More Insights Into AAA Genetics Spur New Questions on Screening

 

(UPDATED) Data from the Million Veteran Program are helping to better define the “genetic architecture” behind abdominal aortic aneurysm (AAA), raising hopes that genetic tests might help pinpoint people who need screening.

The genetic roots of AAA are, as of yet, “incompletely defined,” they say. “In total, 10 previously identified risk loci explain a small fraction of AAA heritability.” This study, describing 14 newly identified loci, brings the total to 24. First author Derek Klarin, MD (Malcom Randall VA Medical Center, Gainesville, FL), and colleagues also created a genome-wide polygenic risk score (PRS) aimed at “targeting of asymptomatic individuals for AAA screening.”

Philip S. Tsao, PhD (VA Palo Alto Health Care System, CA), one of the paper’s two senior authors, said he was pleased by these discoveries. “The hope and hypothesis is that because these are common genetic variants that have a small effect individually on disease, the more individuals you survey the more of these you’ll uncover,” he told TCTMD. “And so it was a pleasant surprise that by testing this many people in such a new and large data set, we were able to replicate [what has] been done before, but also add to those findings.”

The results were published online yesterday in Circulation.

Up to 40% of people with AAA are diagnosed late in their disease course and late diagnosis imparts higher rupture risk at the time of repair. Use of PRS might lead to better screening for AAA, the investigators write. “Our data suggest that extending current screening guidelines to include testing to identify those with high polygenic AAA risk, once the cost of genotyping becomes comparable with that of screening ultrasound, would significantly increase the yield of current screening at reasonable cost.”

One might imagine a scenario in the not-too-distant future where polygenic disease prediction is incorporated into AAA screening algorithms (if it is shown to improve cost-effectiveness of screening), and where genetically-guided therapies are developed and used to prevent progression of AAA. Michael Holmes

The guidelines, of which there are many, vary slightly in their recommendations. In the United States, where this study was based, the US Preventive Services Task Force advises onetime ultrasound screening for AAA in men ages 65 to 75 years who have ever smoked. And the Society for Vascular Surgery, for example, extends that to men 55 years or older with a family history of AAA and women 65 years or older who have smoked or have a family history of AAA.

“In the UK, we screen all males aged 65 for AAA as it reduces mortality from ruptured AAA,” Seamus Harrison, MBBS, PhD (University of Cambridge, England), told TCTMD via email. “That said, some men rupture before the age of 65, the vast majority of men that are screened don’t have a AAA, and mass screening of females is not recommended, so there is scope for improvement and a precision medicine approach could be applicable.”

Michael Holmes, MBBS, PhD (University of Oxford, England), also commenting for TCTMD, predicted that there’s potential—down the road—for clinical use. “One might imagine a scenario in the not-too-distant future where polygenic disease prediction is incorporated into AAA screening algorithms (if it is shown to improve cost-effectiveness of screening), and where genetically-guided therapies are developed and used to prevent progression of AAA,” he wrote in an email.

There’s a lot of work left to do before this happens, however. As Harrison pointed out to TCTMD, “the clinical effectiveness of AAA screening is based upon more than just ‘case-finding’”—targeting people suspected to be at risk.

Blood Pressure and Lipids

The researchers leveraged data on 7,642 cases and 172,172 controls from the Million Veteran Program, run by the US Department of Veterans Affairs, testing around 18 million DNA sequence variants. Individuals with AAA tended to be older, be male, and have been prescribed statin therapy. All were US veterans of European ancestry. They then replicated these results in an independent data set that had been published in Circulation Research in 2017.

Interestingly, 19 of the 24 known AAA risk variants were also associated with aneurysm risk in other vascular territories, most strongly for iliac and lower-extremity aneurysms. Cerebral aneurysms appear to have different genetic origins.

“We also found that several of the DNA sequence variants associated with a range of known risk factors for AAA,” they note. For example, there were links to genes previously associated with LDL cholesterol (LDLR, PCSK9, SORT1, LPA, APOE) and triglycerides (APOA5, TRIB1).

Additionally, they used mendelian randomization to demonstrate a link between AAA risk and diastolic blood pressure (OR per 10-mm Hg increase 1.43; 95% CI 1.24-1.66) but not systolic pressure (OR 1.06; 95% CI 0.97-1.15), suggesting that the former “likely has a causal relationship with AAA development.”

Their PRS included 29 variants and was able to pinpoint individuals at risk of AAA (OR per 1-SD increase 1.26; 95% CI 1.18-1.36), independent of history and smoking risk factors. This score was validated in three additional datasets, including one with 718 AAA cases and 46,380 controls of African ancestry. Although the score was associated with AAA risk across the board, “the effect estimate was diminished most in individuals of African ancestry."

Among European people whose PRS placed them at highest risk (5%), the AAA prevalence was 5.9% overall and 8.6% in men ages 50 or older. By comparison, among people with African ancestry, these rates were 1.7% for the highest-risk group and 2.5% for older men. Most polygenic scores have been developed based on populations that hail from Europe, so the disparity seen here is a reminder that future studies must include diverse genetic backgrounds, the investigators urge.

Caveats and Next Steps

For Harrison, “efforts such as this [study] are extremely welcome given the importance of these phenotypes, particularly in aging populations.” He observed that this is especially true since research on AAA and peripheral vascular disease more broadly “often lags behind that in coronary diseases.”

Holmes agreed that genome-wide association studies can be revealing. ”Certainly, one of the most exciting aspects of conducting [an] analysis such as this,” he said, “is investigating individual genetic loci that are identified as harboring genetic variants that strongly associate with disease and exploring their potential as therapeutic targets.”

He pointed to a 2017 JAMA Cardiology paper, which he led along with Harrison, that found genetically-predicted LDL cholesterol and triglyceride levels go hand in hand with AAA risk. Their work went a step further, looking at whether certain drug targets—for example, an LDL-lowering allele of PCSK9—were themselves associated with lower AAA risk, making the case for treatment strategies.

As for the polygenic risk score, though, Holmes raised some questions.

“The main challenges for putting the polygenic test into clinical use is that although there is strong evidence of an association [between] the polygenic score with AAA, the magnitude of association is relatively modest—with odds ratios of about 1.26 per 1 standard deviation higher PRS,” he observed.

“What this suggests is that the detection rate (or sensitivity) would be rather low for a given false-positive rate,” Holmes explained. “The reason for this is that for common diseases such as AAA, most cases tend to occur in individuals at moderate risk of disease and therefore using a genetic risk score in this way (ie, by focusing on those in the highest 5% of the PRS, as done here) is likely to miss the majority of AAA cases in the population.”

With a specific cut point—here 5%—the score wouldn’t pick up on low- or moderate-risk individuals, what’s known as the “prevention paradox,” he said. “Adopting such an approach is likely to have only a modest effect at preventing morbidity and mortality arising from AAA in the general population.”

Tsao agreed that from a public health perspective, this is a valid point. “And the 5% is somewhat arbitrary,” he said. But comparing people at highest risk, whether that’s 5% or 10%, with those at lower risk allows for proof of concept. “Certainly, you could do the fine-tuning and really see where the score best performs not only in the general population but [also] specific populations, for example different race/ethnic backgrounds or even geographic areas,” Tsao added. “That was again a first attempt.”

The vast majority of this work has been done in individuals of European descent. So we just need to double down, triple down on trying to diversify... Philip S. Tsao

Next steps, said Holmes, could include cost-effectiveness modeling to gauge which subgroups should be offered screening, as well as exploration of therapeutic targets. For example, “given the blood pressure signals identified, the authors might go on to explore individual blood pressure-lowering therapies (through mendelian randomization) to explore whether individual classes of BP-lowering agents are likely to be beneficial or whether lowering BP in AAA is agnostic to drug class,” he suggested. It may be that there’s an optimal combination of drugs aimed at different targets, such as lipids, blood pressure, and inflammation.

“Prevention of AAA as a phenotype will be difficult until there is a more precise way of identifying those at risk,” Harrison cautioned. So far, “it has not been possible to demonstrate a causal relationship between modifiable factors and progression or rupture of AAA—future studies such as the UK Aneurysm Growth Study may begin to answer these important questions.”

It will also be key to establish large biobanks around the world to capture a wide swath of populations, said Tsao. “The vast majority of this work has been done in individuals of European descent. So we just need to double down, triple down on trying to diversify first the recruitment of [underrepresented groups] and express to them the tremendous value and benefit they would have to society as a whole if they were to contribute their samples, and then to really promote that type of work” in developing scores.