Constructing prognostic risk prediction model for atrial fibrillation patients based on the TGF-β1/Smads signaling pathway

HUANG Mao-zhi; ZHENG Jian-ping; YOU Chao-qun; HUANG Qi-lei; CHENG Xian-lu

doi:10.3969/j.issn.1672-5301.2025.02.002

Abstract：Objective To evaluate the risk of poor prognosis in patients with atrial fibrillation(AF)based on the transforming growth factor-β1(TGF-β1)/Smads signaling pathway construct a predictive model accordingly.Method A total of 120 AF patients admitted to Nanping First Hospital from August 2022 to August 2023 were selected as the research subjects,all of whom underwent radiofrequency ablation treatment for the first time.The patients were followed up for 12 months after the surgery and were classified into the poor prognosis group and the good prognosis group based on the recurrence of postoperative AF.The general information and expression levels of TGF-β 1/Smads signaling pathway related factors were compared between two groups.Multivariate Logistic regression model was established to analyze the influencing factors of poor prognosis in patients with AF.Based on the related factors of TGF-β1/Smads signaling pathway,the prognostic risk prediction model of AF patients was constructed and the performance of the model was verified by internal data.At the same time,the clinical decision curve was drawn to analyze its clinical applicability.Results Out of 120 patients,a total of 118 completed follow-up work and 2 withdrew.During the 12-month postoperative period,41 out of 118 patients experienced AF recurrence,accounting for 34.75％.The left atrial diameter(LAD)in the poor prognosis group was higher than that in the good prognosis group;the levels of TGF-β1/Smads signaling pathway related factors[serum galectin-3(Gal-3),TGF-β1,and phosphorylated cell signaling molecule Smad3(p-Smad3)]were higher in the poor(admission)group than in the good prognosis group,with statistical significant differences(P＜0.05).A multivariate Logistic regression model showed that high levels of preoperative serum Gal-3,TGF-β1 and p-Smad3 were the prognostic risk factors for AF patients(OR＞1,P＜0.05).A prognostic risk prediction model was constructed for AF patients based on the results of Logistic regression analysis,and the consistency index(C-index=0.966)of the model discrimination was verified,indicating a good model discrimination.A standard curve display was drawn and the calibration curve fitted well with the ideal curve;the model accuracy was good.The result of drawing the decision curve showed that the net benefit rate within the threshold range of 0.00-1.00 was always greater than 0,indicating good clinical practicality.Conclusion The abnormally high levels of TGF-β 1/Smads signaling pathway related factors before surgery are the risk factors for poor prognosis in AF patients.The prognostic risk column chart prediction model for AF patients based on TGF-β1/Smads signaling pathway related factors has good accuracy and discrimination,and high clinical practicality.