Lunit AI Improves HER2 Therapy Outcome Predictions in Metastatic Colorectal Cancer

Lunit, a leader in AI-powered cancer diagnostics, announced the publication of a new study in the Journal of Clinical Oncology Precision Oncology. In collaboration with Japan's National Cancer Center Hospital East (NCCHE), the study reveals how Lunit's AI-powered pathology solutions, Lunit SCOPE® HER2 and Lunit SCOPE IO, significantly improve HER2 biomarker evaluation and clinical outcome predictions for metastatic colorectal cancer (mCRC) patients undergoing HER2-targeted therapy. The study, based on the TRIUMPH phase II clinical trial, assessed 30 HER2-positive mCRC patients treated with Trastuzumab and Pertuzumab.

Lunit SCOPE HER2 demonstrated an impressive 86.7% accuracy in assessing HER2 immunohistochemistry (IHC) compared to pathologist evaluations, with 100% accuracy in identifying HER2 IHC 3+ cases. Moreover, patients identified by the AI model as having a high proportion of HER2 IHC 3+ tumor cells (AI-H3-high, >50%) exhibited notably better clinical outcomes than those identified through traditional HER2 evaluation methods. The objective response rate (ORR) for AI-H3-high patients was 42.1%, compared to 26.7% for the overall TRIUMPH trial. Additionally, AI-H3-high patients had a progression-free survival (PFS) of 4.4 months, versus 1.4 months for those identified through traditional methods, and an overall survival (OS) of 16.5 months, compared to 4.1 months for AI-H3-low patients.

The study also applied Lunit SCOPE IO for detailed tumor microenvironment (TME) profiling, including lymphocyte, macrophage, and fibroblast densities. Among AI-H3-high patients, those with low stromal TME density (TME-low) showed the most favorable clinical outcomes, with an ORR of 57.1%, PFS of 5.6 months, and OS of 26.0 months.

These findings underscore the transformative potential of AI-powered pathology tools in precision oncology. Lunit's solutions allow for more accurate and detailed evaluation of HER2 status and TME characteristics, enabling better stratification of patients and more reliable predictions of responses to HER2-targeted therapies. This capability can inform more tailored treatment strategies, ultimately improving patient outcomes in mCRC and potentially other HER2-amplified cancers.

Dr. Takayuki Yoshino of NCCHE, the study’s principal investigator, emphasized that AI technology has the potential to redefine how biomarkers are evaluated and treatment responses predicted. Lunit CEO Brandon Suh added that the findings showcase how Lunit's AI-powered solutions provide clinicians with actionable insights to refine treatment strategies, demonstrating AI’s transformative role in precision oncology.

Click here to read the original news story.