LONG TERM BENEFITS OF BLOOD PRESSURE TREATMENT ON THE INCIDENCE OF ATRIAL FIBRILLATION, HEART FAILURE AND CARDIOVASCULAR MORBIDITY AND MORTALITY: 20-YEARS FOLLOW-UP OF ASCOT-LEGACY

Abstract

Objective: Previously, we reported 16-year cardiovascular (CV) mortality associated with either atenolol- or amlodipine-based treatment regimen; but without data on non-fatal CV events. We now report longer morbidity and mortality including the impact of BP-treatment on the incidence of atrial fibrillation and heart failure Design and method: In the ASCOT-Legacy Study, 8580 hypertensive patients (4275 assigned to atenolol+/-diuretic-based and 4305 to amlodipine+/-perindopril-based treatment) were followed in the UK for the maximum duration of 21 years (IQR: 9.1–19.3 years). All fatal/non-fatal CV events during the trial period and the post-trial mortality events were independently adjudicated. Post-trial morbidity events were evaluated using electronic health records. Cox proportional hazards were estimated for the first occurrence of atrial fibrillation, fatal/non-fatal HF, non-fatal/fatal stroke (stroke), non-fatal /fatal-coronary heart disease (CHD), total coronary events and total CV events in two treatment arms. Analyses were adjusted for a-priori confounders (See table). Interaction, if any, with associated statin therapy was evaluated. We also did a sensitivity analysis using only post-trial data. Results: During the in-trial period of 5.5 years, the cumulative mean SBP was marginally higher for those on atenolol-based treatment compared to those on amlodipine-based treatment (138.0 [SD,10.8] and 136.3 [9.9] mm Hg, respectively). Table 1 shows the crude and adjusted hazard ratios (HRs) associated with the two treatment regimens. Those on amlodipine-based (vs. atenolol-based) treatment had significantly reduced risk of atrial fibrillation [0.91, 95% CI, 0.83 to 0.99], total coronary events [092, 0.86 to 0.99], stroke [0.82, 0.72 to 0.93] and total CV events [0.93, 0.88 to 0.98]. There was no significant difference in the incidence of heart failure or CHD, although there was a nominal reduction. No evidence of interaction with statin therapy was noted. Post-trial significant differences were apparent for stroke, total coronary events and total CV events. Conclusions: Allocation to an amlodipine-based treatment has a long-term beneficial CV effect, particularly on stroke and total coronary and CV events. Reduction in the risk new-onset atrial fibrillation may be an important mediator of this legacy effect.