Number of Cerebral Microbleeds and Risk of Intracerebral Hemorrhage After Intravenous Thrombolysis

BACKGROUND AND PURPOSE: Cerebral microbleeds (CMBs) are found in a substantial proportion of patients with ischemic stroke eligible for treatment with intravenous thrombolysis. Until now, there is limited data on the impact of multiple CMBs on occurrence of intracerebral hemorrhage (ICH) after intravenous thrombolysis. METHODS: Between 2008 and 2013, all patients receiving MRI-based intravenous thrombolysis were identified within our prospective thrombolysis register. Number of CMBs was rated on pretreatment T2*-weighted MRI by a rater blinded to clinical data and follow-up. Outcomes of interest were occurrence of symptomatic ICH (sICH) and parenchymal hemorrhage (PH). RESULTS: Among 326 included patients, 52 patients had a single CMB (16.0%), 19 had 2 to 4 CMBs (5.8%), and 10 had ≥5 CMBs (3.1%). Frequency of sICH/PH was 1.2%/5.7% in patients without CMBs, 3.8%/3.8% in patients with a single CMB, 10.5%/21.1% in patients with 2 to 4 CMBs, and 30.0%/30.0% in patients with ≥5 CMBs, respectively (each P for trend<0.01). The unadjusted odds ratio per additional CMB for sICH was 1.19 (95% confidence interval, 1.07-1.33; P<0.01) and for PH was 1.13 (95% confidence interval, 1.03-1.24; P=0.01). Compared with patients without CMBs, both patients with 2 to 4 CMBs (P=0.02/P=0.02) and patients with ≥5 CMBs (P<0.01/P<0.01) had significantly increased odds ratios for sICH and PH, whereas in patients with a single CMB, odds ratios were not significantly increased (P=0.21/P=0.59). The association of CMB burden with sICH/PH remained significant after adjustment for possible confounders (age, age-related white matter changes score, atrial fibrillation, onset-to-treatment time, prior statin use, and systolic blood pressure on admission). CONCLUSIONS: Our findings indicate a higher risk of sICH and PH after intravenous thrombolysis when multiple CMBs are present, with a graded relationship to increasing baseline CMB number.