Diffusion-Weighted Imaging Hyperintensities in Subtypes of Acute Intracerebral Hemorrhage: Meta-Analysis

Stroke. 2019 Jan;50(1):135-142. doi: 10.1161/STROKEAHA.118.021407. Epub 2018 Dec 7.

Abstract

Background and Purpose- Diffusion-weighted imaging (DWI) hyperintensities in intracerebral hemorrhage (ICH) are associated with increased risk of recurrent ICH, cognitive impairment, and death, but whether these lesions are specific to a subtype of ICH remains uncertain. We investigated the association between DWI lesions and ICH subtype and explored the risk factors for DWI lesions. Methods- In a systematic review of ICH studies, we identified those reporting prevalence of DWI lesions. Two reviewers independently assessed study eligibility and risk of bias and collected data. We determined the pooled prevalence of DWI lesions within 90 days after ICH onset for cerebral amyloid angiopathy- and hypertensive angiopathy-related ICH using random-effects meta-analysis. We calculated odds ratios to compare prevalence of DWI lesions by ICH subtype and to assess risk factors for DWI lesions. Results- Eleven studies (1910 patients) were included. The pooled prevalence of DWI lesions was 18.9% (95% CI, 11.1-26.7) in cerebral amyloid angiopathy- and 21.0% (95% CI, 15.3-26.6) in hypertensive angiopathy-related ICH. There was no difference in the prevalence of DWI lesions between cerebral amyloid angiopathy- (64/292 [21.9%]) and hypertensive angiopathy-related ICH (79/370 [21.4%]; odds ratio, 1.25; 95% CI, 0.73-2.15) in the 5 studies reporting data on both ICH pathogeneses. In all ICH, presence of DWI lesions was associated with neuroimaging features of microangiopathy (leukoaraiosis extension, previous ICH, and presence, and number of microbleeds) but not with vascular risk factors or the use of antithrombotic therapies. Conclusions- Prevalence of DWI lesions in acute ICH averages 20%, with no difference between cerebral amyloid angiopathy- and hypertensive angiopathy-related ICH. Detection of DWI lesions may add valuable information to assess the progression of the underlying microangiopathy.

Keywords: cerebral amyloid angiopathy; cerebral hemorrhage; cerebral infarction; leukoaraiosis; prevalence.