Direct Oral Anticoagulants versus Warfarin in Octogenarians with Nonvalvular Atrial Fibrillation: A Systematic Review and Meta-Analysis

Direct oral anticoagulants (DOACs) have been demonstrated to be more effective and safer than vitamin-K antagonist (VKA) for stroke prevention in patients with nonvalvular atrial fibrillation (AF). This meta-analysis aims to assess the effect of DOACS vs. VKA in patients ≥ 80 and AF. Primary endpoints were stroke or systemic embolism and all-cause death. Secondary endpoints included major bleeding, intracranial bleeding, and gastrointestinal bleeding. A random-effects model was selected due to significant heterogeneity. A total of 147,067 patients from 16 studies were included, 71,913 (48.90%) treated with DOACs and 75,154 with VKA (51.10%). The stroke rate was significantly lower in DOACs group compared with warfarin group (Relative risk (RR): 0.72; 95% confidence interval (CI): 0.63–0.82; p < 0.001). All-cause mortality was significantly lower in DOACs group compared with warfarin group (RR: 0.82; 95% CI: 0.70–0.96; p = 0.012). Compared to warfarin, DOACs were not associated with reductions in major bleeding (RR: 0.85, 95% CI 0.69–1.04; p = 0.108) or gastrointestinal bleeding risk (RR: 1.08, 95% CI 0.76–1.53; p = 0.678) but a 43% reduction of intracranial bleeding (RR: 0.47, IC 95% 0.36–0.60; p < 0.001) was observed. Our meta-analysis demonstrates that DOACs are effective and safe with statistical superiority when compared with warfarin in octogenarians with AF.


Introduction
The incidence of nonvalvular atrial fibrillation (AF) increases exponentially with age [1,2], and its prevalence varies from 0.1% among patients younger than 55 years to 9-17.7% in those aged 80 years or older [2][3][4][5][6]. Octogenarians are growing faster than those above age 65, and the global population aged 80 years and over is projected to triple in 2050 [7].
AF is an independent risk factor for stroke, and age per se entails a greater risk of strokes [8]. Furthermore, age is considered a risk factor for CHA2DS2VASc and HASBLED scores [9]. Besides, elderly patients commonly present multiple comorbidities such as dementia, high risk of falls, chronic kidney disease, hypertension, and diabetes, which may lead to greater complexity in their management [10].
Direct oral anticoagulants (DOACs) have been demonstrated to be more effective and safer than vitamin K antagonists (VKA) for stroke prevention in patients with nonvalvular AF [11][12][13][14]. Moreover, DOACs overcome the limitations of VKA, showing fewer drug and dietary interactions and rapid onset of action. Based on these advantages over VKA, DOACs have been recommended as the first-choice therapy to prevent stroke in AF [15][16][17]. However, the potential benefits of DOACs have not been specifically studied in octogenarians with AF because the elderly population was underrepresented in the pivotal randomized controlled trials (RCTs), which excluded these high-risk patients. The rate of patients aged ≥80 and ≥85 years in the pivotal RCTs was 13-17% and 4-7%, respectively [18][19][20][21][22][23].
Considering the current lack of information from clinical trials to make clinical recommendations about the efficacy of oral anticoagulants in the management of patients aged 80 or older with AF, and to provide some insights into this issue, we conducted a meta-analysis to evaluate the clinical benefit of DOACs compared to VKA in octogenarians with nonvalvular AF.

Material and Methods
We performed an intention-to-treat meta-analysis in line with recommendations from the Cochrane Collaboration and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement [24] using standard software (STATA 14.3, STATA Corp, College Station, TX, USA). We included studies that compared DOACs and VKA for patients with nonvalvular AF with an indication for stroke prevention. The primary endpoints analyzed were stroke and systemic embolism, and all-cause death. According to each study's definitions, secondary endpoints were major bleeding, intracranial bleeding, and gastrointestinal (GI) bleeding. The raw numbers of incident endpoints reported in each study were used.
We conducted a systematic search (using PUBMED, EMBASE, and Cochrane Central Register of Controlled Trials (CENTRAL), and Google Scholar), without language restriction, for studies and randomized clinical trials using the Medical Subject Headings terms "dabigatran", "rivaroxaban", "apixaban", "edoxaban", "warfarin", "vitamin K antagonist", "octogenarians", "elderly", or "very elderly". The systematic search was restricted from 2010 to 2020. Information on sample size, treatment type and outcomes were collected from the published papers. For the eligible studies, two authors independently abstracted data into a standardized database; discrepancies in data extraction and quality assessment were resolved by discussion or consensus with a third author. For the studies that evaluated different doses or different drugs each strategy was analyzed as a separate study ( Figure 1).

Data Synthesis and Statistical Analyses
The percentage of variability across studies attributable to heterogeneity beyond chance was estimated using the I2 statistic. A random-effects model was selected because significant heterogeneity was observed. Sensitivity analyses were performed by the analysis of between-study variation using the study-specific standard errors. Potential sources of heterogeneity between trials or treatments were tested by meta-regression analyses and the effect of age, the number of patients treated, type of study (randomization) or propensity score matching, gender distribution, the prevalence of hypertension, diabetes, previous stroke, or heart failure was assessed. We also tested the small-study effect by the Egger test [25].

Sensitivity Analyses and Metaregression
Significant heterogeneity was observed in all endpoints. The funnel plots are presented in Figure 6. Metaregression identified inclusion site in North-America (p < 0.001), the ELDERCARE-AF results (p = 0.023), control arm different than VKA (p = 0.006), and the prevalence of hypertension (p = 0.042) as main sources of heterogeneity for stroke risk reduction. The type of DOAC was the main source of sources of heterogeneity for all-cause mortality (p < 0.001) and major bleeding (p = 0.03) risk reduction. None of the clinical characteristics presented in Table 1 were detected as a source of heterogeneity for gastrointestinal bleeding. No small-study effect was detected for stroke, all-cause mortality, major bleeding, or gastrointestinal bleeding.   Hypertension (p = 0.04) was detected as the leading source of heterogenicity for intracranial bleeding, and a small study effect was also detected (Harbor test p = 0.029).

Discussion
This meta-analysis of currently available evidence from the subanalysis of clinical trials and real-world cohorts endorsed the efficacy and safety of DOACS vs. VKA in octogenarians with AF, associating significant reductions in stroke, all-cause mortality, and intracranial bleeding. Prevalence rates of AF increase significantly with age, ranging from 0.1% among patients younger than 55 years to 9-17.7% in those aged 80 years or older [2][3][4][5][6]. Due to the increased life expectancy, the number of patients aged 80 years or older with AF is expected to rise 4-fold by 2050, representing more than half of all AF cases [2].
Patients with AF aged 75 years or older have a worse prognosis than those between 65 and 74 years, with higher mortality rates and major adverse cardiac events [35]. The classic BAFTA study carried out with warfarin, and the more recent PREFER-AF registry has shown the net clinical benefit of anticoagulation among the elderly or very old population (>85 years) in terms of survival and reduction of thromboembolic events [36,37]. The advent of DOACS has improved prescription rates of oral anticoagulation (OAC) in older people, but OAC remains underutilized in up to 30% of patients with high stroke risk [38]. Every decade increase in age is related to up to a 14% reduction in VKA use, regardless of other stroke risk factors [39]. It has been described that antiplatelet therapy, age over 90 years, falls risk, and nursing home residency are the main reasons for not starting OAC in frail elderly, despite a clear indication that frailty is associated with an increased risk of stroke but not with the risk of major bleeding [40].
There are several important findings in this meta-analysis evaluating the clinical benefit of DOACs in octogenarians with AF. First, compared to VKA, DOACs significantly reduced the risk of stroke, all-cause mortality, and intracranial bleeding in patients with AF aged 80 or older. Moreover, in this group of patients, treatment with DOACs was not associated with greater major or gastrointestinal bleeding risk than VKA. Therefore, old age alone should not be a criterion for withholding OAC therapy with DOACs. Nonetheless, these results do not contradict the fact that the use of OAC in octogenarian patients with AF should be based on an individualized, case-by-case approach with an adequate assessment of stroke risk, bleeding risk, and comorbidities [3].
Our results take a step forward within the currently available evidence, especially because although different meta-analyses analyzed the efficacy and safety of the use of DOACs in elderly patients [3,[41][42][43][44][45][46], there is currently no one that has focused on patients over 80 years of age. Furthermore, a dedicated trial with a head-to-head comparison among DOACs in this group of patients is unlikely. This lack of evidence in octogenarians justifies our meta-analysis that demonstrated the beneficial effect of DOACs on stroke and all-cause mortality. The highest reduction associated with DOACs were for intracranial hemorrhage, a devastating bleeding complication that is reduced by 43%. Octogenarians have a high risk of both bleeding and stroke, which could explain the absence of difference in major bleeding rates in patients treated with DOACs or VKA. Nonetheless, the significant reductions of stroke, all-cause mortality, and intracranial bleeding highlight the net benefit of DOACs vs. VKA in octogenarians with AF.

Limitations of the Study
This meta-analysis might be limited because it mainly included a subanalysis of randomized clinical trials and observational studies and, therefore, evidence might be less reliable than randomized trials. Nonetheless, this also highlights the unmet need for accurately designed and powered clinical trials in elderly patients with AF. We also observed differences in clinical characteristics and meant follow-up in the studies included, which might explain the significant heterogeneity observed in all the endpoints. In the review of the manuscripts we have not found information about treatment crossover between both arms, a significant percentage of crossover could affect the quality of the results. In the AVERROES [20] study, control patients took aspirin while control patients in the ELDERCARE-AF [18] were under placebo. Finally, there is a limitation regarding the analysis of the geriatric profile in the elderly, taking into account that RCTs, based on ethical and practical considerations, usually do not include patients with complex real-life conditions (as frailty, multimorbidity, and polypharmacy) and that there are currently no unified recommendations about the best dose for DOACs in this group of patients.

Conclusions
This meta-analysis of more than 147,000 patients with age >80 and AF demonstrates the superiority of DOACs compared to VKA in stroke prevention and all-cause mortality and intracranial bleeding. Funding: This work was supported by "Instituto de Salud Carlos III" and "Fondos Europeos de Desarrollo Regional FEDER" (grant numbers PI20/00637, and CIBERCV16/11/00486).

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable. Data Availability Statement: Data are available from the authors upon reasonable request.

Acknowledgments:
The authors acknowledge the support of the Sections of Geriatric Cardiology of SEC (Spanish Society of Cardiology). The authors also acknowledge GOC Health Consulting for partial editorial assistance in the elaboration of this manuscript.

Conflicts of Interest:
The authors declare no conflict of interest in the elaboration of this article.