Anticoagulation for atrial fibrillation: Time to revise the paradigm?

Atrial fibrillation (AF) is the most common and clinically important heart rhythm disorder, and the incidence and prevalence of AF increase progressively with age. In the United States, the prevalence of AF increases from less than 1% among adults under 60 years of age to more than 10% in those ≥80 years of age. As a result, over 50% of AF is concentrated in the 7% of the population 75 years of age or older, and it is projected that by 2050, the median age of patients with AF will be 80 years. AF is more common in men than in women at all ages, and it is also more common in whites than in other races and ethnicities. In addition to the rising prevalence of AF in older adults, age is a strong independent risk factor for ischemic stroke in patients with AF. In the CHA2DS2-VASc stroke risk schema, age 65–74 years receives 1 point, and age 75 or older receives 2 points. The risk of stroke in patients with AF is higher in women than in men, and female sex is also assigned 1 point. Current guidelines recommend systemic anticoagulation for almost all men with a CHA2DS2-VASc score of 2 or higher and almost all women with a CHA2DS2-VASc score of 3 or higher. 5 Thus, all men and women age 75 or older with AF merit consideration for anticoagulation, even in the absence of other stroke risk factors. The evidence to support anticoagulation for AF in older adults is compelling. In the Birmingham Atrial Fibrillation Treatment of the Aged study (BAFTA), 973 patients with AF 75 years of age or older (mean 81.5 years, 45% women) were randomized to warfarin with a target international normalized ratio (INR) of 2.0– 3.0 or to aspirin 75 mg daily. The primary endpoint was a fatal or disabling stroke, arterial embolism, or intracranial hemorrhage. After a median follow-up of 2.7 years, patients randomized to warfarin experienced a 52% reduction in the primary endpoint (relative risk 0.48, 95% CI 0.28–0.80, absolute annual risk reduction 2.0%, 95% CI 0.7%–3.2%, p = 0.003). The risk of extracranial hemorrhage did not differ between groups. In subgroup analysis, the benefits of warfarin were consistent in men and women and across age groups, including patients over 85 years of age. More recently, a meta-analysis of over 20,000 patients ≥75 years of age with AF demonstrated that direct-acting oral anticoagulants (DOACs) were associated with a 35% reduction in the risk of stroke or systemic embolization (HR 0.65, 95% CI 0.48–0.87) compared to standard therapy (principally warfarin). In this analysis, the risk of major bleeding varied by agent and dose, but all DOACs have been associated with a reduced risk of intracranial hemorrhage compared to warfarin. An important limitation of these trials is that older adults with multimorbidity, including cognitive impairment and frailty, were either peremptorily excluded from participation or markedly under-represented. Since older adults with multimorbidity, cognitive impairment or frailty are at increased risk for bleeding complications, the balance between benefit and harm of anticoagulation may be altered; as a consequence, the applicability of trial findings to these patients is uncertain. In this issue of the Journal, Wang and colleagues report the results of a study designed to examine the association between anticoagulation and clinical outcomes in older adults with AF and comorbid cognitive impairment or frailty. The authors analyzed data on 1244 patients ≥65 years of age enrolled in the Systematic Assessment of Geriatric Elements (SAGE) AF study from 2016 to 2018. Data were collected prospectively, and patients were followed until 2020 (median follow-up 2.0 years). The primary endpoint was a composite of death or major bleeding. Secondary outcomes included clinically relevant bleeding and the composite outcome of death, major bleeding, or stroke. The authors used a propensity score to adjust for 16 potential confounding covariates, including age, CHA2DS2-VASc score, HAS-BLED bleeding risk score, Charlson Comorbidity Index, falls in the past 6 months, renal function, dual antiplatelet therapy, and number of medications. Cox regression analysis was used to calculate propensityadjusted hazard ratios for the primary and secondary endpoints in patients receiving anticoagulation compared to those not treated with anticoagulants. The mean age of the population was 75.5 years, 49% were women, 42.4% had cognitive impairment (Montreal Cognitive Assessment score <23 out of 30), 13.8% were This editorial comments on the article by Wang et al. in this issue.