Integrin subunit alpha 2 gene polymorphisms,susceptibility to chronic total occlusion of coronary arteries,and disease prediction model:An academic investigation
LYU Tong-wei
ZHANG Jing-yu
KAN Tong
HOU Pan
LI Pan
GUO Zhi-fu
Abstract：Objective To explore the association between integrin subunit alpha 2(ITGA2)gene single nucleotide polymorphisms(SNPs)and the susceptibility to chronic total occlusion(CTO)of the coronary arteries and to establish and optimize a risk prediction model for CTO occurrence based on the risk factors.Methods A single-center retrospective case-control design were employed.The case group(n=264)was selected from the First Affiliated Hospital of the Naval Medical University of the People's Liberation Army of China from August 1,2024 to December 1,2024,in accordance with the guidelines of the Chronic Total Occlusion Academic Research Consortium(CTO-ARC)and the Japanese CTO interventional expert consensus(J-CTO score≥2).The control group(n=236)was strictly defined using computer-assisted quantitative coronary analysis(QCA),excluding individuals with atherosclerosis(defined as stenosis＜20％and Agatston calcium score＜100).The entire cohort excluded individuals with active infections,malignancies,immune diseases and acute vascular lesions.Polymerase chain reaction(PCR)and genotyping techniques were used to analyze the selected SNP rs35235 in the ITGA2 gene.Statistical analysis was performed using R language version 4.2.2 to assess the association between different genotypes and the risk of CTO.Additionally,nine machine learning algorithms,including Support Vector Machine(SVM),Gradient Boosting Machine(GBM)and Random Forest(RF),were employed to establish and optimize a predictive model for CTO occurrence risk.Results The TT genotype of the SNP rs35235 in the ITGA2 gene was identified as a protective factor against CTO(OR=0.33,95％CI 0.16-0.66).ROC curve analysis demonstrated that the predictive performance of the SNP genotype combined with traditional risk factors(AUCl=0.84)was significantly improved compared to that of traditional risk factors alone(AUC2=0.826),with a statistically significant difference(P=0.028).Nine machine learning algorithms-Logistic Regression,SVM,GBM,Neural Network(NN),XGBoost,K-Nearest Neighbors(KNN),LightGBM,CatBoos,and RF-yielded AUC values of 0.845,0.803,0.950,0.879,0.867,0.948,0.932,0.926,and 0.908,respectively,for the CTO risk prediction models.The GBM-based model identified the top seven risk factors for CTO as:gender,low-density lipoprotein,total cholesterol,fibrinogen,SNP genotype,diabetes mellitus,and fasting blood glucose.Conclusion The TT genotype of the ITGA2 gene SNP rs35235 is the protective factor for CTO,providing a new perspective for the genetic study of CTO.As a genetic marker,SNP genotype is not affected by drug intervention,offering a more stable and reliable risk prediction indicator for clinical use.The GBM machine learning algorithm is identified as the best-fitting prediction model.
Keywords：Integrin subunit alpha 2Single nucleotide polymorphismsChronic total occlusionRisk prediction model
Publication Date：2025-06-20
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：7( 490-496 )