Patients with end-stage renal disease on hemodialysis have a high prevalence of atrial fibrillation (AF), ranging from 3.8% to 27% across the different registries.1
AF is associated with an increased risk of systemic thromboembolic events, including ischemic stroke. In patients at a significantly high risk, as indicated by a CHA2DS2-VASc score ≥ 2, oral anticoagulation is recommended.
Patients on hemodialysis with associated AF have high thromboembolic and hemorrhagic event rates, reported in up to 5.61 and 8.89 cases per 100 persons-years,2 respectively, which renders the decision to initiate antithrombotic therapy difficult. In this setting, treatment with vitamin K antagonists has been associated with excess bleeding and worse outcomes, although the results of observational studies diverge markedly and no randomized trials have addressed this question. Furthermore, the clinical guidelines of the main international scientific societies make contradictory recommendations on this topic.3
In addition, experience with direct-acting oral anticoagulants in hemodialysis patients is scarce and the associated bleeding risk remains high. As a result, many hemodialysis patients with AF are left without treatment for the prevention of thromboembolic events.1
In this setting, left atrial appendage occlusion (LAAO) appears to be an attractive alternative, as it provides protection against thromboembolic events, without increasing bleeding risk. However, to the best of our knowledge, only 1 previous study has assessed the early efficacy and safety of LAAO in hemodialysis patients4 and no data on long-term efficacy have been published to date.
We performed a single-center retrospective analysis of patients with nonvalvular AF on hemodialysis that had undergone a LAAO procedure, to assess its long-term efficacy and safety in a real-world cohort of hemodialysis patients.
Between January 2013 and January 2018, 14 patients were identified. The mean age at the time of the procedure was 69.21±11.58 years and 10 (71.4%) patients were male. Baseline thromboembolic and bleeding risk were both high with mean CHA2DS2-VASc score of 4.5±1.45 and mean HAS-BLED of 5.0±0.96. LAAO was recommended because of previous significant bleeding in 11 patients (78.6%), labile international normalized ratio in 5 (35.7%), and hematological disorders in one patient (7.1%).
LAAO was performed with a Watchman device (Boston Scientific) in 7 patients, with an Amulet device (Abbott Vascular) in 6 patients, and with an Ultraseal (Cardia Inc) in 1 patient. Device deployment was successful in all cases and no device-related or periprocedural complications developed, with the exception of an allergic reaction to iodine contrast in 1 patient (Table 1). All patients were successfully discharged home within 2 days and there were no early deaths or complications at 30 days.
Patient Follow-up
| Early follow-up | Number of events | Event description | Median time to event (d) |
|---|---|---|---|
| Periprocedural complications (first 72 h) | 1 | Allergic reaction to iodine contrast | _ |
| In-hospital mortality | 0 | 0 | _ |
| 30-day mortality | 0 | 0 | _ |
| Long-term follow-up | Number of events | Event description | Median time to event (d) |
| Stroke/transient ischemic attack | 0 | 0 | _ |
| Bleeding | 4 | BARC 2 bleeding events: 3BARC 3a bleeding events: 1 | 129 |
BARC, Bleeding Academic Research Consortium.
Postprocedural antithrombotic management included a 45-day period of dual antiplatelet therapy in 12 patients (85.7%), and single antiplatelet therapy with low-dose aspirin thereafter.
During a median follow-up of 585 days, 4 patients had bleeding complications, with 3 BARC 2 minor bleeding events and 2 BARC 3a hemorrhage, requiring transfusion. Importantly, no thromboembolic events were recorded during follow-up.
The usefulness of LAAO in patients with chronic kidney disease has been previously described. LAAO has shown comparable procedural safety among patients with and without chronic kidney disease and has been proven to be effective in significantly reducing cerebrovascular and bleeding rates at all chronic kidney disease stages compared with the expected annual risk.5,6 However, only 1 previous study4 has reported the preliminary feasibility and safety results of LAAO in hemodialysis patients and its long-term efficacy has not been investigated.
The current study offers an initial insight into this clinical problem, by providing long-term follow-up data, which is essential to assess the efficacy of LAAO at preventing thromboembolic events in this very high-risk population.
The present analysis suggests that LAAO could be a safe and effective procedure in hemodialysis patients, in which it may be a reasonable alternative to oral anticoagulation. Nevertheless, careful attention to baseline comorbidities prior to indication of LAAO is of the utmost importance in this population, and further randomized trials are warranted.
CONFLICTS OF INTERESTI. Cruz-González is proctor for Abbot Vascular. No other conflicts of interest exist.