Genetic Risk Score for Endothelial Cell Dysfunction May Help Hone ASCVD Care

A polygenic risk score (PRS) based on genes associated with endothelial cell function identifies patients at a higher risk of coronary artery disease, according to a new study.

The risk score, which is based on 35 single nucleotide polymorphisms (SNPs), also pinpointed patients more likely to benefit from therapies that lower LDL cholesterol.

“This is what’s been termed an ‘actionable genotype,’” said lead investigator Nicholas Marston, MD (Brigham and Women’s Hospital, Boston, MA). “If you identify this genotype in the patient population, you're not only identifying greater risk, but greater benefit from therapy.”

A few years back the American Heart Association issued a scientific statement on the use of PRS in clinical practice, stating they can be used if they help enable earlier and potentially tailored prevention, said Marston.  

“We’re seeing polygenic risk scores, more generally speaking, starting to be used in the clinic,” said Marston. “They can be used in the right setting if using them will impact management, if it will help you make that clinical decision about starting a therapy, and then most importantly, if it will have downstream benefits to the patient. I think this is a nice example of that.”

The mantra that “lower is better” when it comes to LDL-cholesterol levels is widely accepted across the cardiovascular community, and there’s been an increasing recognition that getting to low levels earlier in life has benefits, too. However, there are patients who have higher LDL levels throughout their lifetime yet don’t develop heart disease or coronary atherosclerosis, said Marston. Working with senior investigator Rajat Gupta, MD (Brigham and Women’s Hospital), Marston said they began to suspect that the protection afforded some patients might be related to endothelial cell biology.

“Basically, for cholesterol in the bloodstream to get into the vessel wall and cause plaque, it needs to get through the endothelial cell lining,” he said. “If you have endothelial cell dysfunction, it may be easier for cholesterol to traverse the intima.”     

SNPs for Endothelial Dysfunction

For the study, which was published this week in Nature Medicine, the researchers first took genome-wide SNPs previously linked to coronary artery disease and categorized the variants by biological pathway. From here, they identified 35 SNPs with a documented role in endothelial cell function. There were a number of criteria for identifying endothelial cell variants, including a known relationship between the lead SNP in the locus and endothelial cell function or the causal gene being part of the nitric oxide signaling pathway, among others.

The researchers constructed a polygenic risk score based on these endothelial cell-specific variants (EC-PRS) and tested its association with the risk of incident cardiovascular disease in three cohorts: UK Biobank (UKBB), a primary-prevention cohort of 348,967 patients; JUPITER, a primary-prevention study testing rosuvastatin in 8,749 patients; and FOURIER, a secondary-prevention study testing a PCSK9 inhibitor in 14,298 patients.

The mean age of patients ranged from 56.4 years in UKBB to 66.1 years in JUPITER, while the percentage of female participants ranged from 23.7% in JUPITER to 57.6% in UKBB. LDL-cholesterol levels at baseline ranged from 92 mg/dL in FOURIER to 142 mg/dL in UKBB.

In the UKBB, the EC-PRS was a significant predictor of coronary artery disease, with each standard deviation associated with a 24% higher risk of incident disease. The EC-PRS was only weakly correlated with a conventional PRS for LDL-cholesterol levels. Additionally, those with a high EC-PRS had a 44% greater risk of coronary artery disease than those with a moderate or low EC-PRS (adjusted HR 1.44; 95% CI 1.37-1.51). The risk associated with the EC-PRS remained after adjusting for clinical risk factors and the degree of risk was on par with that seen with systolic blood pressure, inflammation, and glycemic control.

“We were able to identify an axis of risk—endothelial cell dysfunction—that’s not really captured in our current clinical evaluation,” said Marston. “There’s no blood test for this so a genetic score for endothelial dysfunction can act as surrogate for identifying people.”

The risk associated with endothelial cell dysfunction was modified by the concentration of LDL cholesterol. In those with LDL levels of 150 mg/dL, for example, the EC-PRS was strongly correlated with the risk of incident coronary disease. On the other hand, the EC-PRS was not linked with risk of coronary disease when LDL levels were 50 mg/dL.  

In patients with moderate or high LDL-cholesterol levels, explained Marston, there is enough cholesterol in the bloodstream able to make its way into the endothelium if there’s endothelial dysfunction. Endothelial cell function does not appear to matter as much if LDL levels are low, because there’s little in the bloodstream to cross into the endothelium.

Variability in LDL-Lowering Benefits

In both JUPITER and FOURIER, the clinical benefit of LDL lowering with statin therapy or evolocumab (Repatha; Amgen) was greatest in those with a high EC-PRS. In JUPITER, there was a significant interaction between statin therapy and endothelial cell function where those in the highest EC-PRS quintile had a 72% lower risk of incident coronary artery disease events while those in the bottom 80% had just a 29% lower relative risk benefit (P = 0.033 for interaction).

Similarly, individuals with a high EC-PRS in FOURIER had a 33% lower relative risk of major vascular events compared with 9% and 4% lower risks in those with an intermediate or low EC-PRS, respectively (P = 0.03 for trend).

For doctors who might be on the fence about starting a statin or escalating treatment by adding a PCSK9 inhibitor, the EC-PRS can help them “use therapies with a bit more precision,” said Marston. Those with genetic evidence of dysfunctional endothelium appear to be more susceptible to high LDL-cholesterol levels, so these are patients where the focus should be in driving levels down low. It’s also possible that these patients should have lower LDL cholesterol targets, he suggested.

Broadly speaking, Marston said PRS, particularly those related to coronary artery disease, are slowing being incorporated into clinic, and there are numerous examples where the data have been shown to help patients and guide management.

“It’s exciting for those of us in preventive cardiology,” he said.