For HER2 clinical categories, accuracy improved from 66.7 percent without AI to 88.5 percent with AI
By Elana Gotkine HealthDay Reporter
FRIDAY, May 30, 2025 (HealthDay News) — Artificial intelligence (AI)-assisted training improves pathologist accuracy in human epidermal growth factor receptor 2 (HER2) immunohistochemistry (IHC) scoring in patients with breast cancer, according to a study presented at the annual meeting of the American Society of Clinical Oncology, held from May 31 to June 4 in Chicago.
David Mulder, from Mindpeak GmbH in Hamburg, Germany, and colleagues developed a training platform for AI-supported digital HER2 IHC assessment of breast cancer samples for pathologists. One hundred five pathologists participated in masterclass sessions, assessing 20 digital HER2 IHC stained breast cancer cases, both with and without AI assistance. The masterclasses included exam A, a lecture on HER IHC scoring, exam B, discussion of results, and AI-assisted exam C.
The researchers found that pathologists achieved an average agreement of 76.3 percent with reference scores across 1,940 readings without AI (exams A and B), compared to 89.6 percent on exam C (with AI assistance). Accuracy improved from 66.7 percent without AI to 88.5 percent with AI for HER2 clinical categories (null, ultralow, low, positive). In 29.5 percent of readings without AI, misclassification of HER2-ultralow cases as HER2 null occurred; with AI assistance, this decreased to 4.0 percent.
“Our study provides the first multinational evidence that artificial intelligence can help close a critical diagnostic gap and open the door to new therapies like antibody-drug conjugates for a majority of patients who, until recently, had not been offered these options,” lead author Marina De Brot, M.D., Ph.D., from the Camargo Cancer Center in São Paolo, Brazil, said in a statement.
Several authors disclosed ties to biopharmaceutical companies, including AstraZeneca, which funded the study.
Copyright © 2025 HealthDay. All rights reserved.
