Development and Validation of a Predictive Model Based on 17 Genes
A research team at Seoul National University Hospital has become the first in the world to identify a new molecular characteristic that can predict 'hematogenous metastasis'-the spread of gastric cancer cells through the bloodstream to organs such as the liver, lungs, bones, and adrenal glands-at an early stage. The team classified gastric cancer into two subtypes and developed a predictive model using 17 genes, providing a basis for assessing each patient's risk of metastasis in advance.
(From left) Professor Dojung Park, Department of Gastrointestinal Surgery, Seoul National University Hospital, Professor Hyesung Lee, Department of Pathology, Seoul National University Hospital. Seoul National University Hospital
The research team, led by Professor Dojung Park of the Department of Gastrointestinal Surgery and Professor Hyesung Lee of the Department of Pathology at Seoul National University Hospital (with co-first authors Clinical Instructor Seungho Lee and Researcher Jaeun Yoo), conducted a detailed analysis of tumor tissues from 64 gastric cancer surgery patients. They identified molecular subtypes directly associated with hematogenous metastasis and developed and validated a 17-gene-based model to predict the risk of metastasis for individual patients, announcing their findings on September 24.
Gastric cancer is known as one of the most common cancers worldwide. The key factor affecting patient survival is metastasis. Metastasis is largely classified into lymph node, peritoneal, and hematogenous metastasis, with hematogenous metastasis being associated with a poor prognosis. However, until now, it has not been possible to determine in advance which patients are vulnerable to hematogenous metastasis.
The research team performed bulk RNA sequencing using RNA extracted from patient tumors and classified gastric cancer into two subtypes based on gene expression patterns. As a result, they identified a 'stem cell-like' subtype with a high risk of metastasis and a 'gastric mucosal' subtype with a relatively lower risk of metastasis.
In the cohort analysis of 64 patients, those with the stem cell-like subtype had a significantly shorter hematogenous metastasis-free survival (HMFS) (P=0.005), while there was no difference between the two subtypes regarding peritoneal metastasis and overall survival (OS). Multivariate Cox analysis in the same cohort also confirmed that the stem cell-like subtype was an independent prognostic factor, increasing the risk of hematogenous metastasis by approximately 2.9 times.
The team then used a machine learning-based survival model to select 17 key genes closely related to hematogenous metastasis (10 for the stem cell-like subtype and 7 for the gastric mucosal subtype) and developed a 'hematogenous metastasis risk score' that quantifies the risk for each patient. This score is calculated by subtracting the expression values of gastric mucosal genes from those of stem cell-like genes, and patients are classified as high-risk (≥0.15) or low-risk (<0.15) based on a threshold of 0.15.
Validation using three external cohorts (TCGA, GSE66229, GSE84437) and 51 patient-derived xenograft (PDX) models showed that the high-risk group had a significantly shorter HMFS (P=0.002, 0.04, 0.03 for each cohort; PDX P=0.0001).
Furthermore, in patients classified as high-risk for hematogenous metastasis, adjuvant chemotherapy after surgery did not show the expected effectiveness, indicating the need to improve future treatment strategies. Accordingly, the research team analyzed the international cancer cell database (CCLE) and suggested Olutasidenib (an IDH1 inhibitor), MMV-390048 (a PI4K inhibitor), and IACS-10759 (an oxidative phosphorylation inhibitor) as potential therapeutic candidates for which high-risk cancer cells may be sensitive. However, they noted that further preclinical and clinical validation of these drugs is required.
Professor Dojung Park stated, "This study is significant as it is the first in the world to identify molecular subtypes directly associated with hematogenous metastasis in gastric cancer and to develop a predictive model that can assess individual patients' risk of metastasis at an early stage. The ability to precisely evaluate each patient's metastatic risk is expected to provide a crucial foundation for establishing personalized treatment strategies and developing new therapeutics."
The results of this study were published in the latest issue of the international journal, International Journal of Surgery (IF 10.1).
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
