L1 cell adhesion molecule may be a protective molecule for atrial fibrillation in patients with valvular heart disease.

ABSTRACT

Background: Atrial fibrillation (AF) is the most prevalent sustained arrhythmia. L1 cell adhesion molecule (L1CAM) served as a crucial regulator of signaling pathways. This research sought to examine the clinical value and functions of soluble L1CAM in the serum of AF patients. Methods: In total, 118 patients (valvular heart disease patients [VHD, total n = 93; AF n = 47; sinus rhythm (SR) n = 46] and healthy controls [n = 25]) were recruited in this retrospective study. Plasma levels of L1CAM were detected by enzyme-linked immunosorbent assays. The Pearson's correlation approach, as applicable, was used for analyzing the correlations. The L1CAM was shown to independently serve as a risk indicator of AF in VHD after being analyzed by the multivariable logistic regression. To examine the specificity and sensitivity of AF, receiver operating characteristic (ROC) curves and the area under the curve (AUC) were used. A nomogram was developed for the visualisation of the model. We further evaluate the prediction model for AF using calibration plot and decision curve analysis. Results: The plasma level of L1CAM was substantially decreased in AF patients as opposed to healthy control and SR patients (healthy control = 46.79 ± 12.55 pg/ml, SR = 32.86 ± 6.11 pg/ml, AF = 22.48 ± 5.39 pg/ml; SR vs. AF, P < 0.001; control vs. AF, P < 0.001). L1CAM was significantly and negatively correlated with LA and NT-proBNP (LA r = -0.344, P = 0.002; NT-proBNP r = -0.380, P = 0.001). Analyses using logistic regression showed a substantial correlation between L1CAM and AF in patients with VHD (For L1CAM, Model 1 OR = 0.704, 95%CI = 0.607-0.814, P < 0.001; Model 2 OR = 0.650, 95% CI = 0.529-0.798, P < 0.001; Model 3 OR = 0.650, 95% CI = 0.529-0.798, P < 0.001). ROC analysis showed that inclusion of L1CAM in the model significantly improved the ability of other clinical indicators to predict AF. The predictive model including L1CAM, LA, NT-proBNP and LVDd had excellent discrimination and a nomogram was developed. The model had good the calibration and clinical utility. Conclusion: L1CAM was shown to independently serve as a risk indicator for AF in VHD. In AF patients with VHD, the prognostic and predictive effectiveness of models incorporating L1CAM was satisfactory. Collectively, L1CAM may be a protective molecule for atrial fibrillation in patients with valvular heart disease.