Which Anticoagulants Should Be Used for Stroke Prevention in Non-Valvular Atrial Fibrillation and Severe Chronic Kidney Disease?

Philip A. Kalra; Alexandru Burlacu; Charles J. Ferro; Adrian Covic


Curr Opin Nephrol Hypertens. 2018;27(6):420-425. 

In This Article

Vitamin K Antagonists in Patients With Severe Chronic Kidney Disease and Atrial Fibrillation: Potential Importance of the Time in Therapeutic Range

There are no RCT data to confirm that stroke risk or mortality are reduced by any form of anticoagulation in patients with CKD stage 4 or 5 or those receiving dialysis. Tan et al.[10] undertook a systematic review that included 2709 references, which distilled down to a meta-analysis of 20 observational cohorts that included over 56 000 patients with CKD 5D and atrial fibrillation. Warfarin use was not associated with reduction in all-cause stroke (hazard ratio 0.92, 95% CI 0.74–1.16), but it was associated with an increase in all cause bleeding (hazard ratio 1.21, 95% CI 1.01–1.44). In the Taiwanese dialysis registry study of Shih et al.,[5] the 6772 patients with atrial fibrillation had an annual stroke rate of 3.35% but annual mortality rate seven times greater (24.7%). Warfarin use was also not associated with a reduction in stroke rate. Yoon et al. analysed Korean dialysis registry data which included almost 10,000 patients with atrial fibrillation of which 29.3% had warfarin use;[11] 5548 patients were included in a propensity matched analysis to account for baseline differences. After an average follow up of 15.9 months the overall (haemorrhagic and ischaemic) stroke rate was 6.8% in warfarin treated versus 2.3% in untreated patients. Cox proportional hazards analysis showed a 44% increase in haemorrhagic stroke but no difference in ischaemic stroke with warfarin use. No difference in gastrointestinal bleeding was observed in atrial fibrillation patients who were warfarin users or nonusers.

These data are at odds with the findings of the SWEDEHEART registry in which VKA treatment did reduce a composite of all cause death, myocardial infarction and ischaemic stroke with no increased risk of bleeding.[12] Over 50% of the 24,317 participants in this web-based data system had CKD (eGFR < 60 ml/min/1.73 m2) and hazard ratios for better outcomes with VKA were consistent across all strata of CKD including stage 5 CKD. One explanation for the positive outcomes associated with warfarin is provided by another study from Sweden, which analysed outcomes in 7738 patients commencing warfarin for atrial fibrillation (the SCREAM cohort).[13] This concentrated on the impact of the time in the therapeutic range (TTR), which was based upon a median of 21 international normalized ratio (INR) readings per patient. A TTR of 70% is considered reasonable anticoagulant control, but in Sweden the healthcare system usually facilitates TTR of more than 75%; in SCREAM TTR was 83% overall but 70% in patients with eGFR less than 30 ml/min/1.73 m2. Adverse events were twice as frequent in patients with TTR less than 60% compared to more than 75%. The impact of lower renal function on the TTR, and of INRs outside of the target range, has been emphasized by Bonde et al.[14] who aggregated several Danish registry studies to analyse the effects of warfarin use in 10441 patients with atrial fibrillation; 43% had CKD stage 3 and 390 patients (3.7%) stage 4 or worse CKD. Mean TTR was 66.7%, 61.2 and only 51.1% in patients with eGFR more than 60 ml/min/1.73 m2, stage 3 or those with eGFR less than 30 ml/min/1.73 m2, respectively. Using an INR of 2–3 as reference, INR less than 2 was associated with an increased hazard for ischaemic stroke or systemic embolization (hazard ratio 1.45, 95% CI 1.29–1.61) whereas INR of more than 3 was associated with increased bleeding (hazard ratio 1.36, CI 1.13–1.49).

In summary, although VKA have been used for decades as prophylaxis against embolic complications of atrial fibrillation in diverse groups of patients, in contrast to the situation with early stage CKD most studies have shown that their efficacy in severe CKD, particularly dialysis patients, is very limited. This lack of effect is compounded by a significant risk of bleeding. A low TTR may underpin the lack of effectiveness and bleeding associated with VKA, and TTR is more likely to be reduced in patients with severe CKD. Guidance should thus advise very careful consideration before commencing VKA in dialysis patients, but if they are to be used, efforts need to be concentrated on maintaining a high TTR.