Study I: CROMIS-2 (AF) Prospective cohort study of patients anticoagulated after cardioembolic stroke An observational inception cohort study (n=1425) of patients throughout the United Kingdom (UK) - (79 hospitals) started on best practice oral anticoagulant (without prior use) for presumed cardioembolic ischaemic stroke due to non-valvular AF with follow up for the occurrence of intracerebral haemorrhage (ICH) and ischaemic stroke for an average of two years. The main baseline exposures (risk factors of interest) are the presence of cerebral microbleeds (CMBs) on magnetic resonance imaging (MRI), and genetic polymorphisms in candidate genes with potential functional relevance to ICH risk.

Study II: CROMIS-2 (ICH) Observational and genetics study of intracerebral haemorrhage The investigators will also recruit 600 patients admitted to participating centres with ICH (with a target of at least 300 anticoagulant-related ICH cases) and collect DNA to increase the power of the genetic studies. The investigators will collect clinical and imaging data from these ICH cases to investigate risk factors associated with anticoagulant-related ICH compared to non anticoagulant-related ICH.

Background Over the last decade, increasing use of oral anticoagulants to prevent cardioembolic ischaemic stroke due to atrial fibrillation (AF) in an ageing population has led to a five-fold increase in the incidence of anticoagulant-related intracranial haemorrhage (ICH) - a rare but unpredictable and catastrophic complication. Cerebral microbleeds (CMBs) on magnetic resonance imaging (MRI) may predict ICH risk, as may genetic polymorphisms influencing brain small-vessel integrity or anticoagulation stability.

Aims To establish the value of CMBs and genetic factors in predicting symptomatic ICH following best practice oral anticoagulation to prevent recurrent ischaemic stroke due to AF.

Methods CROMIS-2: Study I (AF) - Prospective, multicentre, inception cohort study in 1425 patients with ischaemic stroke due to AF started on best practice oral anticoagulation. Patients will have genetic testing and standardized MRI including Gradient recalled Echo (GRE) at baseline, with follow-up by postal questionnaire (and clinical assessment or medical records surveillance after suspected events), and where possible there will be an in person clinical assessment at 2 years. The investigators will compare the rate of symptomatic ICH between CMB and CMB-free patients and test for associations with plausible candidate genes. The investigators aim to develop and validate a risk model to predict symptomatic ICH following best practice oral anticoagulation to prevent recurrent ischaemic stroke due to AF.

CROMIS-2: Study II (ICH) - An observational study of ICH investigating genetic, clinical and radiological risk factors associated with anticoagulant-related ICH. The investigators will recruit patients admitted to participating centres with ICH (with a target of at least 300 anticoagulant-related ICH cases) and collect DNA to increase the power of the genetic studies. The investigators will collect clinical and imaging data from these ICH cases to investigate risk factors associated with anticoagulant-related ICH compared to non anticoagulant-related ICH.

Expected outcomes A successful predictive model for ICH risk after best practice oral anticoagulation for AF will help to determine whether genetic or CMB screening should be used in clinical practice and future trials. New genetic, clinical and radiological risk factors associated with anticoagulant-related ICH will be identified.

Study I: CROMIS-2 (AF)

Inclusion criteria:

• Adult (≥18y; no upper limit) patients with a clinical diagnosis of non-valvular AF (verified by ECG) with intention to treat with best practice oral anticoagulants (e.g. warfarin)
• Previous ischaemic stroke or TIA diagnosed by treating clinician
• All patients must be able to have GRE MRI before (or within 1 week) of starting best practice oral anticoagulant

Exclusion Criteria:

• Any MRI contraindications
• Previous use of oral anticoagulation
• Definite contra-indication to oral anticoagulation
• Serious head injury (resulting to loss of consciousness)

Study II: CROMIS-2 (ICH)

Inclusion criteria:

• Adult (>18y) patients treated at participating centres with confirmed ICH (confirmed on CT or MRI scans) with or without a history of anticoagulant use at the time of the ICH

Exclusion criteria:

• Known underlying structural cause for ICH (e.g arteriovenous malformation, tumour, cavernoma, intracranial aneurysm, haemorrhagic transformation of an infarct)
• Major head trauma (causing loss of consciousness and though to be sufficient to have caused the ICH) in previous 24 hours

Study I: CROMIS-2 (AF) Prospective cohort study of patients anticoagulated after cardioembolic stroke Study I (AF): 1425 patients from UK centres. All eligible patients with first or recurrent ischaemic stroke and TIA in whom it is decided that best practice oral anticoagulant treatment is to be commenced.

Study II (ICH): The investigators will recruit 600 patients treated at participating hospitals with ICH. Of these patients, 300 ICH cases will be related to anticoagulant use. The investigators will also recruit at least 300 ICH cases not related to anticoagulant use during the study period. Patients seen in outpatient clinics or from existing databases may also be recruited, at centres where these are available.