Incidence of Diffusion-Weighted Imaging Lesions in Patients With Intracerebral Hemorrhage in the Acute and Subacute Time Periods.

Background And Objectives
Diffusion-weighted imaging (DWI) lesions in patients with intracerebral hemorrhage (ICH) are associated with poor outcomes. Knowledge about the underlying pathophysiology is scarce, and it is hypothesized that they are related to either the ICH itself, adverse effects of treatment, or the activity of the underlying small vessel disease (SVD) causing the ICH. We investigated their association with time point of MRI acquisition and underlying SVD type and burden.Methods
In this Swiss multicenter ICH cohort, we enrolled patients who underwent MRI within 15 days after SVD-associated ICH. The primary outcome was presence of DWI lesions. Time point of MRI was investigated as a continuous (days) and dichotomized (hyperacute = MRI on admission vs subacute = MRI during follow-up) variable. We measured cerebral amyloid angiopathy (CAA) and SVD severity using MRI burden scores and defined the type of SVD using CADMUS classification and Boston 2.0 criteria. At 3 months, we assessed functional outcome using the modified Rankin Scale score, recurrent ICH, and ischemic stroke.Results
We included 644 patients (median age 73 years, interquartile range [IQR] 64-79, median SVD burden 1 IQR 1-2; median CAA burden 2 IQR 1-3; 208 patients/32.3% with Boston 2.0 CAA, 431 patients/66.9% with mixed CAA-DPA phenotype according to CADMUS). Among enrolled patients, 16.0% underwent hyperacute MRI and 84.0% underwent subacute MRI (median on day 2 IQR 1-5), and 166 patients (25.8%) had DWI lesions (18.4% with hyperacute MRI vs 27.2% with subacute MRI). We observed no association of presence of DWI lesions with hyperacute MRI (adjusted odds ratio [aOR] 0.61, 95% CI 0.36-1.00, p = 0.073) but with time to MRI in days (aOR 1.07, 95% CI 1.00-1.13, p = 0.007). Higher SVD (aOR 1.33, 95% CI 1.12-1.59, p = 0.001) and CAA (aOR 1.29, 95% CI 1.15-1.44, p < 0.001) burdens were associated with presence of DWI lesions. There was no association between type of SVD (CADMUS) or CAA (Boston criteria) and DWI lesions. There was no association between DWI lesions and functional outcome, recurrent ICH or ischemic stroke at 3 months.Discussion
DWI lesions in patients with ICH are already common at baseline. Their prevalence is higher with a longer time since symptom onset and with higher SVD burden. Overall, these findings suggest a relation with the underlying condition resulting in ICH. The significant number of DWI lesions visible on admission MRI might diminish their use as surrogate outcome in future trials in ICH.