Jonathan Graff Radford

INCIDENCE OF CEREBRAL MICROBLEEDS AND AMYLOID BURDEN: THE MAYO CLINIC STUDY OF AGING

performance on memory tasks before and after treatment with antibodies that reduce the incidence of capillary plugging. Results:About 1.5% of cortical capillaries were transiently stalled in AD mice, leading to an overall reduction in cerebral blood flow (CBF). These stalls were released using an antibody against the neutrophil cell surface protein Ly6G, resulting in an immediate w25% increase in CBF and an immediate improvement in performance on short-term memory tasks. Mice treated with an isotype control antibody showed no improvement in CBF or memory performance. In a second set of experiments we deterimend that reducing capillary stalls and increasing CBF led to improved cognitive performance in mice with relatively advanced AD pathology (17 month old APP/PS1 mice, first cognitive impacts detectible at 8 months). Conclusions: In this study we uncovered leukocyte adhesion in brain capillaries as a mechanism contributing to reduced CBF in AD mouse models and showed that blocking this adhesion leads to immediate cognitive benefits even in advanced stages of disease development.

CHARACTERIZATION OF THE CLINICAL AND NEUROIMAGING FEATURES OF PATIENTS WITH DEMENTIA WITH LEWY BODIES ASSOCIATED WITH VARIANTS IN GLUCOCEREBROSIDASE (GBA)

Background:A significant portion of patients with early-onset Alzheimer’s disease (age of onset <65 years) have non-amnestic presentations, such as logopenic primary progressive aphasia (lvPPA), posterior cortical atrophy (PCA), and the behavioral/executive variant. The NIA-AA criteria for probable AD recognize language, visuospatial, and executive variants but do not specify other presentations for AD.Methods:We reviewed 220 patients with clinically diagnosed early-onset AD byNIA-AA criteria from a subspecialty university clinic. All patients had biomarker-supported clinically probably AD either by fluorodeoxyglucose positron emission tomography (FDG-PET) or cerebrospinal fluid b-amyloid or tau biomarkers. The presenting symptoms and clinical and neuroimaging characteristics were reviewed. Results:We identified 15 (6M/9F) right-handed patients with AD who presented with predominant non-aphasic, non-visuospatial symptoms referable to the left parietal lobe. All lacked family history for an autosomal dominant disorder. Their ages of presentation and onset were 57.82 (8.9) and 55.11 (5.51), respectively. Mini-Mental State at onset was 23.1 (3.67) with retrospective Clinical Dementia Rating Scores of 1.23 (0.53). The predominant presenting symptoms were decreased writing (lexical and transitional motor) (9) decreased manual dexterity (3), and acalculia (5). On examination, the agraphia was usually lexical plus transitional motor; 12 had limb apraxia (10 ideomotor and 6 limb-kinetic); the acalculia was an anarithmetia; and 3 had the full Gerstmann syndrome. Neuroimaging, including FDG-PET was consistent with prominent dysfunction of the left parietal lobe. Conclusions: Patients with earlyonset AD may present with variants such as lvPPA and PCA, and it is now possible to diagnose them with clinical AD using NIA-AA criteria. However, other variant presentations of earlyonset AD need to be considered, including the combination of symptoms referable to disproportionate involvement of the left parietal lobe.

Systems fail before molecules spread: A cascading failure model of Alzheimer’s disease

Triangulated surfaces were geneterated to contour the most likely voxels for each structure. This atlas was mapped diffeomorphically onto each scan, and atrophy in aMCI patients relative to controls was quantified through local change in surface area, a technique known as surface diffeomorphometry. Regions with significant differences between groups were identified using permutation testing, controlling for false positives at 5% familywise error rate. Results: Anatomical changes were detected, including atrophy in the transentorhinal / lateral-entorhinal region (p < 1e-6). Conclusions:This atlas based method for probabilistically identifying topographic changes in the MTL identified regions of atrophy consistent with Braak staging for early Alzheimers disease. This method may also be useful for following the longitudinal progression of disease in the context of a clinical trial.

Predicting survival in dementia with lewy bodies: The importance of hippocampal volume

Background:While Dementia with Lewy bodies (DLB) is associated with shorter survival compared to Alzheimer disease (AD) dementia, antemortem prognosis of disease duration remains problematic. Hippocampal volume is associated with Alzheimer’s disease-related neurofibrillary tangle pathology in DLB patients. Our objective was to determine whether smaller hippocampal volumes are associated with a shorter survival among DLB patients. Methods:Survival analysis for time from onset of cognitive symptoms to death was carried out using Cox proportional hazards models. These analyses used data from 87 consecutive patients with DLB from the Mayo Clinic Alzheimer’s Disease Research Center. All patients underwent 3 Tesla MRI with hippocampal volumetry. Hippocampal volumes were dichotomized into normal and abnormal. The models were adjusted for APOE e4 status, estimated onset age of cognitive problems, and age at MRI. Results: The median age at MRI was 71 years (interquartile range (IQR): 66, 77), and 83% were male. The median time from estimated first cognitive symptom to death was 7.04 (IQR: 5.23, 9.59) Abnormal hippocampal volumes were present in 31 (36%) of the patients. Abnormal hippocampal volumes were significantly associated with higher risk of death (Hazard ratio 2.87 (95% CI 1.40-5.87, p1⁄40.004)). Conclusions: Among patients with clinically diagnosed DLB, those with neuroimaging evidence of hippocampal atrophy have shorter survival times. Hippocampal volumes may improve our ability to predict survival in DLB patients. Therefore among DLB patients, concomitant AD pathology may be associated with a shorter duration of illness, and a subset of DLB patients may benefit from therapeutics targeted at AD.

Intravascular mucinosis: a rare cause of cerebral infarction.

Cerebral infarcts are not infrequent in cancer patients. Of 3,426 cancer patients, 500 (14.6%) had evidence of cerebrovascular disorders at autopsy, 256 (7.5%) representing infarcts [4]. Among rare causes, intravascular mucinosis (IM), the extensive plugging of microvasculature by mucin, occurs in mucin-producing adenocarcinomas and is typically diagnosed postmortem with extensive hemorrhagic infarcts in border-zone distribution [2–4, 8, 9, 12]. We report IM in a 63and a 48-year-old woman, treated for ER/PR positive lobular cancer [AJCC T1 and T2, N0], 19 and 2 years earlier, respectively. Both presented with headache and unusual imaging findings (Fig. 1), raising a broad differential diagnosis including vasculitis and posterior reversible encephalopathy. A left temporal lobe biopsy disclosed IM (Fig. 2) with rare neoplastic cells in Case 1. Both patients progressively deteriorated and expired after 28 and 26 days. Gross and microscopic autopsy findings are illustrated in Figs. 3 and 4. Spinal cord changes, predominantly in the posterior columns, involved the lower cervical-upper thoracic (Case 1) and upper thoracic (Case 2) region. General autopsy disclosed extensive metastatic adenocarcinoma with signet ring features and abundant mucin, involving liver (both cases), lung (Case 1), vertebrae and myocardium (Case 2). A small patent foramen ovale (0.6 cm) was noted in Case 1. Tumor microemboli trapped in distal vascular branches at border-zones may contribute to the pathogenesis of IM [11]. In Case 1, they could have entered the arterial circulation directly from the lung metastases or paradoxically through the patent foramen ovale; and in Case 2, from bone marrow metastases in a fashion similar to fat embolism. Some authors, having found mucin admixed with embolic fat droplets (confirmed by Oil Red O stain) hypothesized that bone marrow and vascular wall necrosis due to radiation, chemotherapy or the tumor itself, could have released mucin and fat droplets into veins, maintained open by the trabecular bone architecture [2]. Similarly, we noted ‘‘clear globules’’ (Fig. 4a, c), but could not confirm their fat embolic nature, since frozen tissue was not available. Tumor cells, mucin and fat globules in the bloodstream could overwhelm the trapping capacity of the pulmonary capillaries and penetrate the arterial circulation [6]. Circulating cells, distributed according to organ blood flow (approximately 20% to the brain) may lodge in major end artery terminations and at the gray–white matter junction, and continue to secrete mucin. Circulating mucin may produce intermittent occlusive events and cause hemorrhagic infarcts by repeated occlusion and reperfusion [1, 5]. Spinal cord border-zones follow a vertical distribution with vulnerability in the thoracic region, as few small radicular branches supply this cord segment [10], and a horizontal distribution, present between anterior and posterior spinal artery territories [6]. In both our cases, hemorrhagic infarcts involved the lower cervical and upper thoracic posterior columns. While this follows a vertical distribution, the posterior column by itself is rarely involved in episodes of ischemia, being richly supplied by posterior radiculo-medullary arteries [7]. However, more tumor cells in the direction of blood flow M. C. Bernardo J. Menke B. Scheithauer C. Giannini (&) Department of Laboratory Medicine and Pathology, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA e-mail: giannini.caterina@mayo.edu