| I. Sensitivity of different sequences of MRI for the diagnosis of CMB and follow-up study of number of CMBObjectiveTo compare the sensitivity of different sequences of MRI for the diagnosis of cerebral microbleed, including SWI(susceptibility-weighted imaging, SWI),DWI,T1WI,T2WI,Flair. and GRE-T2*WI(gradient echo-T2*weighted imaging, GRE-T2* WI), and to observe the number of CMB in 47 cases with CMB at 1 year's follow-up.Patients and methodsEighty three cases with acute cerebral vascular diserses were performed different sequences of MRI, including EPI-SWI(susceptibility-weighted imaging, SWI),DWI,T1WI,T2WI,Flair and GRE-T2*WI(gradient echo-T2*weighted imaging, GRE-T2*WI). A 1.5T-tesla MR machine (GE Corportion) was used to obtain the following sequences by using a 8-channel Total Imaging Matrix head coil: conventional axial DWI,T1WI,T2WI and Flair, T2*-weighted gradient echo(TR/TE=800/26ms,20-degree flip angle,5mm slice thickness, and 1.5 mm gap), and SWI (TR/TE= 57/40ms,5mm slice thickness, no gap, visual fields 24cm×24cm, 8min). All MRI scans were read by 2 neurologists and one neuroradiologist independent raters blinded to clinical history who determined the presence, number, and location of CMBs. The number of CMB in 47 cases with CMB was compared at one year'follow-up.ResultsCMBs in each sequence appeared as round or oval hypointensities with diameter 2-5mm. Of eighty three patients with acute cerebral vascular diseases, the detection sensitivity of CMB was the highest with SWI (60.2%), and the detection sensitivity of CMB was the second highest with GRE (57.3%), while the detection sensitivity of CMB was lower with other convential sequences of T1WI,T2WI,Flair and DWI. The detection sensitivity of CMB with convential sequences was less one third than that with GRE or SWI. The number of CMB with convential sequences were less than that with GRE (P<0.001)The CMBs most frequently detected by GRE were in the basal ganglia, followed by cortex-subcortical region, brain stem and cerebellum area.At one year's follow-up the number of CMB in 47 cases with CMB was compared and most CMB lesions at first scan did not absorb up or disappear but increased. The average number by first scan was 10.62±16.04 while the average number of 1 year later was 11.94±18.31 with a statistically difference (P=0.001). The results showed the microarteriopathy of CMB still developed and silent old hemosiderin cannot be absorbed up, transported and eliminated out. Ⅱ.Prevalence and related risk factors of cerebral microbleeds among different cerebral vascular diseasesObjectiveTo study the prevalence and related risk factors of cerebral microbleeds among different cerebral vascular disease, and to investigate the clinical significance of cerebral microbleeds.Patients and methodsConsecutive 393 in-patients with acute cerebral vascular diseases and 146 control cases were included and related risk factors were registered.348 cases of infarction,25 cases of hemorrhage,20 cases of TIA and 146 control cases were performed gradient echo-T2* weighted imaging.A 1.5T-tesla MR machine (GE Corportion) was used to obtainthe following sequences:axial T2*-weighted gradient echoTR/TE=800/26ms,20-degree flip angle, 5mm slice thickness, and 1.5 mm gap.All MRI scans were read by 2 neurologists and one neuroradiologist independent raters blinded to clinical history who determined the presence, number, and location of CMBs. All risk factors were registered, including sex, age, hypertension, diabetes, addiction of smoking, addiction of alcohol, leukoencephalopathy; atherosclerosis plague of cartotid artery, level of serum lipoprotein, the use of aspine, atrial fibrillation, old silent lacunar infarction, infarction and hemorrhage.Logistic regression analysis was used to assess the relationship of cerebral microbleed with the following variables:age, sex, hypertension, diabetes, addiction of smoking, addiction of alcohol, leukoencephalopathy; atherosclerosis plague of cartotid artery, level of serum lipoprotein, the use of aspine, old silent lacunar infarction, atrial fibrillation, infarction and hemorrhage by enter full variate anaysis. ResultsThe prevalence of cerebral microbleeds in cerebral hemorrhage, cerebral infarction, transient ischemic attack and leukoencephalopathy were 84.0%,45.98%, 25.0% and 63.13% and control was 13.70%. The results of binary logistic regression showed that risk factors of cerebral microbleeds were silent old lacunar infarct (OR=2.350,95% CI 1.426-3.874, P=0.001), hypertension (OR=2.146,95% CI 1.224-3.761, P=0.008), leukoencephalopathy (OR=5.133,95% CI3.121-8.442, P=0.000), infarction (OR=2.206,95% CI 1.175-4.143, P=0.014), hemorrhage (OR =17.751,95% CI 4.781-65.902, P=0.000) and hypercholesterolemia(OR=1.645,95% CI 1.123-2.410, P=0.011). The protective factor was low density lipoprotein (OR=0.516,95% CI 0.336-.794,P=0.003)Ⅲ. Clinical study of cerebral microbleeds among different subtypes of acute ischemic infarctionObjectiveTo study the prevalence and grade of cerebral microbleeds among different subtypes of acute ischemic infarction, and to investigate the clinical significance of cerebral microbleeds.MethodsConsecutive 368 patients with acute ischemic infarct were included and classified according stroke subtypes into atherothrombotic infarction (n=213), small artery disease infarction (n=88), cardioembolic infarction (n=28), undetermined cause infarction (n=19) and transient ischemia attack (n=20). One hundred and forty six patients without cerebral vascular diseases were served as controls. The baseline data were registered and all patients were performed gradient echo-T2* weighted imaging(GRE-T2*WI). The prevalence and grade of microbleeds among different subtypes of stroke were compared. The prevalence rate among the first infarction and recurrent infarction was also analysed.All the data were expressed as mean±values SD for continuous variable. Groups comparison statistical analysis was performed with the ANOVA and comparison between study group and control group were performed by Dunnett-t test. The discontinuous variance was expressed as percentage, group comparison for nonparametric analysis was performed withχ2 test, Kruskal-Wallis H test and Fisher exact test. P<0.05 was considered statistically significant.ResultsAccording to the guidline of modified TOAST criteria of acute cerebral vascular diseases, The prevalence of cerebral microbleeds in different infarction subtypes were higher statistically compared to control(P<0.05). The prevalence of small artery infarction, atherothrombotic infarction, cardioembotic infarction and undetermined etiology infarction were respectively 53.4%,46.0%.28.6% and 36.8%. The prevalenct of TIA group(25.0%) was a little higher than the control (13.7%) without statistically difference. The grade of CMB between different subtypes infarction was compared by K Independents Samples(χ2=282.782, P=0.00) with a statistically difference. The severe grade of CMB of small artery infarction was higher than the other subtypes.The prevalence of cerebral microbleeds in recurrent stroke was higher compared to primary stroke and atherothrombotic infarction group existed statistically different. IV. Relation between hemorrhagic transformation after acute ischemic infarction and cerebral microbleeds and analysis of related risk factors of hemorrhagic thansformationobjectiveTo investigate the relation between hemorrhagic transformation after acute ischemic infarction and cerebral microbleeds and related risk factors of HT, and to analyse possible preventive methods of HT.Patients and methodsConsecutive 348 in-patients with acute cerebral vascular diseases were included and related risk factors were registered, including sex, age, hypertension, diabetes, addiction of smoking, addiction of alcohol, leukoencephalopathy; atherosclerosis plague of cartotid artery, level of serum lipoprotein, the use of aspine, intravenous urokinase thrombolysis, the score of NIHSS and atrial fibrillation. All patients were performed GRE-T2*WI to judge the presence of CMB and hemorrhage transformation. Logistic regression analysis was used to assess the relationship of hemorrhage transformation with the following variables:age, sex, hypertension, diabetes, addiction of smoking, addiction of alcohol, leukoencephalopathy; atherosclerosis plague of cartotid artery, level of serum lipoprotein, the use of aspine, intravenous urokinase thrombolysis, the score of NIHSS and atrial fibrillation by Forward Stepwise (Likelihood Ratio).ResultsHemorrhage transformation occurred in 35 of 348 patients with acute ischemic infarction(10.06%). Ninteen patients experienced HT among 213 cases with atherosclerotic infarction (8.92%). Eleven patients experienced HT among 28 cases with cardioembolic infarction (39.29%) and 5 patients experienced HT among 19 cases with undetermined etiology infarction (26.32%). No case experienced HT among lacunar infarction.The results of univariate binary logistic regression showed that risk factors of HT were the score of NIHSS, atrial fibrillation, intrevous urokinase thrombolysis, protective factors were cerebral microbleeds, leukoencephalopathy, hypercholesterolemia, high density lipoprotein and low density lipoprotein.The results of multivariate binary logistic regression showed that risk factors of HT were cardioembolic infarction (OR-5.338,95% CI 1.926-14.796, P=0.001) undetermined etiology infarction (OR=6.843,95% CI 1.848-25.346, P=0.004), the score of NIHSS (OR=1.181,95% CI 1.085-1.284, P=0.000), diabetes (OR=3.694, 95% CI 1.456-9.371, P=0.006), the protective factor was low density lipoprotein (OR=0.637,95% CI 0.414-0.979, P=0.040)Conclusions1. SWI or GRE imaging can sensitively detect the presence of CMB and can be taken as golden criteria sequences, while other conventional sequences cannot sensitively detect CMB. The number of CMB by GRE or SWI was more than that by other conventional sequences (P=0.000)2. The CMBs most frequently detected by GRE were in the basal ganglia, followed by cortex-subcortical region, brain stem and cerebellum area.3. At one year's follow-up the number of CMB of 47 cases with CMB were compared and most CMB lesions at first scan did not absorb up or disappear but increased. The average number by first scan was 10.62±16.04 while the average number of 1 year later was 11.94±18.31 with a statistically difference (P=0.001). The results show the microarteriopathy of CMB still has developed and silent old hemosiderin cannot be absorbed up, transported and eliminated out.4. The prevalence of acute cerebrovascular diseases goes higher, varies differently according to different subtypes. The prevalence of cerebral microbleeds among hemorrhage, infarction, TIA and leukoencephylopathy are respectively 84.0%, 45.98%,25.0% and 63.13%, and the control is 13.70%.5. Multivariates Logistic regression analysis shows related risk factors of cerebral microbleeds are silent old lacunar infarct, hypertension, infarction, hemorrhage hypercholesterolemia and leukoencephapathy. Low density lipoprotein is the protective factor.6. The prevalence of cerebral microbleeds among different subtypes of infarctions is higher and small artery infarction is the highest (53.4%). The grade of CMB among small artery infarction is more severe.7. The prevalence of cerebral microbleeds in recurrent stroke is higher compared to primary stroke, and the relationship between the relapse of the cerebral infarct and the presence of cerebral microbleeds maybe exist and will be further studied in the future.8. Herrohage transformation occurred in 35 of 348 patients with acute ischemic infarction(10.06%). Eleven patients experienced HT among 28 cases (39.29%) with cardioembolic infarction with highest rate of HT.9. The results of multivariate binary logistic regression show that risk factors of hemorrhage transform are cardioembolic infarction, undetermined etiology infarction, the score of NIHSS and diabetes, the protective factor is low density lipoprotein. CMB doesnot correlate with HT.
|
PDF Full Text Request