White matter lesions and poor outcome after intracerebral hemorrhage A nationwide cohort study

Department of Neurology, Seoul National University Hospital, Seoul, 110-744, Republic of Korea.
Neurology (Impact Factor: 8.3). 05/2010; 74(19):1502-10. DOI: 10.1212/WNL.0b013e3181dd425a
Source: PubMed

ABSTRACT The ability to predict poor outcome is important for patient care and treatment decision-making in cases of intracerebral hemorrhage (ICH). Previous studies have included relatively brief follow-up periods and small numbers of patients, and are limited in terms of considerations regarding individual brain vulnerabilities.
The authors prospectively enrolled 1,321 ICH patients nationwide from 33 hospitals. Clinical, laboratory, and imaging variables, including white matter lesions (WMLs), were collected at admission. Immediate outcome after ICH was measured using total Glasgow Coma Scale (GCS) score at admission, early outcome using 30-day mortality, and long-term outcome using relative risk for mortality. The vital status of included patients was ascertained on December 31, 2006, using Korean National Death Certificates (mean follow-up, 32.6 months).
Of the 1,321 ICH patients included, 352 (27.8%) presented with a moderate GCS score (8.5-12.4) and 249 (19.7%) with a severe GCS score (</=8.4). The mortality rate was 9.1% at day 30 post-ICH and 381 patients (29.8%) had died up to the end of December 2006. Extensive WMLs were associated with severe GCS scores at admission (odds ratio [OR] 2.45, 95% confidence interval [CI] 1.73-3.46), 30-day mortality (OR 2.52, 95% CI 1.33-4.75), and the relative risk for mortality (RR 2.61, 95% CI 1.79-3.82) after adjusting for relevant covariates.
These findings suggest that white matter lesions, which may reflect the vulnerability of individual brains to pathologic insults, should be considered when assessing immediate, early, and long-term outcomes after intracerebral hemorrhage.

  • [Show abstract] [Hide abstract]
    ABSTRACT: Bleeding into the brain or adjacent structures is one of the most devastating neurological conditions, incurring tremendous emotional, financial, and societal costs. Imaging is essential to differentiate variants of hemorrhage, as the clinical features may be insufficient. A comprehensive approach to hemorrhage therefore relies on imaging to disclose pathophysiology, elucidate mechanisms, and thereby open further avenues to effective treatment. Hemorrhage patterns from superficial to deep locations in the brain are surveyed in this work, noting myriad potential causes and the influential pathophysiology of arterial ischemia, venous hypertension, and microvascular dysfunction. Recent progress of imaging studies and novel techniques to evaluate hemorrhage are explored. For decades, only computed tomography was available to define a hematoma without corroborating evidence of other pathology whereas multimodal computed tomography and magnetic resonance imaging, including noninvasive imaging of brain tissue, vessels, and perfusion, have now radically altered clinical practice. Imaging of the blood-brain barrier, cerebral microbleeds, coexistent ischemia, associated vascular lesions, and markers of hemorrhage expansion is possible with routine protocols akin to diagnostic strategies for ischemic stroke. Imaging applications for hemorrhagic transformation, venous thrombosis, and microvascular disorders are considered with a perspective that balances concern for hemorrhage with prevention of ischemia as these processes are often intertwined and clinical conundrums arise. Imminent imaging advances are anticipated with increased use of detailed imaging for hemorrhage and overlap with cerebral ischemia. Numerous questions abound regarding optimal management of hemorrhage and definitive treatments are lacking, yet imaging of pivotal pathophysiology offers tremendous opportunity for future progress in combating this debilitating condition.
    Annals of Neurology 11/2010; 68(5):581-92. DOI:10.1002/ana.22210 · 11.91 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Silent brain infarctions (SBIs), leukoaraiosis (LA), and microbleeds (MBs) are ischaemic silent radiologic abnormalities that act as predictors of subsequent strokes. This study investigated the independent effect of silent radiologic abnormalities on initial stroke severity and short-term outcome. A consecutive series of patients who had their first ischaemic stroke within 72 h of symptom onset were included. Demographic and clinical characteristics were collected on admission, and magnetic resonance imaging was performed to evaluate the ischaemic lesion, SBI, LA, and MB. Factors potentially associated with lower initial stroke severity (admission NIH Stroke Scale 0-5) and good short-term outcome (discharge NIH Stroke Scale 0-5, modified Rankin Scale 0-1) were validated by multivariate analysis. Silent brain infarctions were noted in 82 (45%) of the 182 patients. Although there were no statistically significant differences in stroke subtypes and lesion location, univariate analysis revealed that patients with SBI had reduced stroke severity (P = 0.005) and infarction volume (P = 0.001). After adjusting for covariates, the presence of SBI was independently associated with lower stroke severity and good short-term outcome when the NIH Stroke Scale was used as dependent variable (OR 3.368, 95% CI 1.361-8.332, P = 0.009; OR 3.459, 95% CI 1.227-9.755, P = 0.019, respectively). However, the presence of SBI lost significance when the discharge-modified Rankin Scale was used as dependent variable (P = 0.058). Amongst silent radiologic abnormalities, SBI was the only predictor of reduced stroke severity and infarct volume. Silent brain infarction deserves more attention in evaluating stroke severity.
    European Journal of Neurology 12/2010; 18(7):962-71. DOI:10.1111/j.1468-1331.2010.03282.x · 4.06 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Chronic kidney disease (CKD) is associated with both a risk of adverse vascular outcome and a risk of bleeding. We have tested the hypothesis that in the setting of an acute intracerebral hemorrhage (ICH), CKD is associated with poor outcome and with larger hematoma volume. We examined the association between CKD and ICH characteristics and outcome within a prospective cohort study of consecutive patients hospitalized with an acute stroke and followed for 1 year. CKD was categorized by the estimated baseline glomerular filtration rate into moderate/severe impairment (<45), mild impairment (45-60) and no impairment (>60 ml/min/1.73 m(2)). Among 128 patients with an ICH (mean age = 71.7 ± 12.3 years, 41.4% women) 46.1% had CKD (23.4% mild and 22.7% moderate/severe). Patients with moderate/severe impairment had >4-fold adjusted hazard ratio for mortality over 1 year (4.29; 95% CI = 1.69-10.90) compared to patients with no impairment. The hematoma volumes [median (25-75%)] were 15.3 ml (5.4-37.5) in patients with no impairment, 16.6 (6.8-36.9) in mild impairment and 50.2 (10.4-109.1) in moderate/severe impairment (p = 0.009). The location of the hematoma was lobar in 12% with no impairment, 17% with mild impairment and 39% with moderate/severe impairment (p = 0.02). Patients with moderate/severe impairment exhibited a 2.3-fold higher hematoma volume (p = 0.04) and a >6-fold higher odds of lobar location (95% CI = 1.59-24.02) as compared to no impairment. Further adjustment for antiplatelet use and for presence of leukoaraiosis attenuated the association with hematoma volume (p = 0.15), while moderate/severe impairment was associated with an adjusted OR of 5.35 (95% CI = 1.18-24.14) for lobar location. Presence of moderate/severe CKD among patients with ICH is associated with larger, lobar hematomas and with poor outcome.
    Cerebrovascular Diseases 02/2011; 31(3):271-7. DOI:10.1159/000322155 · 3.70 Impact Factor
Show more