AI Support Significantly Improves HER2 Scoring Accuracy in Breast Cancer Diagnosis
Key Insights
A new study published in JCO Precision Oncology demonstrates that Ibex Medical Analytics (search)' AI technology significantly improves pathologists' accuracy and consistency in HER2 (search) biomarker scoring for breast cancer (search).
The multi-center study showed AI-assisted pathologists achieved higher inter-observer agreement (83.7% vs 75%) and improved accuracy in identifying challenging HER2 (search)-low cases, which is critical for determining eligibility for new targeted therapies.
The "zero-click" decision support tool demonstrated exceptional robustness across multiple labs, antibodies, scanners, and patient demographics, offering particular value in distinguishing between HER2 (search) 0 and 1+ cases.
A clinical study published in JCO Precision Oncology has demonstrated that pathologists supported by Ibex Medical Analytics (search)' AI technology show significant improvement in accuracy and consistency when scoring HER2 (search) biomarkers in breast cancer (search) patients.
The study evaluated Ibex Breast HER2 (search), a "zero-click" decision-support tool that helps pathologists delineate HER2 expression into four standard scores (0, 1+, 2+, and 3+) based on the 2023 ASCO/CAP guidelines. The technology showed particular value in identifying challenging HER2-low cases, which has become increasingly important for treatment decisions.
Improved Accuracy and Consistency in HER2 Scoring
The multi-center study included 120 breast cancer (search) biopsies from medical centers across the US, Europe, and the Middle East. Researchers compared pathologists' performance when using Ibex Breast HER2 (search) versus standard manual HER2 quantification and scoring methods.
Key findings from the study include:
- Pathologists using AI support showed significantly higher inter-observer agreement (83.7%) compared to standard care (75%) across all slides
- For the challenging HER2 0 and 1+ slides specifically, agreement improved dramatically from 69.8% without AI to 87.4% with AI assistance
- When compared to expert consensus, pathologists' accuracy in scoring HER2 0 and 1+ slides improved from 81.9% to 88.8% with AI support
- The AI system demonstrated robust performance across different laboratories, HER2 antibodies, scanners, and patient demographics
Professor Savitri Krishnamurthy, Professor of Pathology and Laboratory Medicine at The University of Texas MD Anderson Cancer Center and principal investigator of the study, emphasized the value of the technology: "Our analysis shows the algorithm performs very well. The study provides evidence that AI can be a useful ancillary tool to aid pathologists in agreeing with expert breast pathologists, as well as creating more concordance amongst themselves for HER2 (search) protein expression at the lower end of the spectrum."
Clinical Significance for Breast Cancer Treatment
HER2 (search) is a protein responsible for division and proliferation of breast cancer (search) cells. Its accurate assessment is critical for identifying patients who may benefit from HER2-directed therapies, including recently approved treatments for HER2-low breast cancers.
The emergence of effective HER2 (search) antibody-drug conjugate therapies, such as Enhertu (trastuzumab deruxtecan), has increased the need for accurate and reproducible HER2 scoring, particularly at lower expression levels. Enhertu was recently approved by both the FDA and EMA for patients with HER2-low metastatic breast cancer (search) (HER2 IHC 1+ or 2+ / ISH-) and has shown efficacy in HER2-ultra-low patients.
Professor Stuart J. Schnitt, Chief of Breast Oncologic Pathology at Dana-Farber Brigham Cancer Center (search) and one of the study's expert breast pathologists, noted: "With the availability of new drugs to treat patients with low levels of HER2 (search) expression, there is a need for a computational image analysis and quantification solution. The AI identifies areas of invasive cancer and categorizes the different classes of HER2 protein expression very clearly."
Integration into Diagnostic Workflow
Ibex Breast HER2 (search) is part of Ibex Breast, an integrated AI solution that detects 54 tissue morphologies in breast H&E slides. The system has been adopted by pathology labs worldwide to facilitate rapid, consistent, and objective diagnosis of breast biopsies and excisions.
Dr. Manuela Vecsler, VP of Clinical and Scientific Affairs at Ibex Medical Analytics (search), commented on the study results: "Our AI-powered solutions are designed together with pathologists, for pathologists, enabling them to leverage cutting-edge technology to make more confident diagnosis. This study reaffirms our commitment to providing the most accurate and reliable tools for pathologists, ultimately improving patient care."
The technology appears particularly valuable in today's clinical environment, where there is increasing pressure to accurately identify HER2 (search)-low cases in a reproducible and objective manner. As new targeted therapies continue to emerge, the ability to precisely stratify patients based on biomarker expression becomes increasingly critical for optimal treatment selection.
Future Implications
The successful application of AI in HER2 (search) scoring represents a significant advancement in digital pathology and precision oncology. As more targeted therapies are developed that depend on precise biomarker quantification, AI-assisted diagnostic tools may become standard components of the pathology workflow.
The Ibex platform includes solutions that are CE-IVD cleared and registered with regulatory authorities in the UK, Australia, and Brazil, while it remains for Research Use Only in the United States pending FDA clearance.