Soylemezoglu, F. | Search bioRxiv

bioRxiv2026

Deciphering the Genomic Landscape of Renal Cell Carcinoma Brain Metastases

Brain metastases from renal cell carcinoma (RCC) remain a major cause of morbidity and mortality, yet the genomic features associated with metastatic dissemination remain poorly understood. Whole-exome sequencing was performed on 72 RCC brain metastasis samples with matched normal. To identify candidate genomic alterations associated with brain metastasis, the genomic alterations detected in the brain metastases were compared against alterations in extracranial metastases from the MSK-ECM cohort (n=137) and primary RCC tumors from TCGA (n=432) by case-control analyses. Candidate alterations were also identified through matched-pair analyses comparing brain metastases with matched primary tumors or extracranial metastases from the same patient (n=25). A random survival forest model incorporating the candidate CNA events was developed to predict overall survival. The candidate CNAs were further evaluated using functional experimental data from MetMap and DepMap. Survival analyses were conducted to assess the prognostic relevance of these alterations. We identified recurrent CNAs enriched in RCC brain metastases, including 4q loss, 7p gain, 7q gain, 8p loss, 8q gain, 9p21.3 deletion, 12q15 amplification, and 14q loss. These alterations were associated with significantly poorer patient survival among RCC patients. A random survival forest model based on these CNA events stratified TCGA-KIRC patients into prognostically distinct risk groups (C-index = 0.64). Among the recurrent CNAs, 8p loss, 8q gain, 9p21.3 deletion were associated with increased incidence of brain metastases across multiple primary cancer types in xenograft mouse models. These alterations were also strongly associated with metastatic progression and poor prognosis across RCC, lung adenocarcinoma, breast cancer, and melanoma. These findings indicate a shared genomic basis for brain tropism and highlight the potential utility of copy-number alterations as biomarkers for risk stratification and clinical decision-making.

genetics↗

bioRxiv2019

Same same but different: a web-based deep learning application for the histopathologic distinction of cortical malformations

We trained a convolutional neural network (CNN) to classify H.E. stained microscopic images of focal cortical dysplasia type IIb (FCD IIb) and cortical tuber of tuberous sclerosis complex (TSC). Both entities are distinct subtypes of human malformations of cortical development that share histopathological features consisting of neuronal dyslamination with dysmorphic neurons and balloon cells. The microscopic review of routine stainings of such surgical specimens remains challenging. A digital processing pipeline was developed for a series of 56 FCD IIb and TSC cases to obtain 4000 regions of interest and 200.000 sub-samples with different zoom and rotation angles to train a CNN. Our best performing network achieved 91% accuracy and 0.88 AUCROC (area under the receiver operating characteristic curve) on a hold-out test-set. Guided gradient-weighted class activation maps visualized morphological features used by the CNN to distinguish both entities. We then developed a web application, which combined the visualization of whole slide images (WSI) with the possibility for classification between FCD IIb and TSC on demand by our pretrained and build-in CNN classifier. This approach might help to introduce deep learning applications for the histopathologic diagnosis of rare and difficult-to-classify brain lesions.

pathology↗