Study Points to Causal Role for Lp(a) in Atrial Fibrillation

Patrice Wendling

April 20, 2022

Although lipoprotein(a) is causally related to coronary artery disease and aortic valve stenosis — two known risk factors for atrial fibrillation (AF) — evidence linking Lp(a) to a causal role in the development of AF has been lukewarm at best.

A recent Mendelian randomization study showed only a nominally significant effect of Lp(a) on AF, whereas an ARIC substudy showed high levels of Lp(a) to be associated with elevated ischemic stroke risk but not incident AF.

A new study that adds the heft of Mendelian randomization to large observational and genetic analyses, however, implicates Lp(a) as a potential causal mediator of AF, independent of its known effects on atherosclerotic cardiovascular disease (ASCVD).

"Why this is exciting is because it shows that Lp(a) has effects beyond the arteries and beyond the aortic valve, and that provides two things," senior author Guillaume Paré, MD, MSc, Population Health Research Institute, Hamilton, Ontario, Canada, told | Medscape Cardiology.

"First, it provides a potential means to decrease the risk because there are all these Lp(a) inhibitors in development," he said. "But I think the other thing is that it just points to a new pathway that leads to atrial fibrillation development that could potentially be targeted with other drugs when it's better understood. We don't pretend that we understand the biology there, but it opens this possibility."

The results were published in the Journal of the American College of Cardiology.

Using data from 435,579 participants in the UK Biobank, the researchers identified 20,432 cases of incident AF over a median of 11 years of follow-up. They also constructed a genetic risk score for Lp(a) using genetic variants within 500 kb of the LPA gene.

After common AF risk factors were controlled for, results showed a 3% increased risk for incident AF per 50 nmol/L increase in Lp(a) at enrollment (hazard ratio [HR], 1.03; 95% CI, 1.02 - 1.05).

A Mendelian randomization analysis showed a similar association between genetically predicted Lp(a) and AF (odds ratio [OR], 1.03; 95% CI, 1.02 - 1.05).

To replicate the results, the investigators performed separate Mendelian randomization analyses using publicly available genome-wide association study (GWAS) statistics from the largest GWAS of AF involving more than 1 million participants and from the FinnGen cohort involving more than 114,000 Finnish residents.

The analyses showed a 3% increase in risk for AF in the genome-wide study (OR, 1.03; 95% CI, 1.02 - 1.05) and an 8% increase in risk in the Finnish study (OR, 1.08; 95% CI, 1.04 - 1.12) per 50 nmol/L increase in Lp(a).

There was no evidence that the effect of observed or genetically predicted Lp(a) was modified by prevalent ischemic heart disease or aortic stenosis.

Further, MR analyses revealed no risk effect of low-density-lipoprotein cholesterol or triglycerides on AF.

Notably, only 39% of Lp(a) was mediated through ASCVD, suggesting that Lp(a) partly influences AF independent of its known effect on ASCVD.

"To me, the eureka moment is when we repeated the same analysis for LDL cholesterol and it had absolutely no association with AF," Paré said. "Because up to that point, there was always this lingering doubt that, well, it's because of coronary artery disease, and that's logical. But the signal is completely flat with LDL, and we see this strong signal with Lp(a)."

Another "Red Flag"

Erin D. Michos, MD, MHS, senior author of the ARIC substudy and associate director of preventive cardiology at Johns Hopkins School of Medicine, Baltimore, said the findings are "another red flag that lipoprotein(a) is a marker we need to pay attention to and potentially needs treatment."

"The fact that it was Mendelian randomization does suggest that there's a causal role," she said. "I think the relationship is relatively modest compared to its known risk for stroke, ASCVD, coronary disease, and aortic stenosis…, which may be why we didn't see it in the ARIC cohort with 12,000 participants. You needed to have a million participants and 60,000 cases to see an effect here."

Michos said she hopes the findings encourage increased testing, particularly with multiple potential treatments currently in the pipeline. She pointed out that the researchers estimated that the experimental antisense agent pelacarsen, which lowers Lp(a) by about 80%, would translate into about an 8% reduction in AF risk, or "the same effect as 2 kg of weight loss or a 5 mm Hg reduction in blood pressure, which we do think are meaningful."

Adding to this point in an accompanying editorial, Daniel Seung Kim, MD, PhD, and Abha Khandelwal, MD, MS, Stanford University School of Medicine, California, say that "moreover, reduction of Lp(a) levels would have multifactorial effects on CAD, cerebrovascular/peripheral artery disease, and AS risk.

"Therefore, approaches to reduce Lp(a) should be prioritized to further reduce the morbidity and mortality of a rapidly aging population," they write.

The editorialists also join the researchers in calling for inclusion of AF as a secondary outcome in ongoing Lp(a) trials, in addition to cerebrovascular disease and peripheral vascular disease.

Unanswered Questions

As to what's driving the risk effect of Lp(a), first author Pedrum Mohammadi-Shemirani, PhD, also from the Population Health Research Institute, explained that in aortic stenosis, "mechanical stress increases endothelial permeability, allowing Lp(a) to infiltrate valvular tissue and induce gene expression that results in microcalcifications and cell death.

"So, in theory, a similar sort of mechanism could be at play in atrial tissue that may lead to damage and the electrical remodeling that causes atrial fibrillation," he told | Medscape Cardiology.

Mohammadi-Shemirani also noted that Lp(a) has proinflammatory properties, but added that any potential mechanisms are "speculative and require further research to disentangle."

Paré and colleagues say follow-up studies are also warranted, noting that generalizability of the results may be limited because AF cases were found using electronic health records in the population-scale cohorts and because few UK Biobank participants were of non-European ancestry and Lp(a) levels vary among ethnic groups.

Another limitation is that the number of kringle IV type 2 domain repeats within the LPA gene, the largest contributor to genetic variation in Lp(a), could not be directly measured. Still, 71.4% of the variation in Lp(a) was explained using the genetic risk score alone, they say.

Paré holds the Canada Research Chair in Genetic and Molecular Epidemiology and Cisco Systems Professorship in Integrated Health Biosystems. Mohammadi-Shemirani is supported by the Frederick Banting and Charles Best Canada Graduate Scholarship from the Canadian Institute of Health Research. Michos reports consulting for Novartis and serving on advisory boards for Novartis, AstraZeneca, Bayer, and Boehringer Ingelheim. Kim reports grant support from the National Institutes of Health and the American Heart Association. Khandelwal serves on the advisory board of Amgen and has received funding from Novartis CTQJ and Akcea.

J Am Coll Cardiol. 2022;79:1579-1590, 1591-1593. Full text, Editorial

Follow Patrice Wendling on Twitter: @pwendl For more from | Medscape Cardiology, follow us on Twitter and Facebook.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.