Glioma subtype classification from histopathological images using in-domain and out-of-domain transfer learning: An experimental study.

We provide in this paper a comprehensive comparison of various transfer learning strategies and deep learning architectures for computer-aided classification of adult-type diffuse gliomas. We evaluate the generalizability of out-of-domain ImageNet representations for a target domain of histopathological images, and study the impact of in-domain adaptation using self-supervised and multi-task learning approaches for pretraining the models using the medium-to-large scale datasets of histopathological images. A semi-supervised learning approach is furthermore proposed, where the fine-tuned models are utilized to predict the labels of unannotated regions of the whole slide images (WSI). The models are subsequently retrained using the ground-truth labels and weak labels determined in the previous step, providing superior performance in comparison to standard in-domain transfer learning with balanced accuracy of 96.91% and F1-score 97.07%, and minimizing the pathologist's efforts for annotation. Finally, we provide a visualization tool working at WSI level which generates heatmaps that highlight tumor areas; thus, providing insights to pathologists concerning the most informative parts of the WSI.

Computational Assessment of Spectral Heterogeneity within Fresh Glioblastoma Tissue Using Raman Spectroscopy and Machine Learning Algorithms.

Understanding and classifying inherent tumor heterogeneity is a multimodal approach, which can be undertaken at the genetic, biochemical, or morphological level, among others. Optical spectral methods such as Raman spectroscopy aim at rapid and non-destructive tissue analysis, where each spectrum generated reflects the individual molecular composition of an examined spot within a (heterogenous) tissue sample. Using a combination of supervised and unsupervised machine learning methods as well as a solid database of Raman spectra of native glioblastoma samples, we succeed not only in distinguishing explicit tumor areas-vital tumor tissue and necrotic tumor tissue can correctly be predicted with an accuracy of 76%-but also in determining and classifying different spectral entities within the histomorphologically distinct class of vital tumor tissue. Measurements of non-pathological, autoptic brain tissue hereby serve as a healthy control since their respective spectroscopic properties form an individual and reproducible cluster within the spectral heterogeneity of a vital tumor sample. The demonstrated decipherment of a spectral glioblastoma heterogeneity will be valuable, especially in the field of spectroscopically guided surgery to delineate tumor margins and to assist resection control.

The Brainbox-a tool to facilitate correlation of brain magnetic resonance imaging features to histopathology.

Magnetic resonance imaging (MRI) has limitations in identifying underlying tissue pathology, which is relevant for neurological diseases such as multiple sclerosis, stroke or brain tumours. However, there are no standardized methods for correlating MRI features with histopathology. Thus, here we aimed to develop and validate a tool that can facilitate the correlation of brain MRI features to corresponding histopathology. For this, we designed the Brainbox, a waterproof and MRI-compatible 3D printed container with an integrated 3D coordinate system. We used the Brainbox to acquire post-mortem ex vivo MRI of eight human brains, fresh and formalin-fixed, and correlated focal imaging features to histopathology using the built-in 3D coordinate system. With its built-in 3D coordinate system, the Brainbox allowed correlation of MRI features to corresponding tissue substrates. The Brainbox was used to correlate different MR image features of interest to the respective tissue substrate, including normal anatomical structures such as the hippocampus or perivascular spaces, as well as a lacunar stroke. Brain volume decreased upon fixation by 7% (P = 0.01). The Brainbox enabled degassing of specimens before scanning, reducing susceptibility artefacts and minimizing bulk motion during scanning. In conclusion, our proof-of-principle experiments demonstrate the usability of the Brainbox, which can contribute to improving the specificity of MRI and the standardization of the correlation between post-mortem ex vivo human brain MRI and histopathology. Brainboxes are available upon request from our institution.

Antiphospholipid antibodies are enriched post-acute COVID-19 but do not modulate the thrombotic risk.

BACKGROUND AND OBJECTIVES: COVID-19-associated coagulopathy, shown to increase the risk for the occurrence of thromboses and microthromboses, displays phenotypic features of the antiphospholipid syndrome (APS), a prototype antibody-mediated autoimmune disease. Several groups have reported elevated levels of criteria and non-criteria antiphospholipid antibodies (aPL), assumed to cause APS, during acute or post-acute COVID-19. However, disease heterogeneity of COVID-19 is accompanied by heterogeneity in molecular signatures, including aberrant cytokine profiles and an increased occurrence of autoantibodies. Moreover, little is known about the association between autoantibodies and the clinical events. Here, we first aim to characterise the antiphospholipid antibody, anti-SARS-CoV-2 antibody, and the cytokine profiles in a diverse collective of COVID-19 patients (disease severity: asymptomatic to intensive care), using vaccinated individuals and influenza patients as comparisons. We then aim to assess whether the presence of aPL in COVID-19 is associated with an increased incidence of thrombotic events in COVID-19. METHODS AND RESULTS: We conducted anti-SARS-CoV-2 IgG and IgA microELISA and IgG, IgA, and IgM antiphospholipid line immunoassay (LIA) against 10 criteria and non-criteria antigens in 155 plasma samples of 124 individuals, and we measured 16 cytokines and chemokines in 112 plasma samples. We additionally employed clinical and demographic parameters to conduct multivariable regression analyses within multiple paradigms. In line with recent results, we find that IgM autoantibodies against annexin V (AnV), ß2-glycoprotein I (ß2GPI), and prothrombin (PT) are enriched upon infection with SARS-CoV-2. There was no evidence for seroconversion from IgM to IgG or IgA. PT, ß2GPI, and AnV IgM as well as cardiolipin (CL) IgG antiphospholipid levels were significantly elevated in the COVID-19 but not in the influenza or control groups. They were associated predominantly with the strength of the anti-SARS-CoV-2 antibody titres and the major correlate for thromboses was SARS-CoV-2 disease severity. CONCLUSION: While we have recapitulated previous findings, we conclude that the presence of the aPL, most notably PT, ß2GPI, AnV IgM, and CL IgG in COVID-19 are not associated with a higher incidence of thrombotic events.

NECAB2 is an endosomal protein important for striatal function.

Synaptic signaling depends on ATP generated by mitochondria. Dysfunctional mitochondria shift the redox balance towards a more oxidative environment. Due to extensive connectivity, the striatum is especially vulnerable to mitochondrial dysfunction. We found that neuronal calcium-binding protein 2 (NECAB2) plays a role in striatal function and mitochondrial homeostasis. NECAB2 is a predominantly endosomal striatal protein which partially colocalizes with mitochondria. This colocalization is enhanced by mild oxidative stress. Global knockout of Necab2 in the mouse results in increased superoxide levels, increased DNA oxidation and reduced levels of the antioxidant glutathione which correlates with an altered mitochondrial shape and function. Striatal mitochondria from Necab2 knockout mice are more abundant and smaller and characterized by a reduced spare capacity suggestive of intrinsic uncoupling respectively mitochondrial dysfunction. In line with this, we also found an altered stress-induced interaction of endosomes with mitochondria in Necab2 knockout striatal cultures. The predominance of dysfunctional mitochondria and the pro-oxidative redox milieu correlates with a loss of striatal synapses and behavioral changes characteristic of striatal dysfunction like reduced motivation and altered sensory gating. Together this suggests an involvement of NECAB2 in an endosomal pathway of mitochondrial stress response important for striatal function.

Single-cell transcriptomics of NRAS-mutated melanoma transitioning to drug resistance reveals P2RX7 as an indicator of early drug response.

Treatment options for patients with NRAS-mutant melanoma are limited and lack an efficient targeted drug combination that significantly increases overall and progression-free survival. In addition, targeted therapy success is hampered by the inevitable emergence of drug resistance. A thorough understanding of the molecular processes driving cancer cells' escape mechanisms is crucial to tailor more efficient follow-up therapies. We performed single-cell RNA sequencing of NRAS-mutant melanoma treated with MEK1/2 plus CDK4/6 inhibitors to decipher transcriptional transitions during the development of drug resistance. Cell lines resuming full proliferation (FACs [fast-adapting cells]) and cells that became senescent (SACs [slow-adapting cells]) over prolonged treatment were identified. The early drug response was characterized by transitional states involving increased ion signaling, driven by upregulation of the ATP-gated ion channel P2RX7. P2RX7 activation was associated with improved therapy responses and, in combination with targeted drugs, could contribute to the delayed onset of acquired resistance in NRAS-mutant melanoma.

Glucosylceramide Synthase Inhibitors Induce Ceramide Accumulation and Sensitize H3K27 Mutant Diffuse Midline Glioma to Irradiation.

H3K27M mutant (mut) diffuse midline glioma (DMG) is a lethal cancer with no effective cure. The glycosphingolipids (GSL) metabolism is altered in these tumors and could be exploited to develop new therapies. We tested the effect of the glucosylceramide synthase inhibitors (GSI) miglustat and eliglustat on cell proliferation, alone or in combination with temozolomide or ionizing radiation. Miglustat was included in the therapy protocol of two pediatric patients. The effect of H3.3K27 trimethylation on GSL composition was analyzed in ependymoma. GSI reduced the expression of the ganglioside GD2 in a concentration and time-dependent manner and increased the expression of ceramide, ceramide 1-phosphate, sphingosine, and sphingomyelin but not of sphingosine 1-phosphate. Miglustat significantly increased the efficacy of irradiation. Treatment with miglustat according to dose recommendations for patients with Niemann-Pick disease was well tolerated with manageable toxicities. One patient showed a mixed response. In ependymoma, a high concentration of GD2 was found only in the presence of the loss of H3.3K27 trimethylation. In conclusion, treatment with miglustat and, in general, targeting GSL metabolism may offer a new therapeutic opportunity and can be administered in close proximity to radiation therapy. Alterations in H3K27 could be useful to identify patients with a deregulated GSL metabolism.

Dilated Virchow-Robin spaces are a marker for arterial disease in multiple sclerosis.

BACKGROUND: Virchow-Robin spaces (VRS) have been associated with neurodegeneration and neuroinflammation. However, it remains uncertain to what degree non-dilated or dilated VRS reflect specific features of neuroinflammatory pathology. Thus, we aimed at investigating the clinical relevance of VRS as imaging biomarker in multiple sclerosis (MS) and to correlate VRS to their histopathologic signature. METHODS: In a cohort study comprising 142 MS patients and 30 control subjects, we assessed the association of non-dilated and dilated VRS to clinical and magnetic resonance imaging (MRI) outcomes. Findings were corroborated in a validation cohort comprising 63 MS patients. Brain blocks from 6 MS patients and 3 non-MS controls were histopathologically processed to correlate VRS to their tissue substrate. FINDINGS: In our actively treated clinical cohort, the count of dilated centrum semiovale VRS was associated with increased T1 and T2 lesion volumes. There was no systematic spatial colocalization of dilated VRS with MS lesions. At tissue level, VRS mostly corresponded to arteries and were not associated with MS pathological hallmarks. Interestingly, in our ex vivo cohort comprising mostly progressive MS patients, dilated VRS in MS were associated with signs of small vessel disease. INTERPRETATION: Contrary to prior beliefs, these observations suggest that VRS in MS do not associate with an accumulation of immune cells. But instead, these findings indicate vascular pathology as a driver and/or consequence of neuroinflammatory pathology for this imaging feature. FUNDING: NIH, Swedish Society for Medical Research, Swiss National Science Foundation and University of Zurich.

Dilated Virchow-Robin Spaces are a Marker for Arterial Disease in Multiple Sclerosis.

Virchow-Robin spaces (VRS) have been associated with neurodegeneration and neuroinflammation. However, it remains uncertain to what degree non-dilated or dilated VRS reflect specific features of neuroinflammatory pathology. Thus, we aimed at investigating the clinical relevance of VRS as imaging biomarker in multiple sclerosis (MS) and to correlate VRS to their histopathologic signature. In a cohort study comprising 205 MS patients (including a validation cohort) and 30 control subjects, we assessed the association of non-dilated and dilated VRS to clinical and magnetic resonance imaging (MRI) out-comes. Brain blocks from 6 MS patients and 3 non-MS controls were histopathologically processed to correlate VRS to their tissue substrate. The count of dilated centrum semiovale VRS was associated with increased T1 and T2 lesion volumes. There was no systematic spatial colocalization of dilated VRS with MS lesions. At tissue level, VRS mostly corresponded to arteries and were not associated with MS pathological hallmarks. Interestingly, dilated VRS in MS were associated with signs of small vessel disease. Contrary to prior beliefs, these observations suggest that VRS in MS do not associate with accumulation of immune cells. But instead, these findings indicate vascular pathology as a driver and/or consequence of neuroinflammatory pathology for this imaging feature.

GD2 Expression in Medulloblastoma and Neuroblastoma for Personalized Immunotherapy: A Matter of Subtype.

Neuroblastoma (NBL) and medulloblastoma (MB) are aggressive pediatric cancers which can benefit from therapies targeting gangliosides. Therefore, we compared the ganglioside profile of 9 MB and 14 NBL samples by thin layer chromatography and mass spectrometry. NBL had the highest expression of GD2 (median 0.54 nmol GD2/mg protein), and also expressed complex gangliosides. GD2-low samples expressed GD1a and were more differentiated. MB mainly expressed GD2 (median 0.032 nmol GD2/mg protein) or GM3. Four sonic hedgehog-activated (SHH) as well as one group 4 and one group 3 MBs were GD2-positive. Two group 3 MB samples were GD2-negative but GM3-positive. N-glycolyl neuraminic acid-containing GM3 was neither detected in NBL nor MB by mass spectrometry. Furthermore, a GD2-phenotype predicting two-gene signature (ST8SIA1 and B4GALNT1) was applied to RNA-Seq datasets, including 86 MBs and validated by qRT-PCR. The signature values were decreased in group 3 and wingless-activated (WNT) compared to SHH and group 4 MBs. These results suggest that while NBL is GD2-positive, only some MB patients can benefit from a GD2-directed therapy. The expression of genes involved in the ganglioside synthesis may allow the identification of GD2-positive MBs. Finally, the ganglioside profile may reflect the differentiation status in NBL and could help to define MB subtypes.

Normal and Pathological NRF2 Signalling in the Central Nervous System.

The nuclear factor erythroid 2-related factor 2 (NRF2) was originally described as a master regulator of antioxidant cellular response, but in the time since, numerous important biological functions linked to cell survival, cellular detoxification, metabolism, autophagy, proteostasis, inflammation, immunity, and differentiation have been attributed to this pleiotropic transcription factor that regulates hundreds of genes. After 40 years of in-depth research and key discoveries, NRF2 is now at the center of a vast regulatory network, revealing NRF2 signalling as increasingly complex. It is widely recognized that reactive oxygen species (ROS) play a key role in human physiological and pathological processes such as ageing, obesity, diabetes, cancer, and neurodegenerative diseases. The high oxygen consumption associated with high levels of free iron and oxidizable unsaturated lipids make the brain particularly vulnerable to oxidative stress. A good stability of NRF2 activity is thus crucial to maintain the redox balance and therefore brain homeostasis. In this review, we have gathered recent data about the contribution of the NRF2 pathway in the healthy brain as well as during metabolic diseases, cancer, ageing, and ageing-related neurodegenerative diseases. We also discuss promising therapeutic strategies and the need for better understanding of cell-type-specific functions of NRF2 in these different fields.

Senescence Is the Main Trait Induced by Temozolomide in Glioblastoma Cells.

First-line drug in the treatment of glioblastoma, the most severe brain cancer, is temozolomide (TMZ), a DNA-methylating agent that induces the critical damage O6-methylguanine (O6MeG). This lesion is cytotoxic through the generation of mismatch repair-mediated DNA double-strand breaks (DSBs), which trigger apoptotic pathways. Previously, we showed that O6MeG also induces cellular senescence (CSEN). Here, we show that TMZ-induced CSEN is a late response which has similar kinetics to apoptosis, but at a fourfold higher level. CSEN cells show a high amount of DSBs, which are located outside of telomeres, a high level of ROS and oxidized DNA damage (8-oxo-guanine), and sustained activation of the DNA damage response and histone methylation. Despite the presence of DSBs, CSEN cells are capable of repairing radiation-induced DSBs. Glioblastoma cells that acquired resistance to TMZ became simultaneously resistant to TMZ-induced CSEN. Using a Tet-On glioblastoma cell system, we show that upregulation of MGMT immediately after TMZ completely abrogated apoptosis and CSEN, while induction of MGMT long-term (>72 h) after TMZ did not reduce apoptosis and CSEN. Furthermore, upregulation of MGMT in the senescent cell population had no impact on the survival of senescent cells, indicating that O6MeG is required for induction, but not for maintenance of the senescent state. We further show that, in recurrent GBM specimens, a significantly higher level of DSBs and CSEN-associated histone H3K27me3 was observed than in the corresponding primary tumors. Overall, the data indicate that CSEN is a key node induced in GBM following chemotherapy.

The Multifaceted Neurotoxicity of Astrocytes in Ageing and Age-Related Neurodegenerative Diseases: A Translational Perspective.

In a healthy physiological context, astrocytes are multitasking cells contributing to central nervous system (CNS) homeostasis, defense, and immunity. In cell culture or rodent models of age-related neurodegenerative diseases (NDDs), such as Alzheimer's disease (AD) and Parkinson's disease (PD), numerous studies have shown that astrocytes can adopt neurotoxic phenotypes that could enhance disease progression. Chronic inflammatory responses, oxidative stress, unbalanced phagocytosis, or alteration of their core physiological roles are the main manifestations of their detrimental states. However, if astrocytes are directly involved in brain deterioration by exerting neurotoxic functions in patients with NDDs is still controversial. The large spectrum of NDDs, with often overlapping pathologies, and the technical challenges associated with the study of human brain samples complexify the analysis of astrocyte involvement in specific neurodegenerative cascades. With this review, we aim to provide a translational overview about the multi-facets of astrocyte neurotoxicity ranging from in vitro findings over mouse and human cell-based studies to rodent NDDs research and finally evidence from patient-related research. We also discuss the role of ageing in astrocytes encompassing changes in physiology and response to pathologic stimuli and how this may prime detrimental responses in NDDs. To conclude, we discuss how potentially therapeutic strategies could be adopted to alleviate or reverse astrocytic toxicity and their potential to impact neurodegeneration and dementia progression in patients.

Correction: Anti-prothrombin autoantibodies enriched after infection with SARS-CoV-2 and influenced by strength of antibody response against SARS-CoV-2 proteins.

[This corrects the article DOI: 10.1371/journal.ppat.1010118.].

From Research to Diagnostic Application of Raman Spectroscopy in Neurosciences: Past and Perspectives.

In recent years, Raman spectroscopy has been more and more frequently applied to address research questions in neuroscience. As a non-destructive technique based on inelastic scattering of photons, it can be used for a wide spectrum of applications including neurooncological tumor diagnostics or analysis of misfolded protein aggregates involved in neurodegenerative diseases. Progress in the technical development of this method allows for an increasingly detailed analysis of biological samples and may therefore open new fields of applications. The goal of our review is to provide an introduction into Raman scattering, its practical usage and also commonly associated pitfalls. Furthermore, intraoperative assessment of tumor recurrence using Raman based histology images as well as the search for non-invasive ways of diagnosis in neurodegenerative diseases are discussed. Some of the applications mentioned here may serve as a basis and possibly set the course for a future use of the technique in clinical practice. Covering a broad range of content, this overview can serve not only as a quick and accessible reference tool but also provide more in-depth information on a specific subtopic of interest.

Anti-prothrombin autoantibodies enriched after infection with SARS-CoV-2 and influenced by strength of antibody response against SARS-CoV-2 proteins.

Antiphospholipid antibodies (aPL), assumed to cause antiphospholipid syndrome (APS), are notorious for their heterogeneity in targeting phospholipids and phospholipid-binding proteins. The persistent presence of Lupus anticoagulant and/or aPL against cardiolipin and/or ß2-glycoprotein I have been shown to be independent risk factors for vascular thrombosis and pregnancy morbidity in APS. aPL production is thought to be triggered by-among other factors-viral infections, though infection-associated aPL have mostly been considered non-pathogenic. Recently, the potential pathogenicity of infection-associated aPL has gained momentum since an increasing number of patients infected with Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) has been described with coagulation abnormalities and hyperinflammation, together with the presence of aPL. Here, we present data from a multicentric, mixed-severity study including three cohorts of individuals who contracted SARS-CoV-2 as well as non-infected blood donors. We simultaneously measured 10 different criteria and non-criteria aPL (IgM and IgG) by using a line immunoassay. Further, IgG antibody response against three SARS-CoV-2 proteins was investigated using tripartite automated blood immunoassay technology. Our analyses revealed that selected non-criteria aPL were enriched concomitant to or after an infection with SARS-CoV-2. Linear mixed-effects models suggest an association of aPL with prothrombin (PT). The strength of the antibody response against SARS-CoV-2 was further influenced by SARS-CoV-2 disease severity and sex of the individuals. In conclusion, our study is the first to report an association between disease severity, anti-SARS-CoV-2 immunoreactivity, and aPL against PT in patients with SARS-CoV-2.

Ablation of lysozyme M-positive cells prevents aircraft noise-induced vascular damage without improving cerebral side effects.

Aircraft noise induces vascular and cerebral inflammation and oxidative stress causing hypertension and cardiovascular/cerebral dysfunction. With the present studies, we sought to determine the role of myeloid cells in the vascular vs. cerebral consequences of exposure to aircraft noise. Toxin-mediated ablation of lysozyme M+ (LysM+) myeloid cells was performed in LysMCreiDTR mice carrying a cre-inducible diphtheria toxin receptor. In the last 4d of toxin treatment, the animals were exposed to noise at maximum and mean sound pressure levels of 85 and 72 dB(A), respectively. Flow cytometry analysis revealed accumulation of CD45+, CD11b+, F4/80+, and Ly6G-Ly6C+ cells in the aortas of noise-exposed mice, which was prevented by LysM+ cell ablation in the periphery, whereas brain infiltrates were even exacerbated upon ablation. Aircraft noise-induced increases in blood pressure and endothelial dysfunction of the aorta and retinal/mesenteric arterioles were almost completely normalized by ablation. Correspondingly, reactive oxygen species in the aorta, heart, and retinal/mesenteric vessels were attenuated in ablated noise-exposed mice, while microglial activation and abundance in the brain was greatly increased. Expression of phagocytic NADPH oxidase (NOX-2) and vascular cell adhesion molecule-1 (VCAM-1) mRNA in the aorta was reduced, while NFκB signaling appeared to be activated in the brain upon ablation. In sum, we show dissociation of cerebral and peripheral inflammatory reactions in response to aircraft noise after LysM+ cell ablation, wherein peripheral myeloid inflammatory cells represent a dominant part of the pathomechanism for noise stress-induced cardiovascular effects and their central nervous counterparts, microglia, as key mediators in stress responses.

Exploiting Gangliosides for the Therapy of Ewing's Sarcoma and H3K27M-Mutant Diffuse Midline Glioma.

The ganglioside GD2 is an important target in childhood cancer. Nevertheless, the only therapy targeting GD2 that is approved to date is the monoclonal antibody dinutuximab, which is used in the therapy of neuroblastoma. The relevance of GD2 as a target in other tumor entities remains to be elucidated. Here, we analyzed the expression of GD2 in different pediatric tumor entities by flow cytometry and tested two approaches for targeting GD2. H3K27M-mutant diffuse midline glioma (H3K27M-mutant DMG) samples showed the highest expression of GD2 with all cells strongly positive for the antigen. Ewing's sarcoma (ES) samples also showed high expression, but displayed intra- and intertumor heterogeneity. Osteosarcoma had low to intermediate expression with a high percentage of GD2-negative cells. Dinutuximab beta in combination with irinotecan and temozolomide was used to treat a five-year-old girl with refractory ES. Disease control lasted over 12 months until a single partially GD2-negative intracranial metastasis was detected. In order to target GD2 in H3K27M-mutant DMG, we blocked ganglioside synthesis via eliglustat, since dinutuximab cannot cross the blood-brain barrier. Eliglustat is an inhibitor of glucosylceramide synthase, and it is used for treating children with Gaucher's disease. Eliglustat completely inhibited the proliferation of primary H3K27M-mutant DMG cells in vitro. In summary, our data provide evidence that dinutuximab might be effective in tumors with high GD2 expression. Moreover, disrupting the ganglioside metabolism in H3K27M-mutant DMG could open up a new therapeutic option for this highly fatal cancer.

Intracerebral endotheliitis and microbleeds are neuropathological features of COVID-19.

Coronavirus disease 19 (COVID-19) is a rapidly evolving pandemic caused by the coronavirus Sars-CoV-2. Clinically manifest central nervous system symptoms have been described in COVID-19 patients and could be the consequence of commonly associated vascular pathology, but the detailed neuropathological sequelae remain largely unknown. A total of six cases, all positive for Sars-CoV-2, showed evidence of cerebral petechial hemorrhages and microthrombi at autopsy. Two out of six patients showed an elevated risk for disseminated intravascular coagulopathy according to current criteria and were excluded from further analysis. In the remaining four patients, the hemorrhages were most prominent at the grey and white matter junction of the neocortex, but were also found in the brainstem, deep grey matter structures and cerebellum. Two patients showed vascular intramural inflammatory infiltrates, consistent with Sars-CoV-2-associated endotheliitis, which was associated by elevated levels of the Sars-CoV-2 receptor ACE2 in the brain vasculature. Distribution and morphology of patchy brain microbleeds was clearly distinct from hypertension-related hemorrhage, critical illness-associated microbleeds and cerebral amyloid angiopathy, which was ruled out by immunohistochemistry. Cerebral microhemorrhages in COVID-19 patients could be a consequence of Sars- CoV-2-induced endotheliitis and more general vasculopathic changes and may correlate with an increased risk of vascular encephalopathy.