%0 Generic
%A Ritter, Michael
%C Heidelberg
%D 2026
%F heidok:36783
%K Räumliche Transkriptomik
%R 10.11588/heidok.00036783
%T Applications of Spatial Transcriptomics in the  Diagnosis of Brain Tumours
%U https://archiv.ub.uni-heidelberg.de/volltextserver/36783/
%X Brain tumours are among the most aggressive cancer types. Due to the vast amount of different brain tu-mour entities, the correct identification of the tumour type is pivotal for assigning affected patients to a suitable treatment strategy. Among the different brain tumours, the glioma subtypes can be particularly challenging to diagnose, because of their infiltrative growth. The lack of tumour specific markers for glio-blastoma (GB), the most aggressive glioma also adds to the challenge of diagnosing gliomas and makes it difficult to determine the tumour content in infiltration zones and distinguish between infiltration zones and reactive nervous tissue. In addition to the infiltrative growth the diagnosis of gliomas can be further im-paired if only miniscule tissue fragments of the tumour are available, or too little DNA was isolated for a molecular diagnostic workup. However, newly developed technologies like spatial transcriptomics might be able to overcome the previously mentioned issues. To test this hypothesis, I applied sequencing-based spa-tial transcriptomics on 43 glioma samples and imaging-based spatial transcriptomics on 17 high-grade gli-oma samples. Single-nuclei data was also generated of these 17 samples to aid in the analysis. The spatial transcriptomics data was benchmarked against the results of the current diagnostic techniques focusing on identification of copy number variations (CNV), immunohistochemistry (IHC) stains and identification of molecular alterations. The CNVs inferred from the spatial transcriptomics data matched the ones generat-ed by the Illumina EPIC array. Using CNVs inferred from the spatial transcriptomics data I could even identi-fy sub clones, which were not detectable by the EPIC array. Furthermore, the inferred IHC stains for Ki67, GFAP and NeuN matched the real IHC stains. Using the spatial transcriptomics data, I was also able to iden-tify molecular alterations on a single gene level such as EGFR amplifications or mutations. Moreover, single infiltrating tumour cells of GB could be robustly identified using imaging-based spatial transcriptomics data. Leveraging the combination of single cell and spatial transcriptomics methods I investigated the processes in the perinecrotic regions and showed a spatially resolved image of the VEGFA signalling and response in GB. In conclusion, I reproduced the results of the majority of the current diagnostic workflow using spatial transcriptomics and minuscule tissue fragments of 5 µm thickness and a 1 mm diameter. Therefore, spatial transcriptomics proved to be a valuable addition to the diagnostic workflow of gliomas.