Development and Validation of a Nomogram for Estimating Stroke Risk in Nonvalvular Atrial Fibrillation Patients With Left Atrial Stiffness: A Single-Center Cross-Sectional Study.

摘要

BACKGROUND: Nonvalvular atrial fibrillation (NVAF) is a prevalent arrhythmia associated with an increased risk of ischemic stroke. This study is aimed at assessing the clinical significance of echocardiographic left atrial (LA) stiffness and developing a nomogram integrating clinical and echocardiographic parameters to evaluate individualized stroke risk in NVAF patients. METHODS: We enrolled 402 consecutive NVAF patients between January 2020 and March 2023. Clinical and echocardiographic data were collected, and patients were randomly assigned to training (n = 281) and validation (n = 121) datasets. Univariate and multivariate logistic regression identified independent stroke risk factors, which were used to construct a nomogram. Model performance was evaluated by receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis. The reclassification ability of the nomogram was assessed using integrated discrimination improvement (IDI) and net reclassification improvement (NRI). RESULTS: Logistic regression identified LA stiffness, LA reservoir strain, hypertension, peripheral vascular disease, and anticoagulation as independent factors associated with stroke. The nomogram based on these factors showed excellent discrimination, with a C-statistic of 0.889 (0.847-0.924) in the training cohort and 0.880 (0.808-0.932) in the validation cohort. Model A (LA stiffness, LA reservoir strain, hypertension, peripheral vascular disease, and anticoagulation) provided better net reclassification and integrated discrimination than Model B (LA stiffness, LA reservoir strain, hypertension, and peripheral vascular disease) despite both outperforming the CHA2DS2-VASc model. LA stiffness was inversely correlated with left atrial appendage (LAA) emptying (R = -0.52, p < 0.01) and filling velocities (R = -0.50, p < 0.01). Patients with LA thrombus or spontaneous echocardiographic contrast exhibited significantly higher LA stiffness (p < 0.05). CONCLUSION: A nomogram incorporating LA stiffness, LASr, hypertension, and peripheral vascular disease can effectively identify the risk of stroke in NVAF patients. Additionally, a secondary model including anticoagulation showed further improved performance but requires prospective validation. Elevated LA stiffness might play a role in the development of stroke in NVAF by influencing LA appendage hemodynamics.