Antihypertensive pharmacogenetic effect of fibrinogen-beta variant-455G>A on cardiovascular disease, end-stage renal disease, and mortality: The GenHAT study

OBJECTIVE: The FGB gene codes for fibrinogen-β, a polypeptide of the coagulation factor fibrinogen, which is positively associated with cardiovascular diseases. Studies show that angiotensin-converting enzyme (ACE) inhibitors lower plasma fibrinogen concentrations, whereas diuretics and calcium-channel blockers do not. As carriers of the FGB-455 minor 'A' allele have higher levels of fibrinogen while ACE inhibitors lower it, we hypothesize that 'A' allele carriers benefit more from antihypertensive treatment with ACE inhibitors than calcium-channel blockers or diuretics, relative to 'GG' genotype individuals. METHODS: The Genetics of Hypertension Associated Treatment (GenHAT) study [ancillary to Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT)] genotyped hypertensive participants for several hypertension-related candidate genes, making this a post-hoc analysis of a randomized trial. In total, 90.1% of the ALLHAT population was successfully genotyped for FGB-455. We included participants (n=30076) randomized to one of three antihypertensive medications (lisinopril, amlodipine, chlorthalidone), with two treatment comparisons: lisinopril versus chlorthalidone and lisinopril versus amlodipine. The primary outcome of ALLHAT/GenHAT was coronary heart disease, defined as fatal coronary heart disease or non-fatal myocardial infarction, and secondary outcomes included stroke, heart failure, all-cause mortality, and end-stage renal disease (ESRD) with mean follow-up time of 4.9 years. Genotype-by-treatment interactions (pharmacogenetic effects) were tested with the Cox regression. RESULTS: Stroke: common 'GG' homozygotes had higher risk on lisinopril versus amlodipine [hazard ratio (HR)=1.38, P<0.001], whereas minor 'A' allele carriers had slightly lower risk (HR=0.96, P=0.76; P value for interaction=0.03). Mortality: 'GG' homozygotes had higher risk on lisinopril versus amlodipine (HR=1.12, P=0.02) or chlorthalidone (1.05, P=0.23), whereas 'A' allele carriers had slightly lower risk (HR=0.92, P=0.33 for lisinopril versus amlodipine; HR=0.88, P=0.08 for lisinopril versus chlorthalidone; P value for interactions 0.04 and 0.03, respectively). ESRD: 'GG' homozygotes had higher risk on lisinopril versus chlorthalidone (HR=1.27, P=0.08), whereas 'A' allele carriers had lower risk (HR=0.64, P=0.12; P value for interaction=0.03). CONCLUSION: There was evidence of pharmacogenetic effects of FGB-455 on stroke, ESRD, and mortality, suggesting that relative to those homozygous for the common allele, variant allele carriers of the FGB gene at position-455 have a better outcome if randomized to lisinopril than chlorthalidone (for mortality and ESRD) or amlodipine (for mortality and stroke). For the models in which a pharmacogenetic effect was observed, the outcome rates among 'GG' homozygotes were higher in those randomized to lisinopril versus amlodipine or chlorthalidone, whereas minor 'A' allele carriers had lower event rates when randomized to lisinopril versus the other medications.