There is substantial inter-individual variability in blood pressure (BP) responses to anti-hypertensive drugs and it is hypothesised that there could be genetic influences. One goal of pharmacogenetics is to aid in treatment selection. However very few variants have been identified. After the success of genome-wide association studies (GWAS) for BP, it is hoped GWAS could also reveal genes associated with BP treatment response.
Design and method:
A discovery analysis was performed including participants from ASCOT and NORDIL studies. Six different analyses were performed for systolic BP (SBP), diastolic BP (DBP) and pulse pressure (PP) responses to monotherapy beta blockers (BBs) and calcium channel blockers (CCBs).
BP response was defined as post-treatment BP minus baseline BP. All analyses adjusted for age, sex, baseline BP and principal components. ASCOT analyses were also adjusted for dose and previous antihypertensive use, and excluded patients taking either BB/CCB drug at baseline, whereas NORDIL participants were previously untreated and subsequently treated on the same drug dose.
Genotyped data was imputed to 1000 Genomes, testing ∼7million common variants. Results from ASCOT and NORDIL were meta-analysed, with total N = 2,556 for BBs and N = 3,369 for CCBs.
Any variants reaching P < 1 × 10-5 for SBP and DBP response to BBs were tested for replication in available ICAPS consortium data (N = 1,328).
Strong correlation between baseline BP and BP response shows patients with higher BP have greater BP reduction post-treatment.
No variants reached genome-wide significance in any of the GWAS analyses, and none replicated in ICAPS data.
Lookups of ∼1,000 published BP-associated variants were performed, but none reached Bonferroni significance level. Ongoing analyses are testing BP genetic risk scores for association with treatment response.
Comparison of GWAS results shows correlation between SBP vs DBP response (r2 = 0.48) and SBP vs PP response (r2 = 0.76), but none between response to BBs vs CCBs.
Despite being the largest GWAS of antihypertensive response, we found no significant associations, showing that larger studies are still required to increase power for discoveries. Model comparisons show different genes are likely to influence response to different antihypertensive drugs, with different genetic architecture.