P3633Incidence of left atrial thrombus prior to direct current cardioversion in patients anticoagulated for 3 weeks or more with warfarin or novel oral anticoagulation

Abstract

Background: The use of anticoagulation is routine prior to direct current cardioversion (DCCV) for atrial fibrillation or atrial flutter. Current guidelines suggest at least 3 weeks of anticoagulation with either warfarin or novel oral anticoagulants (NOAC) to prevent thrombus formation in the left atrial appendage (LAA) and prevent embolic stroke. Transoesophageal echocardiograms (TOE) may be performed to visually confirm and rule out LAA thrombus. Purpose: We retrospectively reviewed the change in practise of anticoagulation prescribed prior to DCCV and ensure its safety in preventing LAA thrombus formation and embolic phenomenon. Materials and methods: We audited the data of all patients undergoing DCCV for atrial fibrillation or flutter since 2012 in the cardiology department. Electronic notes were reviewed for baseline demographics, CHADSVASC score and anticoagulation regime prescribed. TOE images were reviewed by three consultant cardiologists to confirm the findings. Patients were followed up as normal practice and if anticoagulation failed patients were consulted about their compliancy. Results: 354 patients underwent DCCV of which 238 had a TOE. 20 patients were lost to follow up. The mean age was 63 years (range 18-94). The mean CHADsVASC score was 2.04 (SD1.49). Warfarin usage was 75% in 2012 declining to 7% in 2016. NOAC use increased from 6% to 89% in 2016, rivaroxaban being the overwhelming choice with 73%. Of the 238 who underwent TOE, 6 had filling defects suggestive of LAA thrombus. All had ≥3 weeks of formal anticoagulation. 2 were on warfarin and had inadequate INR control. 1 patient was taking apixaban once a day. 3 were on rivaroxaban of which one TOE was re-reported as “suspicious” on review. These patients were compliant and taking rivaroxaban with food. Of the 334 patients with follow up, 2 had MRI confirmed TIA's in the 4 weeks following DCCV. Both had a CHADSVASC of 2, had ≥3 weeks of rivaroxaban prior to and continued anticoagulation following successful DCCV which was proceeded by a TOE. Results and conclusion: Our retrospective dataset shows the changing practise of clinicians, favouring NOAC over warfarin prior to DCCV. Our single centre experience revealed that rivaroxaban, whilst being the favoured NOAC had two confirmed LAA thrombus and one suspicious finding with no obvious cause for the failure of treatment. There were also two TIA events confirmed by MRI with patients on rivaroxaban. Whilst retrospective, this is a real-world experience and warrants further prospective evaluation.