The recent discovery of an association between an Lp(a) gene variant and aortic stenosis once again brings genomics into the limelight. How do we act on findings in genomics? What are the implications of this study?
Dr Topol has no conflicts of interest relevant to this digital technology.
Thanassoulis G, Campbell CY, Owens DS, et al. Genetic associations with valvular calcification and aortic stenosis. N Engl J Med 2013; 368:503-512. Abstract.
Dr Eric Topol: It's time for genomics again. I know this is not easy—it's not a subject that generates a lot of comments and interaction—but there was a very important finding in the February 7, 2013 New England Journal of Medicine about calcific aortic stenosis.
This is a genome scan paper looking at a million or so single nucleotide variants across the genome, not genome sequencing, and finding our friend Lp(a) as the minor allele in about 7% of people strongly associated with an odds ratio of 2.0, the same variants that are known to be associated with coronary artery disease with calcific aortic stenosis.
In well over 6000 individuals, where they had computed tomography of calcium and then replicated across many different cohorts—and not just for calcium but for the need for aortic-valve replacement (for example, in the Swedish cohort and along with other cohorts)—there was unquestionable genomewide significance with multiple points of replication.
They also looked at mitroannular calcification. There was an interferon gene variant that was not consistently replicated, so we have to hold back on making that pronouncement.
But certainly for aortic stenosis, Lp(a) held up an odds ratio over 2, and even adjusted for Lp(a) levels, this Mendelian randomization, basically the genomics play, clearly was influential with a 64% risk of aortic stenosis.
How do we translate these findings into useful actionable clinical information? We don't have a good way to lower Lp(a). And furthermore, we don't know if that's going to translate to preventing aortic stenosis. But we do know that these variants in Lp(a) that have a significant effect on a kringle (and this dysfunctional Lp(a) protein) clearly are incriminated, and we now have the substrate for testing lifelong therapy to suppress Lp(a) to see whether or not we can reduce the toll of aortic stenosis.
We need better agents and we need proof that this strategy would work. That is, by modulating this genetic-burden liability, can it change the natural history to aortic stenosis? Until now, we only knew genomics that were related to bicuspid aortic valve in the notch gene, but we didn't know about run-of-the-mill calcific aortic stenosis. This is a big finding, not yet actionable, but certainly worthy of treatments in the future, developed specifically for these individuals.
I think this is a very important common variant. As far as cardiovascular disease, atrial fibrillation, PITX2 and this one are the ones with the highest penetrance—odd ratios in the case of PITX2was 1.7, the odds ratio here of 2. That is very unusual for common genomic variants, which usually hover around 1.1, 1.2, but not at 2.0 (which is really striking!).
I hope you will cue into this and share your comments regarding Lp(a) showing up incriminated and explaining (at least in part) significant advanced age-related calcific aortic stenosis.
Thanks for your attention.