FC 103: Matched Control Study on Applied Long-Term Anti-Thrombotic Therapies and Clinical Outcome of Atrial Fibrillation in HD-Patients Based on German Network Data Free

Hemodialysis (HD) patients (pts) with atrial fibrillation (AF) are at high risk for cardiovascular events, severe bleeding and rapid vascular/valvular calcification. Thus, antivitamin-K based oral anticoagulation (VK-OAC) for HD patients is debated, since prospective trials are missing and US register studies are often limited to 3–6 months OAC use. We studied AF risk-factors, long-term antithrombotic therapies and clinical outcomes in a large German HD cohort.

We analysed pseudonymized benchmarking data in a German out-patient dialysis center network (Verband Deutsche Nierenzentren, DN) based on quarterly electronically transmitted data. Diagnoses coded by ‘International Classification of Diseases (ICD)’ and ‘Anatomical Therapeutical Chemical (ATC)’ drug codes of adult HD patients 2013–2018.

In 2013, 2753 (18%) of 15 682 HD-pts had AF as coded diagnosis. Baseline CHA2DS2-VASc (4.0/1.5 mean/SD) and HAS-BLED (3.2/0.9) risk-scores indicated high risk for embolism and bleeding. Charlson Comorbidity Index (CCI) was high (6.4/2.8) and median observation was 2.1 years (range 0.01–6 years).

Beside HD-related heparin, four main OAC approaches were applied: no active therapy, VK-OAC+/-aspirin/clopidogrel (VK-OAC+/-Asp/Clop), heparin-based therapy (heparin+/-Asp/Clop) or 1–2 antithrombocyte drugs (Asp/Clop). A total of 959 pts (35%) changed therapy, but 1794 pts (65%) had no change in OAC regime during the 6 year study period and were used for final therapy-related outcome analysis (Table 1).

Both AF risk-scores overestimated de-novo events. Total 6-year event rate was low (7.5%; <1.3% per year) and comparable for all anticoagulant therapies, especially for cerebral ischemic events (2.8%; range 1.2–3.7; NS). All three antithrombotic therapies had similar overall events as without active therapy (7.6 versus 7.3%), including cerebral adverse events (ischemic: 3.0 versus 2.6%; bleeding: 0.7 versus 0.6%; NS).

Survival (Kaplan–Meier) was analysed for matched controls by propensity-score based on mortality risk factors in multivariate Cox regression: age (HR 1.05), sex (female HR 0.78), CCI (HR 1.07) and albumin (HR 0.93). Six-year mortality rates were high (63%) and matched analysis showed significant lower survival (P < 0.001) without anticoagulant (median 1.8 years) or on heparin-based therapy (1.7 years) than on Asp/Clop (2.9 years) or VK-OAC based therapy (2.8 years). This relation was similar for subgroups on age, CCI or changed therapy.

Our large study showed that baseline CHA2DS2-VASc and HAS-BLED scores had no predictive value for clinical events in HD pts with AF.

Cerebral ischemic event rates over 6 years were low (<0.7%/year) and similar for all three antithrombotic therapies and even no active therapy, suggesting major benefit of regular dialysis-related heparin supply.

Since median survival on Asp/Clop is similar to VK-OAC therapy and even 1 year better than on no active or heparin-based therapy, we conclude that antithrombotic therapy in HD patients with AF can effectively be done with Asp/Clop and VK-OAC should be avoided. For future prospective trials, we recommend to apply Asp/Clop as first-line control therapy evaluating new direct-acting oral anticoagulants (DOAK) and/or interventional approaches (i.e. left atrial appendage closure) in HD patients with AF.