Seoul National University Hospital Develops Biomarker to Identify Acute Myeloid Leukemia Subtypes

From the left, Seongsoo Yoon and Youngil Ko, professors of Hematology and Oncology at Seoul National University Hospital, and Hongseok Yoon and Seongyoung Lee, professors of Clinical Genomics.

[Asia Economy Reporter Kim Young-won] A research team at Seoul National University Hospital has presented a biomarker that identifies subtypes of acute myeloid leukemia (AML). Utilizing this biomarker is expected to enable classification of AML patients into three distinct subtypes, allowing for tailored treatments based on their characteristics.

The research team, including Professors Yoon Sung-soo and Ko Young-il from the Department of Hematology and Oncology, and Professors Yoon Hong-seok and Lee Sung-young from the Department of Clinical Genomic Medicine at Seoul National University Hospital, announced on the 25th that they developed a predictive biomarker to identify AML subtypes based on genomic analysis data from 1,350 patients and verified its performance.

Myeloid leukemia consistently accounts for about 60% of leukemia cases in Korea, making it the most prevalent blood cancer domestically. Among these, acute myeloid leukemia, which occurs more frequently in adults, is divided into several types based on biological characteristics. Since treatment response and prognosis vary by subtype, optimized treatment for patients is crucial.

The Seoul National University Hospital research team hypothesized that the biological subtype of AML is determined by the interaction of the BCL2 gene and various related genes, rather than a single gene. Until now, most AML research has focused on individual genes targeting single-gene therapies. Representative genes include the intrinsic apoptosis regulator BCL2 and anti-apoptotic genes MCL1 and BFL1.

To validate their hypothesis, the team applied a non-negative matrix factorization technique to transcriptome data from 1,350 AML patients to extract inherent patterns in the biological data. As a result, they discovered a biomarker called 'BFsigs' that classifies AML into three subtypes (BCL2, MCL1/BCL2, BFL1/MCL1) based on patterns of BCL2-related genes.

Applying BFSigs to three different data cohorts for classification confirmed that all AML patient data could be categorized into the three subtypes proposed by the research team.

Classification of acute myeloid leukemia subtypes using BFSigs markers. All patient data included in each cohort were well divided into three subtypes (purple rectangles).

Furthermore, each subtype classified by BFSigs showed different treatment responses to BCL2-targeted therapies for AML. The research team explained that this is due to the different gene pathways activated in each subtype.

Based on this insight, the team developed a predictive biomarker using BFSigs to forecast treatment responses to AML therapies. Verification of the predictive biomarker’s performance using patient data with known treatment outcomes showed very high accuracy. Treatment responses could be predicted with 87% and 95% accuracy in two separate cohorts, respectively.

The research team also validated the clinical applicability of the BFSigs-based predictive biomarker by utilizing Nanostring technology, which is used for genetic analysis in clinical settings. Analysis of gene patterns from 47 AML patients treated at Seoul National University Hospital using Nanostring technology showed that actual patient genetic data were well classified into the three BFSigs-based subtypes.

Professor Yoon Hong-seok stated, "We successfully confirmed through various validations that the BFSigs-based predictive biomarker has the potential to identify biological subtypes of AML patients," adding, "We expect that further development of this research will enable precision medicine optimized for patients."

Professor Yoon Sung-soo said, "This study is the result of an in-house multidisciplinary research network involving clinicians, physician-scientists, and researchers," and added, "In the future, it will be possible to establish strategies for utilizing BCL2 family genes in AML treatment easily and effectively."

Meanwhile, this research was conducted with support from the Ministry of Health and Welfare’s Research-Centered Hospital Project, and the results were published in the latest issue of the international journal Genome Medicine.

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