While targeted therapies such as FGFR2 inhibitors can offer superior survival outcomes for patients with intrahepatic cholangiocarcinoma (iCCA), it’s estimated that no more than 20% of patients contain the alterations that make them eligible for targeted therapy. Screening for genetic alterations can also be costly for patients and is not commonly utilized.
A recent study published in the British Journal of Cancer has detailed the development of a histopathology-based artificial intelligence (AI)-assisted system to predict genetic alterations in patients with iCCA.
The Genetic Alteration Prediction (GAP) system was developed using multi-instance and self-supervised learning to predict genetic alterations using 2069 whole-slide images (WSIs) from hematoxylin and eosin-stained slides taken from 232 patients with iCCA who had undergone surgery. Another group of 150 patients were used as independent external validators.
The GAP system was found to predict actionable genetic alterations including FGFR2 and IDH, with area under the receiver operating characteristic curves of 0.754 and 0.713, respectively. A decision curve analysis also showed that an AI-assisted strategy can offer better clinical benefits.
This newly-developed AI-assisted genetic alteration screening system can assist in the precise targeted treatment of advanced iCCA.
© 2025 Mashup Media, LLC, a Formedics Property. All Rights Reserved.