'Lunit Scope IO' Demonstrates Possibility of Predicting Colorectal Cancer Recurrence Through AI

Medical AI company Lunit announced on the 27th that the research results predicting colorectal cancer recurrence and treatment response using the AI biomarker ‘Lunit SCOPE IO’ have been published in ‘npj Precision Oncology (IF 10.123)’, a sister journal of the world-renowned scientific journal Nature.

Lunit's 'Lunit Scope IO'
[Photo by Lunit]

Tumor-infiltrating lymphocytes (TIL) refer to lymphocytes gathered around cancer cells that have the function of effectively attacking tumors. Therefore, TIL density is considered an important biomarker indicating cancer treatment prognosis. However, it has been pointed out that manually measuring TIL density by medical staff takes a long time and may result in measurement differences between observers, presenting limitations. Accordingly, TIL evaluation using AI technology to solve this issue is gaining attention as a new alternative.

In this study, the densities of intratumoral tumor-infiltrating lymphocytes (iTIL) and stromal tumor-infiltrating lymphocytes (sTIL) in colorectal cancer patients were evaluated using Lunit SCOPE IO to confirm the predictability of cancer recurrence and survival outcomes. The analysis used slide images from 289 stage 2 and 3 colorectal cancer patients treated at Bundang Seoul National University Hospital between 2009 and 2012.

The study results showed that among the 28 patients whose cancer recurred, sTIL density was significantly lower. Additionally, when patients were divided into four groups based on sTIL density, the top 25% group with the highest density showed the lowest 5-year recurrence rate (1.4%), while the bottom 25% group had a recurrence rate of 17.2%. This indicated a tendency for higher recurrence rates with lower sTIL density and a significant association with recurrence-free survival (RFS). The company explained, "These results confirm the potential of sTIL density measurement using Lunit SCOPE IO as an effective tool for predicting colorectal cancer recurrence."

Furthermore, when patients were classified into three groups?high-risk, intermediate-risk, and low-risk for recurrence?based on Lunit SCOPE IO analysis results, the low-risk group showed a hazard ratio (HR) of 0.11 compared to the high-risk group, suggesting that the classification system using Lunit SCOPE IO could be effective for prognosis prediction.

Seobum Seok, CEO of Lunit, stated, “The recurrence rate for stage 2 and 3 colorectal cancer patients is known to be about 20-30%, and AI-based recurrence prediction will play an important role in improving the survival rate and quality of life of these colorectal cancer patients.” He added, “This study provides important information for personalized treatment strategies and clinical decisions, and Lunit will continue to lead the way in proposing new directions for cancer treatment using AI.”

This content was produced with the assistance of AI translation services.