Oral anticoagulation protects against stroke in atrial fibrillation, heart failure

For lowering the incidence of stroke in patients with either atrial fibrillation (AF) or heart failure (HF) with reduced ejection fraction (HFrEF), oral anticoagulation is the way to go, according to the results of a meta-analysis.

The association of oral anticoagulation with ischaemic stroke, haemorrhagic stroke, myocardial infarction, or fatal bleeding has been consistent in the AF and HFrEF populations. However, there have been differences in incidence of clinical events, with markedly higher mortality in HFrEF trials, the investigators noted.

AF and HFrEF patients are at increased risk of cardioembolic stroke. In the former, oral anticoagulation is strongly recommended (Grade 1A) to reduce the risk of ischaemic stroke. In comparison, oral anticoagulation has not shown superiority over antiplatelet therapy or control for cardiovascular prevention in clinical trials in patients with HFrEF in sinus rhythm, plus guideline recommendations are inconsistent. [Eur Heart J 2020;42:373-498; Circulation 2019;140:e125-e151; Can J Cardiol 2018;34:1371-1392; Circulation 2009;119:1616-1624; N Engl J Med 2018;379:1332-1342]

“For many stroke physicians, anticoagulation decision-making for patients with AF is clear, while the decision in patients with HF is more challenging because of inconsistent guideline recommendations,” the investigators said. [J Cardiac Fail 2010;16:e1-e2; Circulation 2013;128:e240-e327]

“Our analysis offers relevant context for the benefit of oral anticoagulation in HFrEF by comparing directly to a population where the benefit is universally accepted and consistently recommended in guidelines. Clinical context (eg, prior ischaemic stroke), and patient preference, should determine selective prescribing of low-dose or treatment-dose regimens of oral anticoagulant therapy in patients with HFrEF,” they added.

The meta-analysis included 15 AF trials (n=19,332, mean follow-up 23.1 months) and six HFrEF trials (n=9,866, mean follow-up 23.9 months). The components of the primary outcome differed between AF and HFrEF trials, with five HFrEF trials (95.1 percent of the corresponding population) including all-cause mortality in their primary composite endpoint as opposed to only one AF trial (0.38 percent of the corresponding population).

Pooled data showed that control treatments were associated with a lower incidence of all stroke in the HFrEF vs AF trials (3.1 percent vs 7.0 percent) but a higher incidence of all-cause mortality (19.0 percent vs 9.6 percent). [Stroke 2021;doi:10.1161/STROKEAHA.120.033910]

Compared with control, oral anticoagulation significantly reduced the risk of all stroke, with this benefit seen for both AF (odds ratio [OR], 0.51, 95 percent confidence interval [CI], 0.42–0.63) and HFrEF (OR, 0.61, 95 percent CI, 0.47–0.79; I2, 12.4 percent; p_interaction=0.31). Furthermore, the effect of oral anticoagulation on cardiovascular events, mortality, or bleeding was similar in the two populations.

“Based on our findings, use of oral anticoagulation would seem reasonable among patient populations who are considered at high risk of cardioembolic stroke, such as those with recent ischaemic stroke or transient ischaemic attack, where HFrEF is implicated as causal,” the investigators pointed out.

“Although the risk of bleeding in patients with HFrEF is a commonly cited reason to avoid oral anticoagulation, our findings did not reveal major differences in the risk of major or fatal bleeding in patients with AF compared with those with HFrEF (p_interaction=0.17),” they explained.

The investigators, however, acknowledged that oral anticoagulation was associated with significant increase in major haemorrhage in the HFrEF population. Still, there was no excess risk of fatal haemorrhage in either population. [Heart 2019;105:1325-1334]

“We think that our meta-analysis provides an innovative perspective through including clinical trials from both populations in a single analysis, lending insights that may not be apparent through indirect comparisons of individual meta-analyses. Moreover, it permitted us to explore the effect of methodological differences (eg, outcome measures) between trials of patients with AF and those with HFrEF,” they said.