Georg Brunner

Dr. Georg Brunner

orcid

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
16 July 2022 in The American Journal of Pathology
Verified
Journal Article
Open access
GBGeorg Brunner
Georg Brunner
MRMeike S. Roux
Meike S. Roux
TFThomas Falk
Thomas Falk
MBMartina Bresch
Martina Bresch
VBVolker Böhm
Volker Böhm
TMThomas Meiners
Thomas Meiners
Spinal cord injury (SCI) is associated with venous vascular dysfunction below the level of injury, resulting in dysregulation of tissue fluid homeostasis in afflicted skin. The purpose of this study was to determine whether loss of neuronal control in chronic SCI also affects the skin lymphatic system. Morphology of lymphatics was characterized by immunohistochemistry and lymphatic gene expression profiles determined by DNA microarray analysis. In SCI, skin lymphatic function appeared to be impaired, because the ratio of functionally dilated versus collapsed lymphatic vessels was decreased 10-fold compared with control. Consequently, the average lumen area of lymphatic vessels was almost halved, possibly due to the known impaired connective tissue integrity of SCI skin. In fact, collagenases were found to be overexpressed in SCI skin, and dermal collagen structure was impaired. Molecular profiling also suggested an SCI-specific phenotype of increased connective tissue turnover and decreased lymphatic contractility. The total number of lymphatic vessels in SCI skin, however, was doubled, pointing to enhanced lymphangiogenesis. In conclusion, these data show, for the first time, that lymphatic function and development in human skin are under neuronal control. Because peripheral venous and lymphatic vascular defects are associated with disturbed fluid homeostasis, inappropriate wound healing reactions, and impaired skin immunity, they might contribute to the predisposition of afflicted individuals to pressure ulcer formation and wound healing disorders.

15 March 2021 in International Wound Journal
Verified
Journal Article
Open access
PDF file
GBGeorg Brunner
Georg Brunner
MRMeike Roux
Meike Roux
VBVolker Böhm
Volker Böhm
TMThomas Meiners
Thomas Meiners
Patients with spinal cord injury have a predisposition to develop pressure ulcers. Specific characteristics of the patients' skin potentially involved have not yet been identified. The purpose of this investigation was to determine whether loss of neuronal control affects cellular and molecular homeostasis in the skin. Intact afflicted skin, wound edge of pressure ulcers, and control skin were analysed. Platelets, transforming growth factor‐β1, and activin A were identified by immunohistochemistry. Transforming growth factor‐β‐like activity was determined by bioassay, and gene expression by DNA microarray analysis or RT‐PCR. In afflicted skin, enhanced platelet extravasation was detected. Transforming growth factor‐β1 and activin A accumulated in the dermal‐epidermal junction zone. Transforming growth factor‐β‐like activity and activin A expression were increased in intact afflicted skin (compared to control skin) and were further enhanced in pressure ulcers. In vitro, activity was generated by fibroblast‐epithelial cell interactions, which also induced activin A. Thus, loss of neuronal control in spinal cord injury appears to trigger inappropriate wound healing processes in the patients' skin. Plasma leakage and increased transforming growth factor‐β‐like activity combined with shear forces potentially enhance the risk for pressure ulcer formation.

20 May 2017 in Journal of Clinical Oncology
Verified
Journal Article
Closed access
Thilo GambichlerThilo Gambichler
Thilo Gambichler
URUwe Reinhold
Uwe Reinhold
KTKonstantinos Tsagoudis
Konstantinos Tsagoudis
EBErhard Bierhoff
Erhard Bierhoff
TDThomas Dirschka
Thomas Dirschka
ESEggert Stockfleth
Eggert Stockfleth
E21037 Background: Staging of cutaneous melanoma (CM), as defined by the American Joint Committee on Cancer (AJCC), includes determination of the sentinel lymph node (SLN) status as a major prognostic factor. We have previously identified and validated a gene expression risk score (GRS) in primary CM and adjacent stroma, consisting of 7 protective genes and 1 risk-associated gene, which predicts patient survival, independently of AJCC stage. This external validation study evaluated single and combined performance of GRS and SLN in predicting relapse-free survival (RFS). Methods: The study cohort included 203 formalin-fixed, paraffin-embedded (FFPE) -primary CMs (AJCC stages IA-IIIC, n=31 RFS events). Tissue samples and clinical data were provided by two centers, and gene expression was determined externally using real-time RT-PCR (double-blind study). External prospective data analysis followed a pre-specified protocol. Results: Integration Discrimination Improvement (IDI) analysis demonstrated non-inferiority of GRS vs. SLN status in predicting RFS (IDI estimate=-0.8%, p=0.040). In univariate Cox regression analysis, GRS and SLN status predicted RFS (GRS: p=0.011, HR=2.44 [1.2-4.95]; SLN: p=0.038, HR=2.12 [1.03-4.38]). Multivariate Cox regression demonstrated independent contribution of prognostic information by both parameters (GRS: p=0.015, HR=2.40 [1.18-4.89]; SLN: p=0.046, HR=2.11 [1.02-4.37]. IDI analysis demonstrated that combining GRS and SLN status improved prognostic performance, as compared to SLN status alone (IDI estimate=3.2%, p=0.044). When complementing SLN status with GRS (SLN positive or high GRS), sensitivity of relapse detection increased from 38.7% for SLN alone to 67.7%. Conclusions: GRS is a non-invasive predictor of RFS, complementary to SLN analysis. Combining GRS and SLN status significantly improves accuracy of melanoma prognosis and, therefore, may guide follow-up care as well as treatment decisions regarding new adjuvant therapies, e.g. of SLN positive patients with high GRS. [Table: see text]

20 May 2017 in Journal of Clinical Oncology
Verified
Journal Article
Closed access
GBGeorg Brunner
Georg Brunner
TFThomas Falk
Thomas Falk
BEBeyhan Ertas
Beyhan Ertas
CBCarola Berking
Carola Berking
Hans-Joachim SchulzeHans-Joachim Schulze
Hans-Joachim Schulze
NBNorbert Blödorn-Schlicht
Norbert Blödorn-Schlicht
9560 Background: Melanoma staging, as defined by the American Joint Committee on Cancer (AJCC), is limited in its ability to predict outcome. We have previously identified and validated a prognostic gene signature expressed in primary cutaneous melanoma and adjacent stroma. The signature comprises seven protective genes (down-regulated with tumor progression) and one risk-associated gene (up-regulated). A signature-based risk score independently predicts patient survival, across AJCC stages IA-IIIC, in formalin-fixed, paraffin-embedded (FFPE) melanomas (training cohort, n = 125; p = 0.0003, hazard ratio 1.85). The score has been validated in 206 melanomas, selected to be significantly mis-prognosticated by AJCC staging regarding patient survival (40.8% mis-prognostication). In this cohort, the score outperforms AJCC staging (p = 0.0005, hazard ratio 1.41 vs. p = n.s.), correcting 34.9% of AJCC-based mis-prognostications. Methods: Here, we report twofold external validation of the risk score, (i) prognostic performance in silico using the SurvExpress web tool (Aguirre-Gamboa et al., 2013), and (ii) technical performance in vitro(Dermatologikum Hamburg; IMGM Munich). Results: (i) Kaplan Meier analysis and log-rank testing demonstrated that all signature genes combined predicted survival in four different cohorts of metastatic melanoma (from GEO Expression Omnibus or Cancer Genome Atlas; cohorts dichotomized at the median): see table. (ii) The risk score was re-analyzed in melanomas of the training cohort (n=69). The overall concordance of duplicate determinations was 90% (average scores of 1.12 ± 0.14 and 0.97 ± 0.14). Conclusions: In conclusion, we have validated a signature-based FFPE melanoma risk score, complementary to AJCC staging in predicting outcome: (i) Signature genes predicted patient survival in silico(n=449) (ii) The risk score proved to be reproducible and technically robust in vitro. The score improves risk stratification and decision making in melanoma, particularly regarding new adjuvant therapies.

10 February 2022 in arXiv
Verified
Preprint
Open access
PDF file
NKNeerav Karani
Neerav Karani
GBGeorg Brunner
Georg Brunner
EEErtunc Erdil
Ertunc Erdil
SFSimin Fei
Simin Fei
KTKerem Tezcan
Kerem Tezcan
EKEnder Konukoglu
Ender Konukoglu
Performance of convolutional neural networks (CNNs) in image analysis tasks is often marred in the presence of acquisition-related distribution shifts between training and test images. Recently, it has been proposed to tackle this problem by fine-tuning trained CNNs for each test image. Such test-time-adaptation (TTA) is a promising and practical strategy for improving robustness to distribution shifts as it requires neither data sharing between institutions nor annotating additional data. Previous TTA methods use a helper model to increase similarity between outputs and/or features extracted from a test image with those of the training images. Such helpers, which are typically modeled using CNNs, can be task-specific and themselves vulnerable to distribution shifts in their inputs. To overcome these problems, we propose to carry out TTA by matching the feature distributions of test and training images, as modelled by a field-of-experts (FoE) prior. FoEs model complicated probability distributions as products of many simpler expert distributions. We use 1D marginal distributions of a trained task CNN's features as experts in the FoE model. Further, we compute principal components of patches of the task CNN's features, and consider the distributions of PCA loadings as additional experts. We validate the method on 5 MRI segmentation tasks (healthy tissues in 4 anatomical regions and lesions in 1 one anatomy), using data from 17 clinics, and on a MRI registration task, using data from 3 clinics. We find that the proposed FoE-based TTA is generically applicable in multiple tasks, and outperforms all previous TTA methods for lesion segmentation. For healthy tissue segmentation, the proposed method outperforms other task-agnostic methods, but a previous TTA method which is specifically designed for segmentation performs the best for most of the tested datasets. Our code is publicly available.