Differences in lobar microbleed topography in cerebral amyloid angiopathy and hypertensive arteriopathy | Scientific … – Nature.com

Patient selection

We prospectively recruited patients who had suffered symptomatic spontaneous ICH for brain MRI and 11C-Pittsburgh compound B (PiB) PET scans at NTUH between September 2014 and October 2022 (Fig.1)11,12. We excluded patients with potential causes of secondary hemorrhage, including trauma, structural or vascular lesions, brain tumors, severe coagulopathy due to systemic disease or medication, or patients who suffered ischemic stroke with hemorrhagic transformation. Patients were excluded if they could not tolerate imaging studies, including patients with a poor ability to cooperate, hemodynamic instability, or an implanted cardiac pacemaker. A total of 151 patients with ICH fulfilling the enrollment criteria agreed to participate in this study and received brain MRI and PiB PET scans (Fig.1). We excluded patients who did not have any lobar CMBs (n=45) or only had lobar CMBs adjacent to the previous hematoma (n=1). Patients for whom the imaging quality was suboptimal were also excluded (n=14). Thus, a final sample of 91 patients were included in this analysis (Fig.1). As previously described, patients with lobar ICH(s) involving the cerebral cortex and underlying white matter with strictly lobar CMBs and/or cSS were defined as having CAA (n=30) according to the Boston criteria 1.54,5, while patients were defined as having HA (n=61) if the ICH and CMBs were located in both the lobar and deep regions, as we previously proposed11,12. Baseline clinical data collection was performed by the investigators through a comprehensive review of medical records and interviewing each participant. The following clinical variables were systematically recorded for each patient: age, sex, presence of chronic hypertension (defined as clinical diagnosis of hypertension with more than 3months of prescription of anti-hypertensive agents), classes of anti-hypertensive medication, diabetes mellitus, hypercholesterolemia, history of ICH and ischemic stroke, and creatinine clearance value (represented by estimated glomerular filtration rate).

Flowchart of patient enrollment. Of the 151 survivors of spontaneous intracerebral hemorrhage (ICH) who agreed to participate in the study, we excluded patients without any lobar cerebral microbleeds (CMBs; n=45), patients for whom image quality was suboptimal (n=14), and patients with CMBs only close to the hematoma (n=1). In total, 30 patients with cerebral amyloid angiopathy (CAA)-related ICH (defined as strictly lobar ICH and/or CMBs) and 61 patients with non-CAA ICH (mixed deep and lobar ICH/CMBs) were included in the analysis.

Brain MRIs were obtained using a 3-Tesla scanner (Siemens Verio, TIM, or mMR, Siemens Medical Solutions, Malvern, PA, USA). The imaging protocols included T1-weighted imaging, T2-weighted imaging, fluid-attenuated inversion recovery imaging, susceptibility weighted imaging (SWI), diffusion-weighted imaging, apparent diffusion coefficient mapping, and 3D T1-weighted MPRAGE (Magnetization Prepared Rapid Acquisition Gradient Echo) imaging in 1-mm-slice thicknesses. CMBs were defined as lesions with homogeneous round signal loss and a diameter less than 10mm on SWI and were categorized as lobar or deep based on well-validated criteria13,14.

Each lobar CMB was first identified by reviewing SWI sequences. However, due to the limitations of SWI with respect to anatomical discrimination, lobar CMBs were further categorized as intracortical, juxtacortical, or subcortical based on their relative location to the cortex using the corresponding T1-weighted multiplanar reconstruction images, as shown in Fig.2. Intracortical CMBs were defined as CMBs completely located in the gray matter of the cortex (Fig.2A); juxtacortical CMBs were defined as CMBs located on the border of gray-white junctions (Fig.2B); subcortical CMBs were defined as lobar CMBs located in the white matter without reaching the cortex (Fig.2C). Strictly intracortical/juxtacortical lobar CMBs were defined as having no lobar CMB in the subcortical white matter. All MRI scans were independently rated by two investigators (P.-Y. K., 3-year reading experience and P.-T. C., 6-year reading experience) to determine the inter-rater reliability for the presence/absence of intracortical and subcortical CMBs. If there was disagreement between the two readers, the same investigators reached a consensus decision after discussion.

Categorization of lobar CMBs. The location of each lobar CMB was categorized using SWI and the corresponding 3D T1-weighted images. (A) Intracortical CMB: the lobar CMB is completely located in the gray matter of the cortex. (B) Juxtacortical CMB: the CMB is located on the border of a gray-white matter junction. (C) Subcortical CMB: the CMB is completely located in white matter. The borders of the cortex are outlined with dotted lines.

Other MRI markers related to cerebral SVD were evaluated based on the Standards for Reporting Vascular Changes on Neuroimaging criteria8,14,15. Briefly, the presence and number of CMBs and cortical superficial siderosis were evaluated on axial SWI sequences, as previously described13,16. The number of CMBs was calculated in the lobar (i.e., the frontal, temporal, parietal, occipital, and insular cortices) and deep regions (i.e., the brainstem, BG, thalamus, internal capsule, external capsule, corpus callosum, and deep periventricular white matter). Cerebellar CMBs were not evaluated in the current study. Lacunes were evaluated in the supratentorial region and defined as ovoid or round, subcortical, fluid-filled cavities ranging in diameter from 3 to 15mm17,18. WMH volume was calculated based on fluid-attenuated inversion recovery imaging using a semi-automated measure, as we previously described8,19. The volume estimates were performed in the ICH-free hemisphere and multiplied by two. MRI-visible perivascular spaces (PVS) were evaluated on T2-weighted imaging and defined as sharply delineated structures measuring<3mm following the course of perforating or medullary vessels20. The number of PVS were counted in the centrum semiovale (CSO) and basal ganglia (BG) on the side of the brain with more severe involvement. The severity of PVS was rated using a validated visual scale (0=none, 1=<10, 2=1120, 3=2140, and 4=>40)20,21. According to a previously proposed method, we pre-specified a dichotomized classification of high-degree (scale, 3 and 4) and low-degree (scale, 02) PVS20.

PiB was manufactured and handled at the PET center, NTUH, Taipei, Taiwan (specificity activity: 3919GBq/mol). All PET scanning was performed within 3months after acquisition of the MRI. Static PET/CT scans (discovery ST; GE Healthcare, Waukesha, WI) were acquired in three-dimensional mode for 30min starting 40min after the injection of 10mCi 11C-PiB. PET data were reconstructed with ordered set expectation maximization (five iterations; 32 subsets; post filter, 2.57) and corrected for attenuation. Each PiB PET image was realigned, resliced, and manually co-registered to a standardized CT template using PMOD software, as previously described8,15,19. The PET data were semi-quantitatively analyzed and expressed as the average mean standardized uptake value ratios (SUVRs) of the regions of interest using the cerebellar cortex as a reference region. The regions of interest in these spatially normalized images included the frontal, temporal, parietal, and occipital lobes, as defined in the Automated Anatomical Labeling Atlas. Areas of macrobleeds were manually excluded from the SUVR analyses, and the parameters in the specific regions of interest were determined using the ICH-free hemisphere.

Categorical variables are presented as percentages and continuous variables are presented as meanSD or median (interquartile range) based on their distribution. Baseline demographics, clinical, and neuroimaging variables were compared between patients with CAA and HA using the MannWhitney U-test for continuous variables and Fishers exact test for categorical variables. We built multivariable logistic regression models to search for independent associations between CAA/HA and the lobar CMB topography (strictly intracortical/juxtacortical CMBs or subcortical CMBs); model 1 was adjusted for age and sex and model 2 was adjusted for age, sex, and other SVD neuroimaging markers. Additionally, we used the Youden index to determine best cutoff values for the number of lobar CMBs in each category (intracortical, intracortical/juxtacortical, subcortical) to differentiate CAA and HA. The diagnostic value, including sensitivity, specificity, positive predictive value, negative predictive value and the area under curve (AUC), was determined.

To further confirm our hypothesis that CAA is more frequently associated with intracortical lobar CMB topography, we compared the global and regional PiB SUVRs between patients with and without intracortical CMBs and between patients with and without subcortical CMBs using the MannWhitney U-test. The correlations between amyloid retention and the numbers of intracortical or subcortical CMBs were investigated using Spearmans correlation and partial correlation analysis for age adjustment. All statistical analyses were performed using SPSS version 19 (IBM Corp., Armonk, NY, USA). All tests of significance were two-tailed with the threshold for significance defined as P<0.05.

This study was performed with the approval of the institutional review board (201903069RINB) of National Taiwan University Hospital (NTUH) and in accordance with their guidelines. Written informed consent was obtained from all patients or their families in this study.

Visit link:
Differences in lobar microbleed topography in cerebral amyloid angiopathy and hypertensive arteriopathy | Scientific ... - Nature.com