Prevalence, Characteristics, and Outcomes of Undetermined Intracerebral Hemorrhage: A Systematic Review and Meta-Analysis.

Background and Purpose: There are scarce data regarding the prevalence, characteristics and outcomes of intracerebral hemorrhage (ICH) of undetermined (unknown or cryptogenic) etiology. We sought to determine the prevalence, radiological characteristics, and clinical outcomes of undetermined ICH. Methods: Systematic review and meta-analysis of studies involving patients with spontaneous ICH was conducted to primarily assess the prevalence and clinical-radiological characteristics of undetermined ICH. Additionally, we assessed the rates for ICH secondary to hypertensive arteriopathy and cerebral amyloid angiopathy. Subgroup analyses were performed based on the use of (1) etiology-oriented ICH classification, (2) detailed neuroimaging, and (3) Boston criteria among patients with cerebral amyloid angiopathy related ICH. We pooled the prevalence rates using random-effects models, and assessed the heterogeneity using Cochran Q and I2 statistics. Results: We identified 24 studies comprising 15 828 spontaneous ICH patients (mean age, 64.8 years; men, 60.8%). The pooled prevalences of hypertensive arteriopathy ICH, undetermined ICH, and cerebral amyloid angiopathy ICH were 50% (95% CI, 43%­58%), 18% (95% CI, 13%­23%), and 12% (95% CI, 7%­17% [P<0.001 between subgroups]). The volume of ICH was the largest in cerebral amyloid angiopathy ICH (24.7 [95% CI, 19.7­29.8] mL), followed by hypertensive arteriopathy ICH (16.2 [95% CI, 10.9­21.5] mL) and undetermined ICH (15.4 [95% CI, 6.2­24.5] mL). Among patients with undetermined ICH, the rates of short-term mortality (within 3 months) and concomitant intraventricular hemorrhage were 33% (95% CI, 25%­42%) and 38% (95% CI, 28%­48%), respectively. Subgroup analysis demonstrated a higher rate of undetermined ICH among studies that did not use an etiology-oriented classification (22% [95% CI, 15%­29%]). No difference was observed between studies based on the completion of detailed neuroimaging to assess the rates of undetermined ICH (P=0.62). Conclusions: The etiology of spontaneous ICH remains unknown or cryptogenic among 1 in 7 patients in studies using etiology-oriented classification and among 1 in 4 patients in studies that avoid using etiology-oriented classification. The short-term mortality in undetermined ICH is high despite the relatively small ICH volume.