The presentations feature Lunit SCOPE, an AI software that analyzes tissue slide images, developed by Lunit. One finding states that Lunit SCOPE IO—one of Lunit SCOPE product lines—can be used as a new biomarker for the treatment of multiple cancer types. The software analyzes tumor-infiltrating lymphocytes(TIL) in cancer patients’ tissue slides, assigning a score to each criterion. Upon validation with real-world data, the results showed that the higher the score, the better the response to immune checkpoint inhibitor (ICI) treatment, a subtype of immunotherapy.
TIL is known as a potential tumor agnostic biomarker for ICI therapy. Lunit has been validating the clinical application of Lunit SCOPE IO for predicting ICI treatment outcomes in advanced lung cancer. This study expanded the clinical application of Lunit SCOPE IO across multiple cancer types. The study was verified by more than 1,000 real-world patient data of 9 cancer types, which were collected from major institutions including Stanford University Medical Center.
“With multiple researches and long-term studies, we have been validating the effectiveness of an AI-based tissue analysis platform called ‘Lunit SCOPE’ that can help predict a cancer patient’s response to immunotherapy,” said Chan-young Ock, Chief Medical Officer of Lunit.
Lunit will also present a study on Lunit SCOPE PD-L1. AI analysis of PD-L1 expression can generate objective quantification, which can lead to accurately finding subjects for ICI therapy among non-small cell lung cancer (NSCLC) patients.
“Currently, pathologists interpret tissue slides through the naked eye. They assess PD-L1 expression level, which is the current standard for clinical application, but there are limitations. With the help of Lunit SCOPE PD-L1, which was trained with data including PD-L1 expression results of 380,000 cancer cells, we are able to find more patients who would respond to ICI therapy, 50% more, according to our study,” said Ock.
Lunit also announced that one of its abstracts about assessment of breast cancer risk has been selected for the ASCO 2021 ‘Poster Discussion Session’. Around 20% among all poster presentations are selected for this session, through strict review by the ASCO committee.
According to this study, unique parenchymal pattern with future breast cancer risk among breast cancer patients was identified by AI. Among breast cancer patients who developed cancer on one of the breasts, images from the other ‘normal’ side were collected and labeled as ‘high risk’. After training the AI with this dataset, the algorithm was validated on more than 4,000 external cases. The results showed that Lunit’s AI distinguished between high-risk and normal tissue with high accuracy, showing potential to be used as an independent biomarker to select high-risk populations based on mammography alone.
“Through continuous research, Lunit has been presenting groundbreaking findings at ASCO since 2019,” said Brandon Suh, CEO of Lunit. “It is considered a remarkable achievement for a medical AI startup to present four abstracts at such a renowned global medical conference. We are now focusing on the next stage, going beyond academic research to productization of our AI softwares to be used in actual cancer research, and eventually, clinical practice. We are planning to formally launch Lunit SCOPE products this year, taking steps on setting our AI as the new standard of cancer treatment.”
Marking its 57th anniversary, ASCO 2021 will be held online from June 4 to 8 due to COVID-19. More than 45,000 medical professionals from over 150 countries worldwide take part in the meeting to share the latest R&D achievements and insights related to cancer treatment.
Lunit Abstract Information at ASCO 2021
Abstract #2607
Title: Artificial intelligence-powered spatial analysis of tumor-infiltrating lymphocytes predicts survival after immune checkpoint inhibitor therapy across multiple cancer types.
Session Title: Developmental Therapeutics—Immunotherapy
Session Type: Poster Session
Online Demo: demo.scope.lunit.io/io
Abstract #9026
Title: Clinical performance of artificial intelligence-powered annotation of tumor cell PD-L1 expression for treatment of immune-checkpoint inhibitor (ICI) in advanced non-small cell lung cancer (NSCLC).
Session Title: Lung Cancer—Non-Small Cell Metastatic
Session Type: Poster Session
Online Demo: demo.scope.lunit.io/pdl1
Abstract #1568
Title: AI-based imaging biomarker in mammography for prediction of tumor invasiveness.
Session Title: Care Delivery and Regulatory Policy
Session Type: Poster Session
Abstract #10519
Title: Development of AI-powered imaging biomarker for breast cancer risk assessment on the basis of mammography alone.
Session Title: Prevention, Risk Reduction, and Hereditary Cancer
Session Type: Poster Discussion Session
Contact：
Yunyoung Noh, Communications Manager
+82-10-6460-9316
[email protected]
SOURCE Lunit
