New research presented at the 2024 Association for Research in Vision and Ophthalmology (ARVO) meeting earlier this month examined the association between dry eye disease (DED) and genetics.
Sounds pretty straightforward.
Alas; it’s anything but. In fact, numerous studies and research spanning several years have sought to determine if genetics play a role in DED, only recently finding the truth:
Yes, genetics do play a role.
However, the real question is: Exactly what role do genetics play in the development of DED?
So that’s what this study sought to determine?
Indeed. While the investigators of this research referenced twin studies that estimated “a substantial genetic heritability of DED, explaining 30-40% of the variances,” they also called out the difficulty associated with phenotyping DED.
This, they stated, “has limited the application of large-scale genome-wide association studies (GWAS).”
Which leads us to their research …
Yes! Investigators utilized the Veterans Affairs (VA) Million Veteran Program (MPV) biobank to collect data from over 650,000 patients with genetic data accessible via electronic medical records (EMR) system.
Their goal: Evaluate the genetic basis of DED.
How did they do this?
By developing a case-control algorithm—based on International Classification of Diseases (ICD)-9/10 codes and prescription records—that was validated by a manual chart review at three VA eye clinics, per the study authors.
The result: Based on the MVP database of patients, they identified 48,794 DED cases and 29,224 controls.
- Of that number, participants were categorized based on their ancestry and race/ethnicity.
And how were participants classified?
Investigators performed logistic mixed model GWAS analyses for each ancestry group (see below):
- European
- African
- Hispanic
- East Asian
These were adjusted for age, sex, and 10 (unidentified) ancestry principal components; they were then followed by a multi-ancestry meta-analysis assuming fixed effects.
Now the findings.
The chart review revealed positive and negative predictive values of 98% and 93%, respectively.
Interestingly, the following factors were found to be associated with DED (p<0.001):
- Female sex
- Age
- African ancestry
- Charlson Comorbidity Index
- A measure of the relative 1-year risk of mortality based on 17 potential comorbidities from the ICD diagnosis codes
Which comorbid conditions were noted?
Just two:
- Depression (odds ratio [OR] = 2.7 [2.6-2.8])
- Sleep apnea (OR = 2.3 [2.2-2.4])
Now let’s talk about loci
First: A locus (plural loci) is the actual location of a gene or other DNA sequence on the region of a chromosome.
And in these findings: A “novel common variant locus at 3p21.31 (rs9855433, near ZNF445) reached genome-wide significance (GWS).”
- Refresh: GWS is a specific threshold for determining the statistical significance of a reported association between a given trait (ie: DED) and a given single-nucleotide polymorphism.
So what does that mean?
A phenome-wide association scan found this locus variant was consistently associated with the following:
- Depression
- Sedative-hypnotic medication use
- Chronic pain conditions
Plus: They also noted a significance estimate of common variant heritability (read up on that concept here) that was estimated at 12% on a liability scale with a 25% assumed prevalence.
What does this data mean for future research?
The study authors stated that the development and validation of a DED phenotyping algorithm could also be used for other EMR-linked biobanks.
And their conclusions?
As clinical research continues to work toward the difficult process of analyzing DED phenotyping, these authors’ identification of the first DED-specific GWAS locus represents a significant step toward “expanded GWAS meta-analyses and the development of polygenic risk score models” in the future.
The 2024 ARVO meeting was held May 5-8, 2024, in Seattle, Washington.