HYPERTENSIVE PATIENTS WITH GREATER GENETIC RISK RESPOND LESS EFFECTIVELY TO TREATMENT AND ARE MORE LIKELY TO BE TREATMENT RESISTANT

Abstract

Objective: Patients with resistant hypertension (RHTN) have increased cardiovascular disease risk. It is believed there may be genetic causes, although robust associations remain to be validated. Over 1,000 genetic loci have been discovered for BP from genome-wide association studies (GWAS), with application of genetic risk scores (GRS) enabling population risk stratification. However it is unknown whether genetic risk influences treatment response. We investigated whether BP associated loci: (i) are associated with RHTN; (ii) mediate the effect of antihypertensives on BP lowering response. Design and method: We analysed 6,266 patients of European ancestry with genetic data from ASCOT. BP-GRSs were constructed for systolic (SBP), diastolic BP (DBP) and pulse pressure (PP) using all available 1,167 pairwise-independent published BP-associated genetic variants, weighted by their effect estimates from our recent BP-GWAS in N∼750k individuals. We tested the BP-GRS for association with: (i) RHTN, adjusted for known non-genetic predictors: sex, age, BMI, diabetes, randomization treatment arm and left ventricular hypertrophy; (ii) BP response to monotherapy treatment, adjusted for sex, age, baseline-BP, dose and baseline antihypertensive use. All analyses adjusted for 10 principal components of ancestry. Results: Each BP-GRS is significantly associated with RHTN (P = 1.13 × 10-13, 1.96 × 10-5, 1.21 × 10-14 for SBP, DBP, PP) with increased genetic risk leading to increased odds of RHTN. Patients in the top 20% of SBP-GRS have 1.74 times greater odds of RHTN than the lowest 20% (P = 6.96 × 10-10). Additional adjustment for baseline-BP does not change conclusions. The association even remains significant (P = 5.11 × 10-4) with stronger effect (OR = 2.52) when restricted only to 1,037 patients untreated at baseline. The BP-GRS is also significantly associated with SBP and DBP response to amlodipine (N = 1,780): patients in the lowest 20% genetic risk group achieve better lowering response by 3.77mmHg mean difference between SBP pre and post treatment (P = 8.27 × 10-4). Conclusions: Our results confirm the hypothesis for a genetic contribution to RHTN, indicating RHTN patients are likely to be those with highest genetic risk of hypertension, which helps to explain the clinical challenge. We show that those with greater BP genetic risk respond less effectively to treatment and provide evidence that BP genes also influence treatment response.