New Methods for Thromboembolic Risk Stratification in Non-valvular Atrial Fibrillation

Abstract

Thromboembolic risk in patients with atrial fibrillation (AF) is highly variable and is currently assessed through the use of clinical risk schemes, the CHADS2 and CHA2DS2-VASc scores, which perform only slightly better than chance. Transesophageal echocardiogram (TEE) changes are associated with a very high thromboembolic risk and may be used for better understanding the thrombogenic process in this arrhythmia. There are two situations where thromboembolic risk is very acutely increased in patients with AF: cardioversion and catheter ablation. Aims: Goal I: To assess the impact of AF on cerebrovascular mortality at a national level through the review of the currently available evidence. Goal II.A: To evaluate the role of echocardiographic parameters (left atrial – LA - area, volume and speckle-tracking derived strain and strain rate and left ventricle ejection fraction) in the prediction of LA thrombi and other markers of atrial stasis. Goal II.B: To assess the possible contribution of biomarkers (troponin I, mean platelet volume - MPV, red cell distribution width - RDW, mean globular volume – MCV - and C reactive protein) in the understanding of the pathways involved LA stasis. Goal II.C: To clarify the role of chronic kidney disease (CKD), assessed through the estimated glomerular filtration rate (eGFR), as a predictor of thrombogenesis and thromboembolic events. To clarify the benefit and safety of warfarin and the novel oral anticoagulants in patients with CKD and AF. Goal III: To systematically revise all the published evidence concerning dabigatran in the setting of catheter ablation of AF, evaluating its efficacy and safety. Goal IV: To derive the first scheme combining clinical and echocardiographic variables, alongside with biomarkers able to discriminate with high accuracy the presence of LA thrombi. Results: I. AF accounts for a relevant number of ischemic strokes in Portugal and its contribution should not be negleted since it seems not to be decreasing. II.A. Adding LA area and left ventricular ejection fraction to the CHADS2 and CHA2DS2-VASc scores significantly improved their discriminative capability for TEE findings: c-statistics of 0.62 and 0.63, respectively, increased to 0.73 and 0.74 (both P<0.01). AF episode duration, peak negative strain rate and time-to-peak positive strain were independent predictors of LA appendage thrombi or sludge. The c-statistic for the estimated probabilities using this model was 0.89 (95%CI 0.81-0.96; P <0.001). II.B. A direct relation between rising concentrations of troponin I and a higher prevalence of TEE stasis-related changes was found. Adding troponin I to CHADS2 and CHA2DS2-VASc scores improved their predictive ability. MCV and RDW were independent predictors of LA appendage thrombus and dense spontaneous echocardiographic contrast. MPV was an independent predictor of LA appendage thrombus and was incorporated in the predictive models of dense spontaneous echocardiographic contrast, low flow velocities and LA stasis, adding predictive value to clinical, echocardiographic and laboratory variables. II.C. A higher prevalence of dense spontaneous echocardiographic contrast in patients with lower eGFR rate was observed. On multivariate analysis, a predictive value was found for eGFR, that was additive to that of clinical parameters from CHADS2 and CHA2DS2-VASc. The presence of CKD leads to increased risk of thromboembolism (HR = 1.46; 95%CI 1.20-1.76, P=0.0001), particularly in case of terminal CKD (HR = 1.83; 95%CI 1.56-2.14, P<0.00001). Warfarin decreased the incidence of thromboembolic events in patients with non-endstage CKD (HR = 0.39; 95%CI 0.18-0.86, P<0.00001), as well as among patients on dialysis (HR = 0.44; 95%CI 0.26-0.74, P=0.002). In addition, novel oral anticoagulants showed higher efficacy, compared to warfarin (HR = 0.77; 95%CI 0.64-0.93, P=0.006) and aspirin (HR = 0.32; 95%CI 0.19-0.55, P<0.0001) among non-endstage CKD. III. In the setting of cateter ablation of AF, no significant differences were found between patients treated with dabigatran and warfarin regarding thromboembolic events and major bleeding. No difference was found between the 110mg bid and 150mg bid dabigatran dosages concerning these two endpoints. IV. CHADS2 and CHA2DS2-VASc had a modest performance in predicting TEE endpoints displaying a 0.62 c-statistic in average. Using CATES score (C-reactive protein, LA volume, Troponin I, Episode duration and Stroke or embolism) displayed a higher c-statistic: 0.82 for LA appendage thrombus. No patients with thrombus were observed in patients with CATES scores ranging from “0” to “2”, which corresponded to 49.4% of the sample. Conclusion: We have identified new markers and pathways that seem to be involved in this thrombogenic process: clinical variables (CKD as assessed through the eGFR), biomarkers (MPV, MCV, RDW, troponin I and C-reactive protein) and echocardiographic parameters (LA size - area and volume - and deformation). These variables seem to independently increment the discriminative of the clinical variables currently used in risk scores (CHADS2 and CHA2DS2-VASc). In the context of moderate CKD, the novel oral anticoagulants performed favourably, when compared with warfarin. However, in end-stage kidney disease, no data exists concerning the use of these agents. Also, our results seem to ease some of the concerns regarding the use of warfarin in these patients. Despite not clarifying the best regimen for dabigatran in the setting of catheter ablation of AF, our meta-analysis confirms the efficacy and safety of this drug. Using the new CATES scheme, a great amount of patients may eventually be spared TEE before cardioversion.