AI Accurately Predicts Genetic Mutations from Routine Pathology Slides for Faster Cancer Care

Current cancer treatment decisions are often guided by genetic testing, which can be expensive, time-consuming, and not always available at leading hospitals. For patients with lung adenocarcinoma, a critical step in treatment is genetic testing to detect mutations in the tumor’s DNA. These mutations help doctors select personalized treatments, but the genetic tests required are not only costly but also slow and can create barriers to timely care. Now, a new study suggests that artificial intelligence (AI) has the potential to streamline cancer care by accurately predicting genetic mutations from routine pathology slides, reducing the need for rapid genetic testing and improving the efficiency of clinical decision-making.

Researchers at the Icahn School of Medicine at Mount Sinai (New York, NY, USA) and Memorial Sloan Kettering Cancer Center (New York, NY, USA) have developed an AI model that can predict genetic mutations using standard H&E-stained pathology slides. This innovative approach was trained using a large dataset of lung adenocarcinoma pathology slides matched with next-generation sequencing results. The model analyzes these slides, typically used for routine diagnostic biopsies, to detect mutations like those in the EGFR gene. By identifying EGFR mutations, which are crucial for selecting targeted therapies, the AI model can potentially reduce the reliance on expensive and time-consuming genetic tests, offering a faster and more accessible alternative for personalized cancer treatment.

Image: A real-time trial has shown that AI could speed cancer care (Photo courtesy of Campanella, et al., Nature Medicine)

The AI model was tested on live patient samples at Memorial Sloan Kettering Cancer Center as part of a real-time "silent trial." The findings, published in Nature Medicine, showed that the AI could reliably predict EGFR mutations and reduce the need for rapid genetic tests by more than 40%. This is a significant step forward in integrating AI into cancer diagnostics. The researchers plan to expand this AI model to detect other cancer-related mutations and broaden its applications, aiming for regulatory approval and eventual clinical use in diverse healthcare settings. This could ultimately lead to faster diagnosis, more personalized treatment, and better patient outcomes.

“Our findings show that AI can extract critical genetic insights directly from routine pathology slides,” said Gabriele Campanella, PhD, Assistant Professor of the Windreich Department of Artificial Intelligence and Human Health at the Icahn School of Medicine at Mount Sinai. “This could streamline clinical decision-making, conserve valuable resources, and accelerate patients’ access to targeted therapies by reducing reliance on certain rapid genetic tests.”

Related Links:
Icahn School of Medicine at Mount Sina 
Memorial Sloan Kettering Cancer Center Sinai