Chirantan Banerjee, MD

Holliday EG, Traylor M, Malik R, Bevan S, Maguire J, Koblar SA, et al. Polygenic Overlap Between Kidney Function and Large Artery Atherosclerotic Stroke. Stroke. 2014

The human genome project was arguably one of the most important scientific advances of the last decade. It revolutionized our understanding of single nucleotide polymorphisms (SNPs or “snips”) which act as markers of disease genes, as well as help us study their inheritance. Genome wide association studies (GWAS) use DNA microarrays to compare SNPs across thousands of people with and without disease, and have the ability to identify novel associations.




As stroke pathophysiology is heterogeneous, studies have reported the need to study heritability of stroke subtypes separately. The International Stroke Genetics Consortium (ISGC) was formed to study the genetic basis of stroke, and has identified genes associated with large vessel stroke and cardioembolic stroke.
Kidney dysfunction has been associated with ischemic stroke in several epidemiologic cohorts. In the current study by Holliday et al, the ISGC authors investigate whether this association is secondary to underlying shared genetic polymorphisms. One of the major issues with GWAS is that the single SNPs they identify only explain a small portion of the heritability, as individual risk variants have effects too small to be significant statistically. 

The authors of this study try to clarify this issue in the study by aggregating numerous SNPs that pass apriori p-value thresholds into polygenic scores via GWAS. They derived polygenic scores for 3 renal traits: eGFR using serum creatinine, eGFR using cystatin C, and urine albumin to creatinine ratio. The performances of the polygenic scores for these traits were then tested in stroke GWAS datasets. One of the major strengths of the analysis is that they classified stroke cases into three subtypes, large vessel atherosclerosis, cardioembolic, and small vessel disease and tested them separately. 

Polygenic scores associated with higher eGFRcrea were associated with decreased risk of large vessel atherosclerotic stroke, and polygenic scores for higher urine albumin to creatinine ratio were also associated with higher risk of large vessel atherosclerotic stroke.  Subclassifying into stroke subtypes led to decreased power in the study, and the authors believe that the effect estimates are markedly underestimated. 

There are some other interesting hypothesis-generating findings. One such finding is that although CKD is associated with atrial fibrillation epidemiologically, kidney trait polygenic scores in the study were not associated with cardioembolic stroke. Also, there was a trend for association of polygenic score for higher urine albumin to creatinine ratio with small vessel strokes, which may be attributed to underlying heritable small vessel arteriopathy.

These findings are important, as they demonstrate a way to test polygenic genetic variants, which may arguably be more important in a complex heterogeneous entity like ischemic stroke as opposed to individual SNPs. Also, they provide some insight into underlying shared genetic basis of renal dysfunction and large vessel stroke.

As more stroke GWAS datasets become available, these findings will have to be replicated. But stroke genetics will surely be a fascinating area of research over the next several years, as it slowly unravels the underlying genetic underpinnings of cerebrovascular disease, as well as atherosclerosis at large.