Is apolipoprotein B the best measure of CVD risk?

For the bottom tertile of participants, the average apoB was 86.7 mg/dL, and in the top tertile of participants, the apoB was 116 mg/dL, he said, and those who have more apoB have more events. LDL-C levels were nearly the same: 134 mg/dL for the bottom tertile of participants and 136 mg/dL for the top tertile. So the laboratory reporting the result has “reported a biologically false result,” he said, “because they’re not the same in terms of their outcome.”

The same survival result was found for the tertiles of discordant apoB versus non-HDL-C. Average apoB in one tertile was 91.1 mg/dL; non-HDL-C was 159 mg/dL. In the other tertile: apoB, 111 mg/dL; non-HDL-C, 160 mg/dL. People with a high apoB “are in trouble” because they won’t be recognized and treated. “The doctors will be making decisions about giving a drug with inadequate information.”

Dr. Sniderman shared a list of 16 discordance studies, “all of which favor apoB” over LDL-C/non-HDL-C (for example: Kim C-W, et al. Circ J. 2021;85[6]:900–907; Johannesen CDL, et al. J Am Coll Cardiol. 2021;77[11]:1439–1450; Razavi AC, et al. Am J Prev Cardiol. 2021;7:100190). He’s reviewing two additional such studies now. “That’s 18 in a row. That counts as done,” he said.

Even clinical trials favored apoB as the primary CVD marker. “We did a meta-analysis back in 2014, and the bottom line was apoB won,” Dr. Sniderman said, citing one such study (Thanassoulis G, et al. J Am Heart Assoc. 2014;3[2]:e000759).

In a Mendelian randomization analysis, Ference, et al., found in some cholesteryl ester transfer protein statin inhibitors, “if you follow the cholesterol, when you used combined therapy, it was wrong,” Dr. Sniderman said. “It didn’t show you the result, whereas the apoB was always true” (Ference BA, et al. JAMA. 2017;318[10]:947–956). Ference, et al., concluded: “The clinical benefit of lowering LDL-C levels may . . . depend on the corresponding reduction in apoB-containing lipoprotein particles.” They showed, Dr. Sniderman said, that “every lipid-lowering therapy that has worked tracks apoB” (Ference BA, et al. JAMA. 2019;321[4]:364–373).

A Copenhagen-based discordance analysis published last year sought to determine if elevated apoB and/or non-HDL-C are superior to elevated LDL-C in identifying statin-treated patients at residual risk of all-cause mortality and myocardial infarction. The authors found “if the LDL cholesterol is high but the apoB is low, the outcome is good,” Dr. Sniderman said, “whereas if the apoB is high, and the LDL-C is low, the outcome is bad” (Johannesen CDL, et al. J Am Coll Cardiol. 2021;77[11]:1439–1450). Citing the study’s data, he added: “A small number of cholesterol-enriched particles is not bad for you. A larger number of cholesterol-depleted particles is. It’s the number of atherogenic particles” that is key, he said.

A study published this year used the population-based UK Biobank and two international clinical trials—Fourier (a PCSK9 inhibitor plus statin) and Improve-It (ezetimibe plus statin)—to look at whether common measures of cholesterol concentration, TG concentration, or their ratio were associated with CVD risk beyond the number of apoB-containing lipoproteins (Marston NA, et al. JAMA Cardiol. 2022;7[3]:250–256). “But instead of looking for the c-statistic, they did a head-to-head” comparison, Dr. Sniderman said. In the unadjusted analysis performed for the clinical factors, he said, non-HDL-C and apoB were equal. “When they’re adjusted for each other, non-HDL becomes nonsignificant and apoB wins.”­­

Marston, et al., concluded: “In this cohort study, risk of MI was best captured by the number of apoB-containing lipoproteins, independent from lipid content (cholesterol or TG) or type of lipoprotein (LDL or TG-rich). This suggests that apoB may be the primary driver of atherosclerosis and that lowering the concentration of all apoB-containing lipoproteins should be the focus of therapeutic strategies.”

In another study published this year (Odyssey Outcomes), on apoB, residual risk after acute coronary syndrome, and the effects of alirocumab (Praluent), “apoB beats cholesterol and non-HDL cholesterol,” Dr. Sniderman said. In this statin-PCSK9 trial, as in the other, he said, “at baseline and beyond treatment [four months], apoB predicts risk linearly, down to the lower level of detection,” where it’s difficult to measure LDL-C accurately but “you can accurately measure apoB.”

Statins reduce cholesterol more than they do apoB, he noted. ApoB is reduced about 75 percent of the amount that LDL cholesterol is, and non-HDL cholesterol is also reduced more than apoB. “This means that when you’re looking at the LDL cholesterol or non-HDL cholesterol, even if you measured it accurately, it’s wrong,” Dr. Sniderman said. “It either overestimates or underestimates,” especially in the case of LDL-C. “You’re not even getting the LDL cholesterol you would get if you were accurately measuring the number of particles.”

The World Health Organization (1994), AACC (2009, 2013), European Society of Cardiology/European Atherosclerosis Society (2019), and European Federation of Clinical Chemistry and Laboratory Medicine/EAS (2020) have all said apoB can be measured rapidly, inexpensively, and more accurately, particularly at low concentrations, than LDL-C and non-HDL-C, and with standardized methods, using automated, widely available assay systems, Dr. Sniderman said. The same groups said the measurements of triglycerides and HDL-C are not standardized and there is inaccuracy and imprecision in their measurements, he said. While there is a consensus approach to harmonizing the results (“and I salute the people who did that,” he said), “no matter how you change the calculation of LDL cholesterol, based on triglycerides and HDL cholesterol, you can’t improve on measurements that weren’t standardized,” he said. “Moreover, in the age of low levels of cholesterol, the error in HDL cholesterol becomes determinative.”

In a study published in 2020 on the clinical utility of apoB versus LDL-C/non-HDL-C, Dr. Sniderman and coauthors wrote, “Adding apoB to clinical care would increase cost trivially” (Kohli-Lynch CN, et al. Clin Chim Acta. 2020;508:103–108). In their analysis, the cost of a lipid panel was $13.39 and a lipid panel plus apoB, $34.48 (see table). When other screening and statin costs are figured in, the difference in total costs is “trivial,” he said. “This is doing an apoB every single time you do a lipid panel. And I support doing a lipid panel as part of the diagnostic approach to the patient at the beginning, but if you’re following a patient on statin therapy, all you need is apoB.”

Lipoprotein (a), or Lp(a), is a special category of particle and a “particularly poisonous” particle, Dr. Sniderman notes. He uses Lp(a) level as a factor in determining whether to treat apoB because, he says, “if you lower apoB, you will lower the patient’s total risk. And we have found that when the apoB is low, the Lp(a) is not a major risk factor.

“More work needs to be done there, but certainly at the moment,” he continues, “the best therapy for high Lp(a) is a low apoB.”

Whether insurers cover apoB testing depends on what guidelines say, Dr. Sniderman says. “If the guidelines recommended apoB, it would be covered. In fact, a lot of the major insurers already cover apoB, and they do so based on the European guidelines. It’s the American guidelines that are lacking.”

There is no lack of evidence that apoB is better, he insists. “The metabolic studies of the apoB lipoproteins are clear. The pathophysiology of atherosclerosis is clear. The cross-sectional studies that were done [decades ago] were the first evidence, and they turned out to be right.” The prospective observational studies are numerous, and the discordance analyses are definitive, he said. “And now with the Mendelian randomization trials, I think we’re there.

“It is fair to ask for overwhelming proof. But I submit to you there is now overwhelming proof.”

Amy Carpenter Aquino is CAP TODAY senior editor.