Rationale for the trial

Scientific evidence

Guidelines, a systematic review, and expert opinions have confirmed therapeutic uncertainty

Recent guidelines and a working party have not found high quality evidence to enable them to make strong recommendations about whether and when to start full dose oral anticoagulation (OAC) for people with atrial fibrillation (AF) after intracranial haemorrhage (ICH). Our recent Cochrane systematic review did not identify any publications of completed randomised controlled trials comparing full dose OAC to other interventions for people with AF after intracranial haemorrhage. Several narrative reviews, surveys of experts and debates have been unable to reach agreement on the optimal management of patients in AF with intracranial haemorrhage associated with full dose OAC.

Variation in everyday clinical practice

Uncertainty about the effects of full dose OAC after intracranial haemorrhage is likely to have contributed to variation in clinical practice around the world. Although 75-96% of patients with intracranial haemorrhage and AF have a CHA₂DS₂-VASc score of ≥2, in the past only the proportion of these patients starting full dose OAC after intracranial haemorrhage has varied markedly between Europe (14%), Sweden (9-11%), Denmark (17%), Germany (24%), USA (48%), and Japan (91%).

Recent evidence that full dose oral anticoagulation appears to be a promising intervention for people with AF after ICH

Associations observed in recently-published non-randomised studies suggest that starting full dose OAC may be superior to not starting OAC in patients with intracranial haemorrhage (especially ICH) and indications (especially AF) for full dose OAC . The vast majority of patients with AF and a history of intracranial haemorrhage have used Vitamin K antagonist (VKA) OAC in the epochs of the published observational studies.

Unknown influence of radiographic biomarkers of cerebral small vessel disease

Brain microbleeds on MRI are radiographic biomarkers of cerebral small vessel diseases that are associated with VKA use in the general population and at the time of Intracerebral haemorrhage (ICH). Use of VKA OAC , but perhaps not DOAC, seems to be associated with an increase in the number of brain microbleeds on repeat MRI after one year. Brain microbleeds are risk factors for recurrent ICH in people with ICH, but they are also risk factors for ischaemic stroke. Uncertainty remains about whether these biomarkers might modify the effects of full dose OAC on recurrent intracranial haemorrhage or ischaemic stroke, particularly when cerebral amyloid angiopathy is the suspected small vessel disease underlying lobar ICH, although OAC use after ICH did not appear to differ between lobar and non-lobar ICH. Careful sub-group analysis of randomised controlled trials will be the best way of investigating heterogeneity/modification of treatment effects by microbleed presence/distribution/burden.

Randomised trials of starting versus not starting full dose oral anticoagulation for people with AF after intracranial haemorrhage are needed

The magnitude and precision of the observed associations between starting full dose OAC and good outcome in observational studies are not sufficiently ‘dramatic’ by the Oxford Centre for Evidence Based Medicine’s criteria to rule out the possibility of selection bias and confounding by indication accounting for the observed associations.

Therefore, the next step to improve the outcome for patients with AF who survive intracranial haemorrhage is to seek definitive proof of the overall benefit of full dose OAC in randomised controlled trials, and to investigate heterogeneity of treatment effect in sub-groups (e.g. by brain MRI biomarkers).

This article was published on 19 Jul, 2017