Abstract
BACKGROUND: Tissue inhibitor of metalloproteinase-1 (TIMP-1) levels is strongly associated with cardiac extracellular matrix accumulation and atrial fibrosis. Whether serum levels of TIMP-1 are associated with atrial fibrillation (AF) recurrence following radiofrequency catheter ablation (RFCA) remains unknown. MATERIALS AND METHODS: Serum TIMP-1 levels of patients with AF before they underwent initial RFCA were measured using ELISA. Univariate and multivariate-adjusted Cox models were constructed to determine the relationship between TIMP-1 levels and AF recurrence. Multivariate logistic regression analyses were performed to determine predictors of AF recurrence. RESULTS: Of the 194 enrolled patients, 61 (31.4%) had AF recurrence within the median 30.0 months (interquartile range: 16.5-33.7 months) of follow-up. These patients had significantly higher baseline TIMP-1 levels than those without AF recurrence (129.8 ± 65.7 vs. 112.0 ± 51.0 ng/ml, P = 0.041). The same was true of high-sensitivity C-reactive protein (3.9 ± 6.0 vs. 1.9 ± 2.8 ng/ml, P = 0.001). When a TIMP-1 cutoff of 124.15 ng/ml was set, patients with TIMP-1 ≥ 124.15 ng/ml had a higher risk of recurrent AF than those with TIMP-1 < 124.15 ng/ml (HR, 1.961, 95% CI, 1.182-2. 253, P = 0.009). Multivariate Cox regression analysis revealed that high TIMP-1 was an independent risk factor for AF recurrence. Univariate Cox regression analysis found that substrate modification surgery does not affect AF recurrence (P = 0.553). Subgroup analysis revealed that female sex, age < 65 years, hypertension (HTN), body mass index (BMI) ≥ 24 kg/m(2), CHA2DS2-VASc score < 2, HAS-BLED score < 3, and EHRA score = 3 combined with high TIMP-1 level would perform well at predicting AF recurrence after RFCA. CONCLUSION: Elevated preoperative TIMP-1 levels are related to a higher risk of AF recurrence and can independently predict AF recurrence following RFCA.