Clinical Imaging Of Intraventricular Hemorrhage Following Intracerebral Hemorrhage

PART ONE THE SLICE SCORE:A NOVEL SCALE MEASURING INTRAVENTRICULAR HEMORRHAGE SEVERITY AND PREDICTING POOR OUTCOME FOLLOWING INTRACEREBRAL HEMORRHAGEBackground and Purpose Nearly half of patients with spontaneous intracerebral hemorrhage(ICH),intraventricular extension of the hematoma occurs as a lethal complication leading to worse outcome and neurological deficits.The dynamics of IVH extent plays an important role in the course of ICH patients with IVH,were an independent predictor of poor outcome with increased mortality.To quantify extent of intraventricular hemorrhage with a novel,simple IVH severity score named Slice score,and to explore and compare its performance in predicting worse outcomes in ICH.Materials and Methods We retrospectively analyzed 652 scans from 326 subjects.A new scoring system for IVH severity was proposed and termed Slice score.The correlations between measured IVH volume and Slice score,original Graeb,LeRoux,and IVH score(IVHS)were compared.The association between these scores and clinical outcomes were evaluated using logistic regression.We then identified clinical thresholds of Slice score by balancing the probability of prediction and accuracy.Primary outcome was defined as90-day poor outcome(modified Rankin score≥4)and secondary outcome was 90-day mortality.Results Of 326 ICH patients,122(37.4%)had poor outcome and 59(18.1%)died at 3 months.The Slice score showed the highest correlation with measured IVH volume(R=0.73,R~2=0.54,P<0.001).The observed area under the curve were similar among the Slice,original Graeb,LeRoux score,and IVH score for poor outcome(0.633,0.633,0.632,0.634,respectively),and for mortality(0.660,0.660,0.660,0.656,respectively).All IVH scales were independently associated with 90-day poor outcome and mortality with close odds ratio in adjusted models(all odds ratio>1.07,all P<0.05).Multivariate Analyses of categorized Slice score revealed optimal thresholds of 6 and 12 for primary and secondary outcomes(odds ratio 4.20,95%confidence interval 1.82-10.02,P=0.001;odds ratio 5.41,95%confidence interval 1.66-17.43,P=0.005,respectively).Conclusions The Slice score correlated highly with the IVH volume,was a reliable scale for measuring IVH severity and could be an easy-to-use tool for predicting 90-day poor outcome and mortality in ICH.PART TWO COMBINATION OF NONCONTRAST COMPUTED TOMOGRAPHIC BLEND SIGN AND BLACK HOLE SIGN IMPROVES PREDICTIONS OF HEMATOMA EXPANSION AND POOR OUTCOME AFTER INTRACEREBRAL HEMORRHAGEBackground and Purpose Prediction of hematoma expansion(HE),which occurs in≈30% ICH patients,is of significance to direct intensive therapy in potential HE-prone patients for less death.Reliable neuroimaging predictors of hematoma enlargement have been validated in several studies.Of which novel Noncontrast computed tomographic Blend sign(BS)and Black hole sign(BHS)have been validated separately to predict hematoma expansion(HE)in intracerebral hemorrhage(ICH).However,the integrated impact of these markers on neurological deterioration remains unaddressed.Here we aimed to demonstrate a strategy combining BS and BHS to improve risk stratification in HE and poor outcome.Materials and Methods We retrospectively analyzed 277 ICH patients from July 2011 to December 2016.Hematoma expansion was defined hemorrhage growth >6ml or >33%.Poor outcome(modified Rankin Scale score ≥3)was assessed at 3 months.The presence of BS and BHS were analyzed interactively with four potential combinations: BS and BHS negative(BS-BHS-),BS and BHS positive(BS+BHS+),BS positive with and without BHS positive(BS+BHS-,BS-BHS+).Results Of patients with HE,BS+BHS+ accounts for 10.2%,BS-BHS+ for23.9%,BS+BHS-for 28.4% and BS-BHS-for 37.5%(all P<0.001).For poor outcome,the rate of 4 interactive combinations were 7.0%,16.8%,18.2% and 58.0%,respectively(all P<0.001).Multivariate analysis confirmed BS+BHS+(OR=37.24,P=0.002)as the highest risk factor in HE prediction when compared with BS-BHS+(OR=9.49,P<0.001)and BS+BHS-(OR=15.50,P<0.001).Firth penalized regression demonstrated BS-BHS+ and BS+BHS-as independent predictors for poor outcome with close risk levels(OR=5.28,P=0.004;OR=6.91,P<0.001,respectively).Conclusions The combination of BS and BHS especially BS+BHS+ stratify a high-risk ICH population with potential HE and refine estimates of poor prognosis,is meaningful in clinical use and research.