Abstract 193: Cerebral Cavernous Malformation Location and Mode of Presentation Predict the Risk of Hemorrhage During Their Untreated Clinical Course: Individual Patient Data Meta-Analysis
INTRODUCTION: Cerebral cavernous malformations (CCM) are the second commonest incidental finding on brain MRI and may cause symptomatic intracranial hemorrhage (ICH). However, the risk and predictors of ICH from CCM remain uncertain.
HYPOTHESIS: Presentation with ICH or non-hemorrhagic focal neurological deficit (FND) attributable to CCM, brainstem CCM location, female sex, increasing age and multiple CCM predict ICH occurrence during follow-up without CCM treatment.
METHODS: Three hospital-based cohorts and two cohorts from a population-based study provided individual patient data on clinical course from CCM diagnosis until either first CCM treatment or last available follow-up. We used survival analysis of each cohort to estimate the 5-year risk of symptomatic ICH or new FND, multivariable Cox regression to identify baseline predictors of outcome, and random-effects models to pool estimates in meta-analysis.
RESULTS: Among 988 adults who experienced 62 ICHs during 3,232 person-years of follow-up, clinical presentation with ICH or FND (hazard ratio [HR] 7.4, 95% confidence interval [CI] 2.9-19.2) and brainstem CCM location (HR 5.7, 95% CI 3.2-10.3) were associated with a higher risk of a first ICH within five years of CCM diagnosis, but age, sex and CCM multiplicity were not. The 5-year estimated risk of ICH during untreated follow-up was 2.4% (95% CI 1.0-3.8) for 566 adults without ICH/FND from CCM outside the brainstem, 5.5% (95% CI 0-12.9) for 51 adults without ICH/FND from brainstem CCM, 10.6% (95% CI 5.1-16.1) for 198 adults with ICH/FND from CCM outside the brainstem, and 25.7% (95% CI 18.3-33.1) for 173 adults with ICH/FND from brainstem CCM. In secondary analyses of first ICH or FND, event rates increased but predictors remained unchanged.
CONCLUSION: Mode of clinical presentation and CCM location are independently associated with ICH or FND within five years of CCM diagnosis, which can inform decisions about CCM treatment.
Author Disclosures: M.A. Horne: None. K.D. Flemming: None. I. Su: None. C. Stapf: None. R.D. Brown: None. T.J. Christianson: None. R. Agid: None. K. terBrugge: None. R. Willinsky: None. S. Maxwell: None. G.D. Murray: None. R. Al-Shahi Salman: None.
- © 2015 by American Heart Association, Inc.