Junior commentary
In the ELDERCARE-AF trial, the investigators shed light on very elderly AF patients with multimorbidity - a growing patient group in the clinic, yet immensely understudied in previous clinical trials. When warfarin, in The Birmingham Atrial Fibrillation Treatment of the Aged Study (BAFTA), was concluded to be superior to aspirin in thromboembolic prevention and equally safe, the study participants were all elderly ≥75 y, however, they were outpatients implying a lower risk (1). In the pivotal DOAC trials, elderly patients ≥75 y constituted between 31-44% of the study populations with the corresponding numbers being only 4-17% for ≥80 y olds (2). It is widely known that elderly patients constitute a highly heterogenous patient group regarding clinical characteristics with increasing prevalence of comorbidities and organ dysfunction and a significant increase in cardiovascular and bleeding risk, with age (3). To this extent, the conduction of the ELDERCARE-AF trial, where focus lies on very elderly patients with high thromboembolic risk as well as particularly high bleeding risk, is of great importance. The study population of ELDERCARE-AF was indeed of high risk with mean values on age 86,6 y, creatinine clearance 36,3 milliliters (ml)/minute (min), CHA2DS2 VASc score 4,9, HAS-BLED score 2,3 and 35% with a history of falls within the past year and 40,9% considered frail. Unlike the DOAC trials, in which patients with significant organ dysfunction, e.g. severe renal and liver failure, were excluded, the investigators of ELDERCARE-AF contribute to fill a knowledge gap of thromboembolic prevention in elderly patients with dual risk of TE and bleeding, by focusing solely on this vulnerable patient group in their trial.
Subgroup analyses of elderly patients ≥75 y in the phase 3 DOAC trials, showed beneficial effects of DOACs compared to warfarin, results that were consistent with the efficacy results in the overall study populations (4). The studies of safety outcomes did, however, not yield as convincing results in the elderly, except for the incidence of intracranial hemorrhages (ICH) that were significantly lowered also in the higher age category, with DOAC. The dose of 15 mg of edoxaban was chosen based on the fact that it gives a similar plasma concentration of edoxaban in patients with a creatinine clearance between 15-30 ml/min as 30-60 mg of edoxaban yield when creatinine clearance is above 50 ml/min (5). Furthermore, clinically, dose reduction did not seem to alter the efficacy of edoxaban in elderly patients in the ENGAGE AF-TIMI 48 trial, however, it did reduce the incidence of major bleeding, a reduction that increased with age (6). Nevertheless, major bleeding rates were similar in the ELDERCARE-AF trial compared to rates of major bleeding in subanalyses of elderly in the ENGAGE AF-TIMI 48 trial but the incidence of bleedings in general (major or clinically relevant nonmajor bleeding) and particularly GI bleedings were, in the ELDERCARE-AF trial, overall very high and higher with edoxaban compared to placebo (17,7% vs. 10,7%; HR 1,65; 95% CI 1,20-2,27). Notably, also less severe bleedings and especially GI bleedings seem to be associated with a significant short-term mortality in elderly. In a Swedish population-based cohort study, a surprisingly high incidence of 90-day mortality was seen among patients ≥75 y having suffered a GI bleed (16,2%) compared to mortality after ischemic stroke (25,1%) and ICH (31,6%), irrespective of prevalence of antithrombotic treatment (7). Therefore, also secondary safety outcomes and especially GI bleedings, are important to consider in clinical risk evaluation for OAC treatment in the elderly.
Clinical guidelines of today recommend DOAC for prevention of TE in all patients with AF including elderly (8). Although conducting trials on very elderly patients is complicated, with higher rates of discontinuation (nearly one third in the ELDERCARE-AF trial), the clinical impact, with results generalizable to a wider elderly patient group, should be considered important enough to justify the effort. Hence, the ELDERCARE-AF trial will hopefully lead to the conduction of more trials focusing on the patients who challenge us the most in the clinic. Even though treatment with low-dose edoxaban did not reduce mortality compared to placebo in this trial, the significant reduction in thromboembolic events with edoxaban is noteworthy. Considering the impact that occurrence of stroke has on quality of life, it is clinically important to have the opportunity to offer this frail patient group appropriate stroke prevention.
Finally, the ELDERCARE-AF investigated a Japanese population with certain clinical features attributable to East Asian populations, e.g. low body weight, and, from the ENGAGE AF-TIMI 48 trial, high incidence of both thromboembolic and bleeding events compared to non-East Asian populations (9). It is of clinical interest to see whether low-dose edoxaban gives similar efficacy and perhaps improved safety in western populations with a higher body weight.
Senior commentary
Elderly patients with bleeding risk factors are often left non-anticoagulated and treated with aspirin or no antithrombotic therapy. This is common in many Asian countries, where prescribers fear bleeding more than the potential beneficial reductions in stroke and mortality (10, 11). The ELDERCARE-AF trial was designed to answer the question whether treatment with an oral anticoagulant (even a low dose DOAC; in this case, edoxaban) would be better than placebo in these clinically fragile patients, ie. let’s treat with something rather than nothing. The trial shows a reduction in stroke but no significant difference in major bleeding.
Stroke and bleeding risks are not static but ‘dynamic’, changing with ageing and incident comorbidities (12). Clinicians also tend to forget that appropriate use of bleeding risk in clinical practice allows mitigation of modifiable bleeding risk factors and the scheduling of high bleeding risk patients for early review and follow-up (8). This was clearly evident in the prospective mAFA-II trial, where appropriate and responsible use of the HAS-BLED score in the intervention arm was associated with less major bleeding and an increase in anticoagulation uptake at 1 year (13). Ultimately the best clinical outcomes are with an integrated or holistic care approach to AF management, based on the ABC (Atrial fibrillation Better Care) pathway (A/Avoid stroke; B/Better symptom management; C/Cardiovascular risk and comorbidity optimization) which has been proven in independent observational (14, 15) and clinical trial cohorts (15-17), hence recommended in the ESC guidelines (8).