Comparing anticoagulants for AF: Which OAC is best for patients with dementia?
Comparative Effectiveness and Safety of Oral Anticoagulants by Dementia Status in Older Patients With Atrial Fibrillation
Introduction and methods
Background
To reduce the risk of embolic stroke, most older patients with AF should be treated with oral anticoagulation (OAC) [1-6]. While there is a need for an optimal stroke prevention strategy for AF patients who are living with dementia, this population was greatly underrepresented in RCTs showing the efficacy and safety of direct OACs versus warfarin [7-10].
Aim of the study
The authors examined the comparative safety and effectiveness of 4 OACs (apixaban, dabigatran, rivaroxaban, and warfarin) by assessing the risks of ischemic stroke and major bleeding by dementia status in older AF patients.
Methods
In this retrospective comparative effectiveness study, data from 3 large nationwide claims databases in the USA (Optum Clinformatics Data Mart, IBM MarketScan Research Database, and Medicare fee-for-service Parts A, B, and D) were pooled. The dataset comprised 1,160,462 patients ≥65 years diagnosed with AF (mean age: 77.4 years; SD: 7.2) who filled a new prescription for a specific OAC; 7.9% of them had a dementia diagnosis. Based on the medication, the authors established 3 comparative cohorts: warfarin versus apixaban (n=501,990), dabigatran versus apixaban (n=126,718), and rivaroxaban versus apixaban (n=531,754). To adjust for potential confounding, 1:1 propensity score matching was used..
Outcomes
The primary endpoint was a composite outcome of ischemic stroke or major bleeding events in the 6 months following the OAC dispensing date. Secondary endpoints included the individual components of the primary endpoint (i.e., ischemic stroke and components of major bleeding events, including gastrointestinal bleeding, other extracranial bleeding, and intracranial hemorrhage) and all-cause mortality.
Main results
Warfarin versus apixaban
- In the group of patients with dementia, warfarin users showed a higher incidence rate of the primary composite endpoint compared with apixaban users (95.7 vs. 64.2 events per 1000 person-years (PYs); adjusted HR (aHR): 1.5; 95%CI: 1.3–1.7). The magnitude of the benefit associated with apixaban on the HR scale was similar in the group of patients without dementia (51.1 vs. 36.3 events per 1000 PYs; aHR: 1.5; 95%CI: 1.4–1.5; P=0.92 for heterogeneity).
- However, on the rate difference scale, the magnitude of the benefit was greater for patients with dementia compared with those without dementia (29.8 events per 1000 PYs; 95%CI: 18.4–41.1 vs. 16.0 events per 1000 PYs; 95%CI: 13.6–18.4; P=0.02 for heterogeneity).
- Higher incidence rates and treatment effect heterogeneity in rate difference for patients with versus without dementia were also observed for major bleeding events, including intracranial and extracranial bleeding (overall rate difference: 24.3 events per 1000 PYs; 95%CI: 14.3–34.3 vs. 13.1 events per 1000 PYs; 95%CI: 10.1–16.1; P=0.04 for heterogeneity), and all-cause mortality (51.7 events per 1000 PYs; 95%CI: 31.8–71.5 vs. 8.1 events per 1000 PYs; 95%CI: 5.0–11.2; P<0.001 for heterogeneity) but not for ischemic stroke (P=0.46 for heterogeneity).
Dabigatran versus apixaban
- The incidence rate of the primary composite endpoint was higher in patients with dementia taking dabigatran compared with those taking apixaban (84.5 vs. 54.9 events per 1000 PYs; aHR: 1.5; 95%CI: 1.2–2.0). A similar benefit of apixaban was seen in patients without dementia (33.6 vs. 29.0 events per 1000 PYs; aHR: 1.2; 95%CI: 1.1–1.4; P=0.08 for heterogeneity).
- The magnitude of the benefit associated with apixaban was again greater on the rate difference scale for patients with versus without dementia (29.6 events per 1000 PYs; 95%CI: 11.6–47.6 vs. 5.8 events per 1000 PYs; 95%CI: 1.1–10.4; P=0.01 for heterogeneity).
- Of the secondary endpoints, treatment effect heterogeneity in rate difference by dementia status was only observed for major bleeding events (18.6 events per 1000 PYs; 95%CI: 3.4–33.7 vs. 1.9 events per 1000 PYs; 95%CI: –4.4 to 8.2; P=0.05 for heterogeneity), in particular for gastrointestinal bleeding but not for intracranial hemorrhage.
Rivaroxaban versus apixaban
- Treatment with rivaroxaban versus rivaroxaban was associated with a higher incidence rate of the primary composite endpoint in both patients with dementia (87.4 vs. 68.5 events per 1000 PYs; aHR: 1.3; 95%CI: 1.1–1.5) and those without dementia (45.2 vs. 29.8 events per 1000 PYs; aHR: 1.5; 95%CI: 1.4–1.7; P=0.04 for heterogeneity ).
- The magnitude of the benefit associated with apixaban did not differ significantly on the rate difference scale between patients with versus without dementia (20.5 events per 1000 PYs; 95%CI: 9.8 –31.1 vs. 15.9 events per 1000 PYs; 95%CI: 11.4–20.3; P=0.43 for heterogeneity).
- For all-cause mortality, the absolute reduction associated with apixaban was greater in patients with dementia compared with those without (44.4 events per 1000 PYs; 95%CI: 24.1–64.6 vs. 6.3 events per 1000 PYs; 95%CI: 3.6–9.1; P<0.001 for heterogeneity).
Conclusion
In this retrospective comparative effectiveness study comprising >1 million AF patients aged ≥65 years in the USA, treatment with apixaban was associated with a lower rate of ischemic stroke or major bleeding on the relative scale compared with dabigatran, rivaroxaban, or warfarin, mostly irrespective of dementia status. However, in terms of absolute rate reduction, the magnitude of the benefit associated with apixaban was greater in patients with dementia compared with those without dementia, particularly for major bleeding. The authors believe “[t]hese results support the use of apixaban for anticoagulation therapy in vulnerable patients with frailty, particularly those living with dementia.”
Share this page with your colleagues and friends: