NEW YORK (GenomeWeb) – By integrating pharmacodynamic modeling into a GWAS analysis, researchers have identified and validated in independent cohorts five loci significantly associated with asthma patients' responses to inhaled glucocorticoid medication.
If the results, published last week in the Pharmacogenomics Journal, can be validated further, they could potentially inform more personalized strategies for asthma treatment, to tailor more optimal doses for patients based on measuring these genetic alterations to predict response.
Rongling Wu, the study's senior author, and a professor at Penn State College of Medicine, told GenomeWeb that the asthma study is just the first application of the team's pharmacodynamic modeling GWAS strategy, and that this tool could also be useful for detecting drug response variants in other disease contexts.
"We combine not only genetics and the genotyping results, but also pharmacodynamic principles into the model, which means that not only are the results biologically more relevant, but also we can increase the statistical power," Wu said.
In the asthma study, Wu and other Penn State colleagues, along with researchers from several other institutions genotyped an initial set of 120 patients from the dose of inhaled corticosteroids with equisystemic effects (DICE) trial, for 909,622 SNPs. In their analysis, the team then increased the power of their GWAS by incorporating mathematical models of drug response at different doses.
Unlike GWAS of common complex diseases, studies of drug response are often limited by relatively small numbers of samples, which can make it difficult to identify trait-associated loci at genome-wide significance.
The Penn State group's method involves modeling the relationship between dose and drug response through mathematical equations based on repeated measures of response at multiple doses in each subject. This allows for many more points of data to be considered in the GWAS analysis, increasing the likelihood of identifying a drug response-associated variant.
In the asthma study, the team's pharmacodynamic modeling GWAS approach identified five loci that were associated with response to inhaled glucocorticoids, rs6924808 on chromosome 6, rs10481450 on chromosome 8, rs1353649 on chromosome 11, rs12438740 on chormosme 15, and rs2230155 on chromosome 15.
All five of the identified SNPs showed a recessive effect, in which homozygotes for the mutant alleles showed increased response to treatment over both heterozygous and homozygous wild-type subjects.
The researchers then set out to replicate and validate the results in three independent cohorts, from the IMPACT, SOCS, and SLIC trials, representing several hundred additional patients.
All five SNPs remained significant in at least one of the three trial groups. One, chr15 rs2230155 was non-significant in SOCS, while two others, chr8 rs10481450 and chr11 rs1353649, were only marginally significant in IMPACT and SOCS respectively.
Analyzing all the cohorts pooled together, the team found that all but ch15 rs2230155 showed significant association with glucocorticoid response overall. In general, mutant alleles at the SNPs tended to increase pulmonary function of the asthma patients studied by between 30 and 300 percent, the authors wrote.
According to Wu and his coauthors, the group's findings could potentially help tailor and optimize glucocorticoid dosages for individual asthma patients, potentially coupled with previously observed genetic and epigenetic associations with asthma drug response, or with lung function in asthma in general.
For example, several years ago researchers from Brigham and Women's Hospital identified a single locus in the glucocorticoid-induced transcript 1 gene that was associated with poor response to these asthma therapies.
A new study funded by the National Heart, Lung, and Blood Institute early last year is also investigating how genetics influence response to inhaled corticosteroids specifically in African-Americans with asthma.
For the Penn State team's results to be clinically applicable, either alone or coupled with additional markers of drug response, the findings will have to be further validated and confirmed in additional cohorts.