Mélanie Lubrano

Background: Glioblastoma (GBM), the most aggressive primary brain tumor, has a median survival of approximately 15 months. Twenty percent of patients survive beyond three years, but known clinical factors like age, performance status, resection extent, and MGMT promoter methylation status do not fully explain the observed outcomes. Objective: Our objective was to identify novel histology derived biomarkers associated with end-of-spectrum overall survival (OS) to provide novel biological insight with a translational potential. Methods: We analyzed a total of 748 GBM patients from 3 different cohorts, uniquely enriched in long survivors (n=98 with overall survival (OS) > 5y including n=196 with OS≥3y), with clinical data and H&E slides obtained from the primary tumor at baseline. We propose an interpretable machine learning (ML) methodology for the discovery of histological biomarkers. Our method learned to segment each H&E slide into three distinct regions associated with long-term survival, short-term survival, and non-informative tissue. We characterized these regions by integrating unsupervised learning, nuclei segmentation, blood vessels detection, pathologist annotations, and multimodal data including spatial transcriptomics from n=31 patients of the GBM MOSAIC dataset to discover fully interpretable biomarkers. Results: Our OS prediction model using histology and clinical data as input achieved an area under the curve (AUC) of 0.85 for the classification of patients between OSPLIN2 expression and lipid accumulation- is consistently enriched in histological areas predictive of poor prognosis. Conclusion: Our interpretable ML methodology identified a novel prognostic impact of biological processes and cell types according to distinct tumor regions of GBM. These results pave the way for spatially-informed biomarkers to improve risk stratification and for personalized spatially-targeted therapeutic strategies. Key highlights: Our ML model identified histological biomarkers predicting prognosis independently from known clinical factors The region of lowly infiltrated white matter enriched in malignant cells including a mesenchymal-like phenotype is predictive of poor prognosis Angiogenesis is increased in areas predictive of long survival in main non-necrotic tumor regions. The subtype of macrophages expressing PLIN2 and associated with increased lipid metabolism was associated with poor prognosis in all GBM regions. Highlights:

Background: Diagnosis of head and neck (HN) squamous dysplasias and carcinomas is critical for patient care, cure, and follow‐up. It can be challenging, especially for grading intraepithelial lesions. Despite recent simplification in the last WHO grading system, the inter‐ and intraobserver variability remains substantial, particularly for nonspecialized pathologists, exhibiting the need for new tools to support pathologists.Methods: In this study we investigated the potential of deep learning to assist the pathologist with automatic and reliable classification of HN lesions following the 2022 WHO classification system. We created, for the first time, a large‐scale database of histological samples (>2000 slides) intended for developing an automatic diagnostic tool. We developed and trained a weakly supervised model performing classification from whole‐slide images (WSI). We evaluated our model on both internal and external test sets and we defined and validated a new confidence score to assess the predictions that can be used to identify difficult cases.Results: Our model demonstrated high classification accuracy across all lesion types on both internal and external test sets (respectively average area under the curve [AUC]: 0.878 (95% confidence interval [CI]: [0.834–0.918]) and 0.886 (95% CI: [0.813–0.947])) and the confidence score allowed for accurate differentiation between reliable and uncertain predictions.Conclusion: Our results demonstrate that the model, associated with confidence measurements, can help in the difficult task of classifying HN squamous lesions by limiting variability and detecting ambiguous cases, taking us one step closer to a wider adoption of AI‐based assistive tools.

Background: Diagnosis of head and neck (HN) squamous dysplasias and carcinomas is critical for patient care cure and follow-up. It can be challenging, especially for grading intraepithelial lesions. Despite recent simplification in the last WHO grading system, the inter- and intra-observer variability remains substantial, particularly for non-specialized pathologists, exhibiting the need for new tools to support pathologists. Methods: In this study we investigated the potential of deep learning to assist the pathologist with automatic and reliable classification of HN lesions following the 2022 WHO classification system. We created, for the first time, a large-scale database of histological samples (>2000 slides) intended for developing an automatic diagnostic tool. We developed and trained a weakly supervised model performing classification from whole slide images (WSI). We evaluated our model on both internal and external test sets and we defined and validated a new confidence score to assess the predictions which can be used to identify difficult cases. Results: Our model demonstrated high classification accuracy across all lesion types on both internal and external test sets (respectively average AUC: 0.878 (95% CI:[0.834-0.918]) and 0.886 (95% CI: [0.813-0.947])) and the confidence score allowed for accurate differentiation between reliable and uncertain predictions. Conclusions: Our results demonstrate that the model, associated with confidence measurements, can help in the difficult task of classifying head and neck squamous lesions by limiting variability and detecting ambiguous cases, taking us one step closer to a wider adoption of AI-based assistive tools.

In computational pathology, predictive models from Whole Slide Images (WSI) mostly rely on Multiple Instance Learning (MIL), where the WSI are represented as a bag of tiles, each of which is encoded by a Neural Network (NN). Slide-level predictions are then achieved by building models on the agglomeration of these tile encodings. The tile encoding strategy thus plays a key role for such models. Current approaches include the use of encodings trained on unrelated data sources, full supervision or self-supervision. While self-supervised learning (SSL) exploits unlabeled data, it often requires large computational resources to train. On the other end of the spectrum, fully-supervised methods make use of valuable prior knowledge about the data but involve a costly amount of expert time. This paper proposes a framework to reconcile SSL and full supervision, showing that a combination of both provides efficient encodings, both in terms of performance and in terms of biological interpretability. On a recently organized challenge on grading Cervical Biopsies, we show that our mixed supervision scheme reaches high performance (weighted accuracy (WA): 0.945), outperforming both SSL (WA: 0.927) and transfer learning from ImageNet (WA: 0.877). We further shed light upon the internal representations that trigger classification results, providing a method to reveal relevant phenotypic patterns for grading cervical biopsies. We expect that the combination of full and self-supervision is an interesting strategy for many tasks in computational pathology and will be widely adopted by the field.