman, Edwards professor of cardiology and professor of medicine, McGill University, speaking at the AACC annual meeting and recently with CAP TODAY, said, “When your LDL or non-HDL cholesterol are measured, you may not have an accurate idea of the risk posed to you.” The true cardiovascular disease risk may be higher or lower, “and in this day and age,” he said, “that should be unacceptable.”

In fact, measurement of apoB, which is the sum of all atherogenic particles, should be the primary marker, he said. “It’s not the whole story. Lp(a) matters, triglycerides matter for pancreatitis and type three hyperlipoproteinemia. Moreover, apoB is certainly not the end of the road in characterization,” he said. But without an apoB measurement, “you haven’t even begun to take the first step on the road,” and it’s one that “will take you a long way down the road.”

ApoB is better than other cholesterol markers because the number of apoB particles in the lumen is the primary determinant of the number that get into the arterial wall and become trapped, said Dr. Sniderman, who is also director of the Mike Rosenbloom laboratory for cardiovascular research, Royal Victoria Hospital, Montreal. And trapping of the apoB particles within the wall is the fundamental cause of atherosclerosis. Cholesterol within the trapped particles is strongly proatherogenic, and the relationships demonstrated between risk and the levels of LDL-C and non-HDL-C are true. So too is the correlation between lowering levels of LDL-C and non-HDL-C and better outcomes. “But it’s also equally true that cholesterol within the wall only got there within an apoB particle. It got transported in and trapped.”

Cholesterol is “not the only poison in the particle,” Dr. Sniderman said. “The apoB, when it’s oxidized and breaks and is degraded, is strongly proatherogenic,” as are the phospholipids when they’re oxidized. “It’s the particle that matters.”

Smaller, cholesterol-poor apoB particles are trapped more avidly within the arterial wall than larger apoB particles and bind more easily to the glycosaminoglycans, he said. Larger, cholesterol-rich apoB particles deposit more cholesterol when trapped within the arterial wall. “VLDL particles are atherogenic—they tend to have more cholesterol in them.” But when looked at in totality, “they tend to be equally atherogenic within the limits of detection.”

“Maybe there are some differences,” Dr. Sniderman said. “But the tools we have to measure differential atherogenicity don’t allow us to separate anything out. So at the moment, apoB is the accurate sum of all the atherogenic particles, and that’s the most important measure you can make.”

The 2019 European Society of Cardiology/European Atherosclerosis Society guidelines for the management of dyslipidemias to reduce cardiovascular risk concluded that apoB is a more accurate marker of risk and adequacy of therapy than LDL-C or non-HDL-C, Dr. Sniderman said. And they said that apoB can be measured inexpensively—using widely available, automated, standardized methods—and more accurately, particularly at low concentrations, than LDL-C or non-HDL-C. That apoB can be measured inexpensively and accurately is also in AACC’s 2009 and 2013 lipoprotein guidelines, he noted. “But is it being measured in the U.S.? No. What is one of the primary arguments against it? That it can’t be measured more accurately. And that’s nonsense.”

The United States lipid community “led the world” on cholesterol, he said, with major achievements: linking cholesterol—in particular LDL-C—to the risk of atherosclerotic cardiovascular disease, developing algorithms to estimate the short-term risk of atherosclerotic CVD, demonstrating that the LDL receptor pathway is a genetic cause of familial hypercholesterolemia and a major determinant of the levels of LDL-C in plasma, developing statin therapies (Al Alberts of Merck), and creating scientific training opportunities for non-American scientists, among them Dr. Sniderman. “You did it. This is the most positive, wonderful country when you’re on your game,” he said.

However, neither the U.S. nor Canadian lipid communities have led the world in realizing the limitations of the LDL-receptor pathway as a determinant of the concentrations of LDL-C and apoB in plasma, he said (Sniderman AD, et al. J Am Coll Cardiol. 2022;79[10]:1023–1031). They have not accepted the now “overwhelming evidence” that apoB is a more accurate marker of atherosclerotic CVD risk and a more accurate index of the adequacy of lipid-lowering therapy than LDL-C or non-HDL-C, and that it can be measured more accurately, particularly at low concentrations.

As evidence, Dr. Sniderman cites many prospective observational studies that found apoB to be superior to LDL-C in assessing CVD risk, including the Québec Cardiovascular Study, Northwick Park Heart Study, Framingham Heart Study, Copenhagen City Heart Study, the Interheart and Interstroke studies, and others. Two prospective studies (Emerging Risk Factors Collaboration and Atherosclerosis Risk in Communities) found apoB equal to LDL-C in risk assessment. “But primarily they were based on whether it changed the c-statistic, and that’s irrelevant,” he said. They found “total cholesterol was just as good as anything, which is nonsense.”

Five Mendelian randomization studies found apoB superior to LDL-C in predicting coronary heart disease (Richardson TG, et al. PLoS Med. 2020;17[3]:e1003062; Zuber V, et al. Int J Epidemiol. 2021;50[3]:893–901; Yuan S, et al. Ann Neurol. 2020;88[6]:1229–1236; Levin MG, et al. Circulation. 2021;144[5]:353–364; Richardson TG, et al. Lancet Healthy Longev. 2021;2[6]:e317–e326). Seven prospective observational studies found apoB to be equal to non-HDL-C as a marker, and nine prospective observational epidemiological studies found it to be superior to non-HDL-C. “There are more that show apoB wins,” he noted. (For example: Steffen BT, et al. Arterioscler Thromb Vasc Biol. 2015;35[2]:448–454; Marston NA, et al. JAMA Cardiol. 2022;7[3]:250–256; O’Donnell MJ, et al. J Stroke. 2022;24[2]:224–235.)

In a 2019 paper, Dr. Sniderman and coauthors illustrated discordance in LDL-C and apoB when the apoB particles contain an average mass of cholesterol, when they are cholesterol-enriched, and when they are cholesterol-depleted (Sniderman AD, et al. JAMA Cardiol. 2019;4[12]:1287–1295). “Simply because the absolute hazard ratio for LDL-C/non-HDL-C are equal in a study does not mean they predict risk equally in an individual with cholesterol-loaded or cholesterol-depleted apoB particles,” he said.

To show what that means to an individual, Dr. Sniderman created the profile of a patient with cholesterol-depleted apoB particles who had an LDL-C of 111 mg/dL, non-HDL-C of 134 mg/dL, and apoB of 115 mg/dL. He assigned the patient a hazard ratio of 1.20 and two standard deviations. “This is small dense LDL,” he said.

“In any individual, when apoB particles are either cholesterol-loaded or -depleted, the risk predicted by LDL-C/non-HDL-C and apoB will not be equal,” he said. The predicted increase in the individual’s CVD risk was 44 percent based on the patient’s LDL-C and non-HDL-C, whereas the apoB-predicted increased risk was 73 percent, Dr. Sniderman said.

In patients who are discordant based on the amount of cholesterol in their apoB particles, the apoB and cholesterol markers will predict differently, even if the hazard ratio is the same, he said. “That’s the core idea in discordance analysis.” Conventional statistical methods were not designed to deal with highly correlated variables like LDL-C, non-HDL-C, and apoB. “And that’s been used as an argument for saying they’re equivalent,” Dr. Sniderman said. Discordance analysis was designed to directly contrast risk predictions by highly correlated variables.

Dr. Sniderman and coauthors studied a Framingham offspring cohort to determine the additional value of apoB beyond LDL-C or non-HDL-C as a predictor of coronary heart disease (Pencina MJ, et al. Eur J Prev Cardiol. 2015;22[10]:1321–1327). “We separated the population into tertiles,” he explained. One-third of the population had cholesterol-enriched particles, one-third had cholesterol-depleted particles, and the middle third had an average amount of cholesterol.

LDL-C and apoB are highly correlated variables, Dr. Sniderman said, “but there can be significant discordance,” meaning “that for a specific value of one, there’s a range of values for the other.”