We aim to assess the risk of thrombus-associated events (TAE) in patients with heart failure (HF) without atrial fibrillation (AF) and develop an effective scoring system for a risk stratification model. This retrospective study included 450 patients (median age 64.0years, interquartile range [55.0, 75.0]; 31.6% women) hospitalized for HF without AF and atrial flutter, but with a left ventricular ejection fraction (LVEF)≤55% and New York Heart Association (NYHA) functional class of III-IV. A median follow-up of 47months was conducted. In the present study, TAE during follow-up was independently associated with both all-cause death [hazard ratio (HR) 1.756, 95% confidence interval (CI) 1.324-2.328, P<0.001] and readmission for HF (HR 1.574, 95% CI 1.122-2.208, P=0.009) after adjustment for covariates. Hypertension (HR 1.573, 95% CI 1.018-2.429, P=0.041), atrial arrhythmia excluding AF (AAexAF) (HR 2.041, 95% CI 1.066-3.908, P=0.031), previous ischaemic stroke (HR 2.469, 95% CI 1.576-3.869, P<0.001), and vascular disease (HR 1.658, 95% CI 1.074-2.562, P=0.023) were independently associated with TAE. Age (HR 1.021, 95% CI 1.008-1.033, P=0.001), previous ischaemic stroke (HR 1.685, 95% CI 1.248-2.274, P=0.001), LVEF ([10, 25] vs. [40, 55]) HR 1.925, 95% CI 1.311-2.826, P=0.001; (25, 40] vs. (40, 55] HR 1.084, 95% CI 0.825-1.424, P=0.563), and creatinine clearance rate (Ccr) (HR 0.991, 95% CI 0.986-0.996, P=0.001) were independently associated with composite events of TAE and death (TAE-D). CHA2DS2VASc modestly predicted 5-year TAE [area under the receiver operating characteristic curves (AUC) 0.660, P<0.001 compared with 0.5] and TAE-D (AUC 0.639, P<0.001 compared with 0.5). (C)ACE, formed by incorporating AAexAF, LVEF, and Ccr into CHA2DS2VASc, had higher AUC for predicting 5-year TAE (0.694 vs. 0.660, P=0.018) and TAE-D (0.708 vs. 0.639, P<0.001) compared with CHA2DS2VASc. In patients with HF with reduced ejection fraction (HFrEF), (C)ACE and (C)ACEN [formed by incorporating NYHA into (C)ACE] had higher AUC compared with CHA2DS2VASc in predicting 5-year TAE (0.700 and 0.707 vs. 0.649, P=0.013 and 0.030, respectively) and TAE-D (0.712 and 0.713 vs. 0.622, P<0.001 and <0.001, respectively). The AUC did not improve statistically from (C)ACE to (C)ACEN (0.700 vs. 0.707, P=0.600 for TAE; 0.712 vs. 0.713, P=0.917 for TAE-D). In HF without AF, TAE during follow-up was associated with adverse prognoses. The independent risk factors of TAE or TAE-D improved CHA2DS2-VASc predictive ability, especially in patients with HFrEF. Our findings provide new evidence for TAE risk stratification in HF without AF, potentially guiding prophylactic anticoagulation.
Read full abstract