Diagnostic Utility of Brain MRI in Spontaneous Intracerebral Hemorrhage: A Retrospective Cohort Study and Meta-Analysis

Background: The diagnostic yield of brain Magnetic Resonance Imaging (MRI) in spontaneous intracerebral hemorrhage (ICH) is unclear. We performed both an independent single-center retrospective cohort study and a meta-analysis to assess the detection rate of secondary lesions on MRI in patients with spontaneous ICH. Methods:In the retrospective cohort study, we examined 856 consecutive patients with spontaneous ICH. Brain MRI scans on admission and follow-up were assessed for secondary lesions. We also examined clinical and CT radiographic variables associated with secondary lesions in univariable analysis. In the meta-analysis we searched PubMed and EMBASE for articles investigating the secondary lesion detection rate on brain MRI in spontaneous ICH. Random-effects models were used to calculate the pooled estimate of secondary lesion detection rate. The meta-analysis used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Results:Of the 856 patients with ICH, 481 (56%) had at least one brain BRI performed [7014 years, 270(56% male)]. 462(54%) had an admission MRI and 138(16%) had both admission and follow-up MRIs. The detection rate of secondary lesions on admission MRIs was 24/462(5.2%). 4/127(3.1%) patients with a negative admission MRI had a lesion identified on follow-up MRI. No clinical or radiographic variables were associated with a secondary lesion on MRI using univariable analysis. The meta-analysis included 5 studies total (4 identified in the PubMed and EMBASE searches as well as our own cohort study) comprising 1147 patients with spontaneous ICH who underwent brain MRI. The pooled detection rate of secondary lesions was 11% (95%CI: 7-16%). Conclusions:No predictors of secondary lesion detection were identified in our cohort study. Prospective studies are required to better understand the diagnostic utility of MRI in spontaneous ICH.