Discovered a New 'Biomarker' to Gauge Success of Immunotherapy in Lung Cancer Treatment

Professor Lee Se-hoon’s Team at Samsung Seoul Hospital Proposes 'Tumor-Infiltrating Lymphocytes' as Auxiliary Indicator
Higher Immune Cell Density Increases Treatment Response Rate and Extends Survival Period

[Asia Economy Reporter Lee Gwan-joo] A new path has opened to predict in advance which patients will respond to lung cancer immunotherapy using tumor-infiltrating lymphocytes (TILs).

Researchers including Professors Lee Se-hoon and Park Se-hoon from the Department of Hematology-Oncology at Samsung Medical Center, Professor Choi Yoon-ra from the Department of Pathology, Professor Kim Hyo-jin from the Department of Pathology at Seoul National University Bundang Hospital, and Lunit's Chief Medical Officer Ok Chan-young announced on the 21st that using ‘Lunit Scope IO,’ it is possible to predict the effectiveness of immunotherapy in non-small cell lung cancer based on the distribution of tumor-infiltrating lymphocytes.

Recently, the insurance coverage for the lung cancer immunotherapy drug pembrolizumab has been expanded. Compared to existing treatments, it is known to extend patients' survival periods, but since treatment effects vary among patients, it must be used cautiously. Currently, the expression rate (TPS) of a specific protein called PD-L1 on cancer cells is used; if it is 50% or higher, immunotherapy alone is considered effective. However, PD-L1 alone is insufficient to identify patients who can expect treatment benefits, making the search for new ‘biomarkers’ urgent.

The research team focused on tumor-infiltrating lymphocytes (TILs), immune cells that reside in the tumor microenvironment and attack cancer tissue. Although these immune cells have been recognized for their potential as key indicators determining the success or failure of cancer treatment, it has been challenging to examine each finely divided cancer tissue to determine how many TILs are present.

As a solution, the team employed an artificial intelligence model. Referring to the pathological experts' readings analyzing lung cancer tissue, they trained the AI model to check the distribution of TILs in patients' cancer tissues instead of manually reviewing pathology slides. Using the AI model, the team classified the immunological phenotype of lung cancer into active and inactive (further subdivided into excluded and deficient) based on the density and distribution of TILs, then analyzed the effectiveness of immunotherapy in 518 patients treated at Samsung Medical Center and Seoul National University Bundang Hospital for each case.

As a result, patients with active TILs showed significantly better response rates to immunotherapy. This was consistent in independent cohorts from both Samsung Medical Center and Seoul National University Bundang Hospital, as well as in combined analyses. The median overall survival was also improved in active patients at 24.8 months, compared to 14 months for excluded patients and 10.6 months for deficient patients. Progression-free survival was longer in active patients at 4.1 months, versus 2.2 months for excluded patients and 2.4 months for deficient patients.

Notably, when conditions were the same, the effect was more pronounced in patients who received immunotherapy as first-line treatment. The median overall survival for active patients was 38 months, significantly longer than 11.9 months for inactive patients. Progression-free survival was also extended in active patients at 15.6 months, compared to 4.8 months for inactive patients. Professor Lee Se-hoon stated, “Finding the most suitable treatment for patients is the first step toward treatment success. If this new biomarker is used as an auxiliary tool, more patients can expect better treatment outcomes, and even those who might otherwise be overlooked can have treatment opportunities.”

This study was published in the recent issue of the ‘Journal of Clinical Oncology (IF=44.54),’ one of the world's top journals related to cancer.

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