BACKGROUND: The impact of direct oral anticoagulants (DOAC) on haematoma size after intracerebral haemorrhage (ICH) compared to no-anticoagulation is controversial and prospective data are lacking. METHODS: The investigator-initiated, multicentre, prospective RASUNOA-prime study enrolled patients with non-traumatic ICH and atrial fibrillation while on a DOAC, vitamin K antagonist (VKA) or no anticoagulation (non-OAC). Neuroimaging was reviewed centrally blinded to group allocation. Primary endpoint was haematoma expansion (≥ 6.5 ml or ≥ 33%, any new intraventricular blood or an increase in modified Graeb score by ≥ 2 points) between baseline and follow-up scan within 72 h after symptom onset. RESULTS: Of 1,440 patients screened, 951 patients with ICH symptom onset less than 24 h before admission were enrolled. Baseline scans were performed at a median of 2 h (IQR 1-6) after symptom onset. Neurological deficit and median baseline haematoma volumes (11 ml
IQR 4-39) did not differ among 577 DOAC, 251 VKA and 123 non-OAC patients. Haematoma expansion was observed in DOAC patients in 142/356 (39.9, 95%-CI 34.8-45.0%), VKA in 47/155 (30.3, 95-CI 23.1%-37.6%), versus non-OAC in 22/74 (29.7, 19.3-40.1%). Unspecific reversal agents in DOAC-ICH (212/356, 59.6%) did not affect the haematoma expansion rate compared to no-antagonization. CONCLUSION: Baseline haematoma volume and risk of haematoma expansion did not differ statistically significantly in patients with and without DOAC.