In high-risk older patients with atrial fibrillation, rhythm control reduces cardiovascular death and stroke, but at a price

Référence

Kirchhof P, Camm AJ, Goette A, et al, for the EAST-AFNET 4 Trial Investigators. Early rhythm-control therapy in patients with atrial fibrillation. N Engl J Med 2020;383(14):1305-1316.

Sommaire

Previous studies that compared rate control and rhythm control had mixed results. In this study, 2789 older adults with an onset of AF within the past year were recruited. Participants had to be older than 75 years, have had a recent transient ischemic attack (TIA) or stroke, or have at least 2 of the following: older than 65 years, female sex, heart failure or left ventricular hypertrophy, hypertension, diabetes, chronic kidney disease, or severe coronary disease. At baseline, the mean age of participants was 70 years, 46% were women, 12% had a previous TIA or stroke, 12% had chronic kidney disease, 28% had heart failure, 88% had hypertension, and 44% had valvular heart disease. The groups were balanced at baseline and analysis was by modified intention to treat of all patients who had at least one follow-up assessment. Overall, this was a very-high-risk group of patients. The patients were randomized to receive rhythm control using medications or ablation, or rate control to manage symptoms. In the rhythm control goup, recurrent AF triggered additional attempts to cardiovert the patient. In the rhythm control group, after 2 years 19.4% had undergone ablation, 21% were taking flecainide, 17.7% were taking amiodarone or dronedarone, and 35% were taking no anti-arrhythmic. In the rate control group, after 2 years only 7% had undergone ablation and 5.7% were taking an anti-arrhythmic drug. Approximately 90% in both groups were taking anticoagulants after 2 years. The study was stopped early due to the detection of an efficacy signal after a median follow-up of 5.1 years. The primary outcome was a composite of 2 important things (stroke and cardiovascular death) and 2 much less important things (hospitalization for heart failure and hospitalization for acute coronary syndrome [ACS]). This composite was less likely in the rhythm control group (3.9 vs 5.0 per 100 person-years; NNT = 90 over 5 years to prevent one event). With regard to the important outcomes, the likelihoods of cardiovascular death (1.0 vs 1.3 per 100 person-years; NNT = 67 over 5 years) and stroke (0.6% vs 0.9% per 100 person-years; NNT = 67 over 5 years) were significantly lower in the rhythm control group, but the magnitude of those benefits was relatively small. Hospitalizations for heart failure and ACS were numerically less likely in the rate control group, but this difference was not statistically significant. The SF-12 mental score was significantly lower in the rhythm control group (-1.2 points; 95% CI -2.04 to -0.37), while patients in that group were significantly more likely to be in sinus rhythm (82.1% vs 60.5%; P < .05; NNT = 5). Patients in the rhythm control group had significantly more serious adverse events attributed to therapy, such as drug-induced bradycardia, toxicity due to the anti-arrhythmic drugs, pericardial tamponade, or major bleeding. There was no difference in the likelihood of being symptomatic and no difference in hospital days or in other health scores.