Refine
Year of publication
Document Type
- Article (46)
Language
- English (46)
Has Fulltext
- yes (46)
Is part of the Bibliography
- no (46)
Keywords
- glioblastoma (10)
- glioma (3)
- Brain metastasis (2)
- EMT (2)
- Glioblastoma (2)
- angiogenesis (2)
- anti-angiogenic therapy (2)
- brain metastases (2)
- metabolism (2)
- pericytes (2)
Institute
- Medizin (46) (remove)
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.
Immune profile and radiological characteristics of progressive multifocal leukoencephalopathy
(2021)
Background and purpose: Progressive multifocal leukoencephalopathy (PML) constitutes a severe disease with increasing incidence, mostly in the context of immunosuppressive therapies. A detailed understanding of immune response in PML appears critical for the treatment strategy. The aim was a comprehensive immunoprofiling and radiological characterization of magnetic resonance imaging (MRI) defined PML variants.
Methods: All biopsy-confirmed PML patients (n = 15) treated in our department between January 2004 and July 2019 were retrospectively analysed. Data from MRI, histology as well as detailed clinical and outcome data were collected. The MRI-defined variants of classical (cPML) and inflammatory (iPML) PML were discriminated based on the intensity of gadolinium enhancement. In these PML variants, intensity and localization (perivascular vs. parenchymal) of inflammation in MRI and histology as well as the cellular composition by immunohistochemistry were assessed. The size of the demyelinating lesions was correlated with immune cell infiltration.
Results: Patients with MRI-defined iPML showed a stronger intensity of inflammation with an increased lymphocyte infiltration on histological level. Also, iPML was characterized by a predominantly perivascular inflammation. However, cPML patients also demonstrated certain inflammatory tissue alterations. Infiltration of CD163-positive microglia and macrophage (M/M) subtypes correlated with PML lesion size.
Conclusions: The non-invasive MRI-based discrimination of PML variants allows for an estimation of inflammatory tissue alterations, although exhibiting limitations in MRI-defined cPML. The association of a distinct phagocytic M/M subtype with the extent of demyelination might reflect disease progression.
Aims: In primary central nervous system tumours, epithelial-to-mesenchymal transition (EMT) gene expression is associated with increased malignancy. However, it has also been shown that EMT factors in gliomas are almost exclusively expressed by glioma vessel-associated pericytes (GA-Peris). In this study, we aimed to identify the mechanism of EMT in GA-Peris and its impact on angiogenic processes.
Methods; In glioma patients, vascular density and the expression of the pericytic markers platelet derived growth factor receptor (PDGFR)-β and smooth muscle actin (αSMA) were examined in relation to the expression of the EMT transcription factor SLUG and were correlated with survival of patients with glioblastoma (GBM). Functional mechanisms of SLUG regulation and the effects on primary human brain vascular pericytes (HBVP) were studied in vitro by measuring proliferation, cell motility and growth characteristics.
Results: The number of PDGFR-β- and αSMA-positive pericytes did not change with increased malignancy nor showed an association with the survival of GBM patients. However, SLUG-expressing pericytes displayed considerable morphological changes in GBM-associated vessels, and TGF-β induced SLUG upregulation led to enhanced proliferation, motility and altered growth patterns in HBVP. Downregulation of SLUG or addition of a TGF-β antagonising antibody abolished these effects.
Conclusions: We provide evidence that in GA-Peris, elevated SLUG expression is mediated by TGF-β, a cytokine secreted by most glioma cells, indicating that the latter actively modulate neovascularisation not only by modulating endothelial cells, but also by influencing pericytes. This process might be responsible for the formation of an unstructured tumour vasculature as well as for the breakdown of the blood–brain barrier in GBM.
Simple Summary: Therapeutic antibodies are an integral part of treatment regimens for metastasized colorectal cancer. In KRAS wildtype tumors both bevacizumab and cetuximab are active. While bevacizumab has previously been shown to induce tumor hypoxia, we here report that EGFR inhibition by cetuximab protects colon cancer cells from hypoxia-induced cell death. This effect appears to be responsible for the inferior efficacy of a treatment sequence of bevacizumab followed by cetuximab versus an inverse sequence that we observed in a colorectal cancer mouse model. It also offers a mechanistic explanation for effects observed in clinical trials such as underadditive or even detrimental effects when combining bevacizumab and cetuximab (CAIRO2 trial) and the superior efficacy of first line cetuximab (FIRE-3 trial) under chemotherapy backbones in colorectal cancer.
Abstract: Monoclonal antibodies like cetuximab, targeting the epidermal growth factor receptor (EGFR), and bevacizumab, targeting the vascular endothelial growth factor (VEGF), are an integral part of treatment regimens for metastasized colorectal cancer. However, inhibition of the EGFR has been shown to protect human glioma cells from cell death under hypoxic conditions. In colon carcinoma cells, the consequences of EGFR blockade in hypoxia (e.g., induced by bevacizumab) have not been evaluated yet. LIM1215 and SW948 colon carcinoma and LNT-229 glioblastoma cells were treated with cetuximab, PD153035, and erlotinib and analyzed for cell density and viability. The sequential administration of either cetuximab followed by bevacizumab (CET->BEV) or bevacizumab followed by cetuximab (BEV->CET) was investigated in a LIM1215 (KRAS wildtype) and SW948 (KRAS mutant) xenograft mouse model. In vitro, cetuximab protected from hypoxia. In the LIM1215 model, a survival benefit with cetuximab and bevacizumab monotherapy was observed, but only the sequence CET->BEV showed an additional benefit. This effect was confirmed in the SW948 model. Our observations support the hypothesis that bevacizumab modulates the tumor microenvironment (e.g., by inducing hypoxia) where cetuximab could trigger protective effects when administered later on. The sequence CET->BEV therefore seems to be superior as possible mutual adverse effects are bypassed.
Linking epigenetic signature and metabolic phenotype in IDH mutant and IDH wildtype diffuse glioma
(2020)
Aims: Changes in metabolism are known to contribute to tumour phenotypes. If and how metabolic alterations in brain tumours contribute to patient outcome is still poorly understood. Epigenetics impact metabolism and mitochondrial function. The aim of this study is a characterisation of metabolic features in molecular subgroups of isocitrate dehydrogenase mutant (IDHmut) and isocitrate dehydrogenase wildtype (IDHwt) gliomas. Methods: We employed DNA methylation pattern analyses with a special focus on metabolic genes, large-scale metabolism panel immunohistochemistry (IHC), qPCR-based determination of mitochondrial DNA copy number and immune cell content using IHC and deconvolution of DNA methylation data. We analysed molecularly characterised gliomas (n = 57) for in depth DNA methylation, a cohort of primary and recurrent gliomas (n = 22) for mitochondrial copy number and validated these results in a large glioma cohort (n = 293). Finally, we investigated the potential of metabolic markers in Bevacizumab (Bev)-treated gliomas (n = 29). Results: DNA methylation patterns of metabolic genes successfully distinguished the molecular subtypes of IDHmut and IDHwt gliomas. Promoter methylation of lactate dehydrogenase A negatively correlated with protein expression and was associated with IDHmut gliomas. Mitochondrial DNA copy number was increased in IDHmut tumours and did not change in recurrent tumours. Hierarchical clustering based on metabolism panel IHC revealed distinct subclasses of IDHmut and IDHwt gliomas with an impact on patient outcome. Further quantification of these markers allowed for the prediction of survival under anti-angiogenic therapy. Conclusion: A mitochondrial signature was associated with increased survival in all analyses, which could indicate tumour subgroups with specific metabolic vulnerabilities.
Impact of Docetaxel on blood-brain barrier function and formation of breast cancer brain metastases
(2019)
Background: Breast cancer (BC) is the most frequent malignant tumor in females and the 2nd most common cause of brain metastasis (BM), that are associated with a fatal prognosis. The increasing incidence from 10% up to 40% is due to more effective treatments of extracerebral sites with improved prognosis and increasing use of MRI in diagnostics. A frequently administered, potent chemotherapeutic group of drugs for BC treatment are taxanes usually used in the adjuvant and metastatic setting, which, however, have been suspected to be associated with a higher incidence of BM. The aim of our study was to experimentally analyze the impact of the taxane docetaxel (DTX) on brain metastasis formation, and to elucidate the underlying molecular mechanism.
Methods: A monocentric patient cohort was analyzed to determine the association of taxane treatment and BM formation. To identify the specific impact of DTX, a murine brain metastatic model upon intracardial injection of breast cancer cells was conducted. To approach the functional mechanism, dynamic contrast-enhanced MRI and electron microscopy of mice as well as in-vitro transendothelial electrical resistance (TEER) and tracer permeability assays using brain endothelial cells (EC) were carried out. PCR-based, immunohistochemical and immunoblotting analyses with additional RNA sequencing of murine and human ECs were performed to explore the molecular mechanisms by DTX treatment.
Results: Taxane treatment was associated with an increased rate of BM formation in the patient cohort and the murine metastatic model. Functional studies did not show unequivocal alterations of blood-brain barrier properties upon DTX treatment in-vivo, but in-vitro assays revealed a temporary DTX-related barrier disruption. We found disturbance of tubulin structure and upregulation of tight junction marker claudin-5 in ECs. Furthermore, upregulation of several members of the tubulin family and downregulation of tetraspanin-2 in both, murine and human ECs, was induced.
Conclusion: In summary, a higher incidence of BM was associated with prior taxane treatment in both a patient cohort and a murine mouse model. We could identify tubulin family members and tetraspanin-2 as potential contributors for the destabilization of the blood-brain barrier. Further analyses are needed to decipher the exact role of those alterations on tumor metastatic processes in the brain.
Glioblastoma (GBM), WHO grade IV, is the most aggressive primary brain tumor in adults. The median survival time using standard therapy is only 12–15 months with a 5-year survival rate of around 5%. Thus, new and effective treatment modalities are of significant importance. Signal transducer and activator of transcription 3 (Stat3) is a key signaling protein driving major hallmarks of cancer and represents a promising target for the development of targeted glioblastoma therapies. Here we present data showing that the therapeutic application of siRNAs, formulated in nanoscale lipopolyplexes (LPP) based on polyethylenimine (PEI) and the phospholipid 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), represents a promising new approach to target Stat3 in glioma. We demonstrate that the LPP-mediated delivery of siRNA mediates efficient knockdown of Stat3, suppresses Stat3 activity and limits cell growth in murine (Tu2449) and human (U87, Mz18) glioma cells in vitro. In a therapeutic setting, intracranial application of the siRNA-containing LPP leads to knockdown of STAT3 target gene expression, decreased tumor growth and significantly prolonged survival in Tu2449 glioma-bearing mice compared to negative control-treated animals. This is a proof-of-concept study introducing PEI-based lipopolyplexes as an efficient strategy for therapeutically targeting oncoproteins with otherwise limited druggability.
Background. Extracts from Viscum album L. (VE) are used in the complementary cancer therapy in Europe for decades. VE contain several compounds like the mistletoe lectins (MLs) 1-3 and viscotoxins and also several minor ingredients. Since mistletoe lectin 1 (ML-1) has been described as the main component of VE harboring antitumor activity, purified native or recombinant ML-1 has been recently used in clinical trials. MLs stimulate the immune system, induce cytotoxicity, are able to modify the expression of cancer-associated genes, and influence the proliferation and motility of tumor cells.
Objective. In this study our goal was to determine anticancer effects of the VE ISCADOR Qu, of recombinant ML-1 (Aviscumine), and of native ML-1 in the treatment of glioblastoma (GBM), the most common and highly malignant brain tumor in adults. Additionally we were interested whether these drugs, used in combination with a temozolomide-(TMZ)-based radio-chemotherapy, provide synergistic effects.
Methods. Cell culture assays, ex vivo murine hippocampal brain slice cultures, human GBM cryosections, and a xenograft orthotopic glioblastoma mouse model were used.
Results. In cells, the expression of the ML receptor CD75s, which is also expressed in GBM specimen, but not in normal brain, correlates with the drug-induced cytotoxicity. In GBM cells, the drugs induce cell death in a concentration-dependent manner and reduce cell growth by inducing cell cycle arrest in the G2/M phase. The cell cycle arrest was paralleled by modifications in the expression of cell cycle regulating genes. ML containing drugs, if combined with glioma standard therapy, provide synergistic and additive anticancer effects. Despite not reaching statistical significance, a single intratumoral application of Aviscumine prolonged the median survival of GBM mice longer than tumor irradiation. Moreover, intratumorally applied Aviscumine prolonged the survival of GBM-bearing mice if used in combination with irradiation and TMZ for further 6.5 days compared to the radio-chemotherapy.
Conclusion. Our results suggest that an adjuvant treatment of glioma patients with ML-containing drugs might be beneficial.
Background: Despite significant advances in the understanding of glioblastoma genetics and biology, survival is still poor. Hypoxia and nutrient depletion in the tumour microenvironment induce adaptive signalling and metabolic responses, which can influence sensitivity to therapeutic regimens. DNA damage-inducible transcript 4 (DDIT4) is a protein induced by hypoxia and in response to DNA stress. Mechanistically, DDIT4 inhibits mammalian target of rapamycin complex 1 (mTORC1) signalling by activation of the tuberous sclerosis 1/2 (TSC1/2) complex.
Methods: Using short hairpin RNA-mediated gene suppression as well as doxycycline-regulated gene induction, we developed a glioblastoma cell model to study effects of DDIT4 under conditions of the glioblastoma microenvironment and therapy.
Results: We found an intact DDIT4-mTORC1 signalling axis in human glioblastoma cells that was inducible by hypoxia. Temozolomide and radiotherapy also induced DDIT4 and repressed mTORC1 activity in some glioblastoma cell lines. DDIT4 gene suppression sensitised glioma cells towards hypoxia-induced cell death, while DDIT4 overexpression protected them. Additionally, in clonogenic survival analyses, DDIT4 induction conferred protection from radiotherapy and temozolomide, while DDIT4 gene suppression sensitised cells.
Conclusions: We identified DDIT4 as a cell-intrinsic regulator for adaptive responses and therapy resistance in glioblastoma cells which may interfere with cell death induction by temozolomide, radiotherapy or hypoxia by inhibiting mTORC1 activity.
Akt and mTORC1 signaling as predictive biomarkers for the EGFR antibody nimotuzumab in glioblastoma
(2018)
Glioblastoma (GB) is the most frequent primary brain tumor in adults with a dismal prognosis despite aggressive treatment including surgical resection, radiotherapy and chemotherapy with the alkylating agent temozolomide. Thus far, the successful implementation of the concept of targeted therapy where a drug targets a selective alteration in cancer cells was mainly limited to model diseases with identified genetic drivers. One of the most commonly altered oncogenic drivers of GB and therefore plausible therapeutic target is the epidermal growth factor receptor (EGFR). Trials targeting this signaling cascade, however, have been negative, including the phase III OSAG 101-BSA-05 trial. This highlights the need for further patient selection to identify subgroups of GB with true EGFR-dependency. In this retrospective analysis of treatment-naïve samples of the OSAG 101-BSA-05 trial cohort, we identify the EGFR signaling activity markers phosphorylated PRAS40 and phosphorylated ribosomal protein S6 as predictive markers for treatment efficacy of the EGFR-blocking antibody nimotuzumab in MGMT promoter unmethylated GBs. Considering the total trial population irrespective of MGMT status, a clear trend towards a survival benefit from nimotuzumab was already detectable when tumors had above median levels of phosphorylated ribosomal protein S6. These results could constitute a basis for further investigations of nimotuzumab or other EGFR- and downstream signaling inhibitors in selected patient cohorts using the reported criteria as candidate predictive biomarkers.