South Korea’s Lunit Teams Up with AstraZeneca, Creating AI for Lung Cancer Diagnostics
South Korea-based medical AI developer Lunit (Seoul, South Korea) has partnered with AstraZeneca (Cambridge, UK) to develop innovative AI-powered digital pathology solutions to predict the presence of druggable driver mutations in non-small cell lung cancer (NSCLC).
AstraZeneca will incorporate Lunit’s AI-powered Scope Genotype Predictor into its pathology workflow. This technology aims to simplify genomic testing and help identify key mutations that can guide research and treatment for cancers like NSCLC.
Genomic testing for cancer is often resource-intensive and time-consuming, delaying urgent treatment. To address this, AstraZeneca and Lunit are developing an AI-based screening tool that uses deep learning algorithms to analyze Hematoxylin and Eosin (H&E) slide images to predict the presence of “druggable genomic alterations” within the patient’s cells.
H&E is a type of staining commonly used in microscopy to visualize tissue samples. Hematoxylin stains cell nuclei a blue or purple color, while Eosin stains the cytoplasm and other tissue components pink or red, helping to clearly distinguish the structure of organelles. This common staining method allows the AI to identify patterns and abnormalities associated with certain genetic mutations.
The tool aims to predict the likelihood of tumor harboring driver mutations in genes such as epidermal growth factor receptor (EGFR), which has been associated with NSCLC. The EGFR gene codes for proteins that are involved in cell growth and division. Mutations in EGFR can lead to uncontrolled cell growth, which is a hallmark of cancer.
“We are excited to partner with AstraZeneca, a leader in oncology therapeutics, to develop and evaluate this groundbreaking technology,” added Brandon Suh, CEO of Lunit.
“The integration of Lunit SCOPE Genotype Predictor as a screening test into pathology workflows promises to improve the opportunity for patients to benefit from appropriate targeted therapy, ultimately improving patient outcomes and streamlining the treatment planning process.”
The results generated by Lunit would be completed faster than molecular tests, enabling physicians to prioritize further targeted molecular testing (using standard molecular testing techniques) of tumor samples on patients that have a high probability of harbouring druggable driver mutations, such as the EGFR mutation in NSCLC.
Some of these targeted molecular tests include next-generation sequencing, polymerase chain reaction, or fluorescence in situ hybridization. By streamlining the process, Lunit enables faster identification of actionable mutations, allowing physicians to tailor treatments for a more personalized patient experience.
“This collaboration with Lunit underscores our commitment to advancing precision medicine in oncology,” said Kristina Rodnikova, Head of Global Oncology Diagnostics, Oncology Business Unit at AstraZeneca.
“Tools like this will help to address unmet needs by optimizing diagnostic workflows for NSCLC patients and, ultimately, improve their outcomes.”
Once their joint solution is validated, the two companies plan to deploy it in real-world settings to enhance molecular testing for NSCLC.
