Abstract 4519: June 2nd, 2024 - 9:00 AM-12:00 PM
Association of machine learning (ML)–derived histological features with transcriptomic molecular subtypes in advanced renal cell carcinoma (RCC)
Previously, transcriptomic analysis in the Phase 3 IMmotion 151 (Im151) trial identified 7 molecular subtypes that showed differential outcomes to Atezolizumab+Bevacizumab (A+B) vs Sunitinib (S) treatment. In this abstract, in collaboration with Genentech, human interpretable features (HIFs), including blood vessels, immune cells, fibroblasts, tissue morphologies, and nucleus shape, extracted from H&E-stained whole slide images (WSI) from Im151 and Im150, were used to identify positively associated HIFs within each subgroup in the Im151 WSI and then validated in Im150 molecular subgroups. 169 HIFs were differentially enriched across 3 molecular subsets in both datasets. Our results suggest that clinically relevant RCC subtypes may be extracted directly from H&E-stained WSI and may complement gene expression-based patient stratification and selection strategies.
Abstract 8539: June 3rd, 2024 - 1:30-4:30
Correlation of immune phenotypes derived from H&E-stained whole slide images with prognosis and response to checkpoint inhibitors in NSCLC.
Collaborator: Incendia Therapeutics
The classification of tumors as inflamed, excluded or desert based on spatial patterns of tumor infiltrating lymphocytes (TILs) is a potential biomarker of patients likely to respond to checkpoint inhibitors (CPI). In this abstract, in collaboration with Incendia Therapeutics, PathExplore IOP is used to classify immune-phenotype (IP) based on patch level TIL distribution in tumor core and periphery from H&E images. Survival analysis indicates Immune inflamed phenotype is associated with improved PFS in CPI-treated NSCLC patients independent of PD-L1 status.
