Abstract 3212: Aggravated Intraventricular Hemorrhage and White Matter Lesions: A Nationwide Cohort of Intracerebral Hemorrhage Patients

Abstract

Background. Intraventricular hemorrhage (IVH) secondary to intracerebral hemorrhage (ICH) is an important indicator of poor outcome. Location and size of hematoma have been known to associate with IVH, but other clinical and neuroimaging predictors are yet to be discovered. Among them, white matter lesion (WML), which reflects accumulation of chronic sublethal ischemic injuries and periventricular edema, may signify vulnerability of individual brain, hence being associated with development and aggravation of IVH. Methods. A total of 1,604 ICH patients were collected through 33 stroke centers in South Korea with nationwide coverage. We excluded 199 subjects whose computed tomography (CT) scans were taken more than 48 hours after symptom onset, 89 subjects whose images were of poor quality, 42 subjects who had primary IVH without ICH and 12 subjects who were misclassified. Finally, 1,262 acute ICH patients were included in our analysis dataset. Clinical, laboratory, and imaging information was collected during admission. Experienced stroke neurologists independently reviewed CT scans of included patients for rating WML (van Swieten method) and IVH scores, with good inter-rater agreements. IVH score is a recently proposed semi-quantitative and validated method for estimating volume of IVH. Functional outcomes included Glasgow coma scale score and mortality status at the end of 2006. Results. Among the 1,262 ICH patients, 343 subjects (27.2%) were found to have IVH. Patients with IVH are older and had higher level of blood glucose and ICH volume. The proportion of subjects with extensive WML was significantly higher in patients with IVH then those without IVH (33.8% vs. 16.3%; P <0.01). The association between WML and IVH remained significant after controlling for relevant covariates (OR 1.43 and 95% CI 1.02 - 2.02 for mild WML; OR 2.66 and 95% CI 1.83 - 3.87 for extensive WML; compared to no WML). Estimated mean of IVH score of extensive WML group was also significantly higher than that of no WML group (estimated mean ± standard error; 11.5 ± 0.5 vs 9.4 ± 0.5; P =0.01 from analysis of covariance with Tukey’s post-hoc multiple comparison). For outcomes after ICH, presence of IVH was associated with increased odds of having surgical management (OR 3.48 and 95% CI 2.54 - 4.76), lower Glasgow coma scale score (OR 1.90 and 95% CI 1.46 - 2.48), and higher risk of long-term mortality (HR 1.30 and 95% CI 1.03 - 1.64). Conclusion. From this study, we documented that WML was associated with development and increased volume IVH in acute ICH patients. WML would be considered in diagnosis and management of ICH patients with IVH.