A new study examines the development of two machine learning models to classify the immunophenotype of a cancer specimen.
The digital pathology approach presented can characterise and classify cancer immunophenotypes in a reproducible and scalable fashion, holding promise for the application of such a method to identify patients that may benefit from immunotherapy in non-small cell lung cancer (NSCLC), according to the study published in the peer-reviewed journal AI in Precision Oncology.
The cellular composition of the tumour immune microenvironment is a key contributor to the response of the tumour to immunotherapy.
TGF-ß signaling is known to promote immune-exclusion, where CD8+ T cells are in the surrounding stromal tissue but not within the tumour itself.
PODCAST EXCLUSIVE TO THE ONCOLOGY NETWORK | LISTEN NOW
To better identify patients who are immune-excluded, Rui Wang, from Sanofi, and coauthors, developed two machine learning models to quantify CD8+ cell positivity and classify the immunophenotype of a cancer specimen in patients with NSCLC.
“Our results support the potential use of machine learning-predicted cancer immunophenotypes to identify patients that may benefit from immunotherapy and/or TGF-ß blockage in NSCLC,” concluded the investigators.
“This research points towards improvements in patient identification for drug candidacy, utilising AI and machine learning to pinpoint precise biomarkers for immunotherapy in NSCLC. It signifies progress towards personalised medicine, promising treatments tailored to individual patient profiles for greater effectiveness and minimised side effects. Essentially, it emphasises the importance of directing new treatments to the right patients, paving the way for a new era of precision in cancer care,” says Douglas Flora, MD, Editor-in-Chief of AI in Precision Oncology.
