A team of Dana-Farber researchers has identified a potential new way to assess clinically valuable features of clear cell renal cell carcinoma (ccRCC; pictured) using image processing with deep learning.
An AI-based assessment tool evaluates two-dimensional pictures of a tumour sample on a pathology slide and identifies previously underappreciated features, such as tumour microheterogeneity, that could help predict whether a tumour will respond to immunotherapy. Their results suggest that pathology slides contain important biological information about ccRCC tumours – and possibly all types of tumours – that could be valuable for understanding more about the biology of the cancer.
The work, which is described in Cell Reports Medicine, is part of a broader effort at Dana-Farber – a world leading centre for cancer research and treatment - to use AI in biologically grounded ways to transform cancer care and cancer discovery. “This is an example of the growing convergence of AI and cancer biology,” says co-senior author Eliezer Van Allen, MD, Chief of the Division of Population Sciences at Dana-Farber. “It represents a major opportunity to measure key features of the tumour and its immune microenvironment at the same time. These measures could help drive not only biological discovery but also potentially guide cancer care.”
Renal cell carcinoma is among the 10 most common cancers worldwide. The clear cell subtype (ccRCC) accounts for 75-80% of metastatic cases. Some tumours are sensitive to immune checkpoint inhibitors (ICIs), but so far there are no measures that predict whether a ccRCC tumour will respond to immunotherapy with an ICI. As part of diagnosis, pathologists analyse pathology slides of tumour samples that have been stained to reveal the structures of cells. A routine measure is nuclear grade, which indicates how far tumour cells deviate from normal cells.
Nyman, who collaborated with Van Allen, Dana-Farber pathologist Sabina Signoretti, MD, and Toni Choueiri, MD, Director of the Lank Center for Genitourinary Oncology at Dana-Farber, on the project, first trained an AI model to assess a tumour’s nuclear grade. The AI model was not only able to assess nuclear grade, but also to identify differences in grade across a tumour sample.
The finding inspired the team to expand their deep learning model to quantify tumour microheterogeneity and immune properties, such as immune infiltration, across the slide. They found that features such as tumour microheterogeneity and immune infiltration were associated with improved overall survival among patients taking immune checkpoint inhibitors. The tumours that responded to ICIs had both higher levels of tumour microheterogeneity and denser infiltration of lymphocytes in high-grade regions.
The tool is not ready for clinical use, but as a next step, the team is testing it in an ongoing clinical trial involving combination immunotherapy as first-line treatment in patients with ccRCC. The team also plans to explore whether these visual clues in pathology slides are related to molecular features of the tumour, such as alterations in genes.
- Nyman J, Denize T, Bakouny Z, et al. Spatially aware deep learning reveals tumour heterogeneity patterns that encode distinct kidney cancer states. Cell Reports Med.2023;4(9) doi:10.1016/j.xcrm.2023.101189
