Role of Anticoagulant Therapy in Hemorrhagic Transformation of Acute Ischaemic Stroke in Patients with Non-Valvular Atrial Fibrillation: Reply to Commentaries

We would like to thank George Vilanilam and colleagues for their interest in our article [1]. As they correctly state, it is plausible that treatment with intravenous thrombolysis (IVT) could have influenced the association between baseline treatment with direct oral anticoagulants (DOACs) and symptomatic hemorrhagic transformation (sHT) of the cerebral infarct found in our cohort. We did not assess this specific association as it was not part of the main objective of our study. Nonetheless, we have some data not included in our article that could be of interest regarding this matter. In patients treated with IVT, the rate of sHT was higher in patients treated with DOACs than in those treated with vitamin K antagonists but without statistical significance [1 (25%) vs. 11 (12%); P = 0.41]. In patients not treated with IVT, the rate of sHT was significantly lower in those treated with DOACs [3 (1.25%) vs. 35 (3.86%); P = 0.045]. The association between previous anticoagulant treatment and sHT disappeared when it was adjusted for IVT (Cochran–Mantel– Haenszel test; odds ratio, 0.413; 95% confidence interval, 0.15–1.16; P = 0.094). These results should be interpreted cautiously because of several factors: (i) our study was not designed for assessing this association, (ii) IVT treatment was not randomly assigned and (iii) the number of patients ‘exposed’ to IVT and DOACs was extremely small (four cases). We did not assess the effect of cerebral microbleeds on sHT because its detection requires a cerebral magnetic resonance imaging scan, which is not a standard of care for acute ischaemic stroke in our setting, although we believe that this is a factor that should be considered in future research. Similarly, we did not measure HAS-BLED scale score because inferring the score from clinical notes is unreliable (i.e. international normalized ratio lability is not usually included in the patient’s history) and, perhaps more importantly, the HAS-BLED scale has not been validated for patients being treated with DOACs [2] (just for those treated with vitamin K antagonists), making any comparison between study groups meaningless.