Use of Non-Vitamin K Antagonist Oral Anticoagulants in Patients with Heart Failure and Atrial Fibrillation: Does Concomitant Kidney Disease Change our Practice?

This article refers to ‘Severity of renal impairment in patients with heart failure and atrial fibrillation: implications for non-vitamin K antagonist oral anticoagulant dose adjustment’, by N.M. Hawkins et al., published in this issue on pages 1162–1171. Heart failure (HF) and chronic kidney disease (CKD) are closely related, with CKD being an established independent marker of poor prognosis in patients diagnosed with HF.1 In the Acute Decompensated Heart Failure Registry (ADHERE), for example, 30% of the hospitalized patients had renal insufficiency;1 of this cohort, 21% had serum creatinine levels >2.0 mg/dL, 9% had serum creatinine >3.0 mg/dL, with 5% on dialysis. Thus, the term ‘cardio-renal syndrome’ has been coined to describe this close relationship between the heart and the kidney, where patients with established LV dysfunction develop glomerular hypoperfusion and subsequent impairment in renal function.2 Similarly, AF and CKD commonly co-exist in the same patient, and the presence of AF increases the relative risk of stroke in subjects with CKD at least to the same extent as in AF populations with normal renal function.3, 4 Although CKD is a significant risk factor for stroke and major bleeding, CKD per se does not improve the predictive ability of the CHA2DS2-VASc score for high-risk patients, given that CKD is associated with its risk factor components.5 Furthermore, based on these considerations, stroke prevention with oral anticoagulation (OAC) remains essential in AF patients with CKD, with evidence of a clear positive net clinical benefit in such patients.6 However, patients with CKD still have a higher risk of both intracranial and extracranial haemorrhage.7 A fine balance is therefore needed to shift the overall risk-to-benefit ratio towards stroke prevention with minimal possible risk of harm from serious bleeding with OAC use, and, if a vitamin K antagonist (VKA) is used, the importance of good quality anticoagulation control [as reflected by the time in therapeutic range (TTR)] is crucial, given the strong inverse relationship between TTR and thrombo-embolism and bleeding. The TTR can be related to common clinical factors, and simple risk scores, such as SAMe-TT2R2, have been validated to predict patients effectively who will be able to maintain good anticoagulation control with a VKA.8, 9 Large-scale registry data analyses have shown that TTR in patients with HF is often suboptimal, and this has been shown in previous studies to predispose to heightened stroke and bleeding risk.10 Warfarin regrettably has numerous drug and food interactions, which accounts for repeated patient visits to anticoagulation clinics to check the international normalized ratio (INR), and this in turn can lead to hesitation on the part of patients to initiate anticoagulation in the first instance.11 An alternative to the VKAs would be the non-VKA oral anticoagulants (NOACs), but there are no clinical trials in patients with severe CKD to guide clinicians as to the best practice in such circumstances. All the randomized trials excluded patients with creatinine clearance <25–30 mL/min, and, thus, observational studies and meta-analyses are important in providing some insight into the optimal management of such patients. In this issue of the journal, Hawkins et al. retrospectively analysed data from the Candesartan in Heart Failure Assessment in Reduction in Mortality and Morbidity program (CHARM).12 In CHARM, patients with HF were randomized to candesartan or placebo, but in this post-hoc retrospective analysis, the authors aimed to identify patients with a diagnosis of AF and to highlight changes in renal function during follow-up, and, where appropriate, the need for dose adjustment of NOACs based on creatinine clearance calculations. Serum creatinine was checked at baseline, 6 weeks, 14 months, and 26 months, and patients with AF found at baseline visit or during the study were identified. Estimated glomerular filtration rate (eGFR) was then calculated using three methods: Cockcroft–Gault (CG), simplified Modification of Diet in Renal Disease (MDRD), and Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI). Patients with AF and HF tended to have more co-morbidities and a lower mean eGFR from the outset of the study. At baseline, a quarter of the study population with AF had at least moderate renal impairment (eGFR <50 mL/min). One-third showed a decline in renal function at completion of the study, and approaching a half of patients had such deterioration at least once during follow-up. Specifically, one-fifth of patients with AF developed at least moderate worsening of renal impairment across the four visits. However, very few patients developed renal impairment significant enough to justify cessation of NOAC treatment. Nevertheless, a sizeable fraction of participants receiving NOACs required a dose reduction, which was needed in 6% of AF subjects treated with apixaban and in 16% for those receiving rivaroxaban or edoxaban. Hawkins et al. explain that in the USA, dabigatran dose reduction is only indicated for severe renal impairment (eGFR 15–30 mL/min), potentially affecting a similar proportion of patients to the apixaban criteria (6.2% at baseline and 5.8% during follow-up). In contrast, rivaroxaban and edoxaban dose reduction is recommended for moderate or severe renal impairment (15–50 mL/min), affecting 27.3% of patients at baseline and a further 16.4% during follow-up. European guidelines are similar to those in the USA for the prescribing of rivaroxaban and apixaban, but more stringent for edoxaban and dabigatran. Applying such guidelines would have prohibited the use of dabigatran in 6.7% and 5.8% of patients at baseline and follow-up, respectively, with a dose reduction in 14.4% and 7.8%. This study highlights the importance of regular renal function monitoring of patients prescribed NOACs. Indeed, patients with HF and AF show a greater decline in renal function compared with patients without AF. This underscores the importance of regular renal function monitoring of patients taking NOACs. However, such monitoring should not deter the use of NOACs in patients in need of anticoagulation. Rather, this should further encourage the use of such agents that are non-inferior (if not superior) for stroke protection whilst providing a reduced major bleeding risk when directly compared with warfarin; indeed, large real-world studies back up observations from the randomized trials.13, 14 For example, a recent large German registry highlighted the superior efficacy and relative safety of rivaroxaban and dabigatran in ‘real-world’ data over a 2-year period.15, 16 Furthermore. rates of discontinuation of therapy were lower with rivaroxaban when compared with warfarin in this cohort. In a retrospective analysis in Asian patients, a direct comparison between NOACs and warfarin in 1319 patients with renal dysfunction has recently been reported, which found that patients with renal dysfunction faired significantly worse on warfarin.17 The composite clinical outcomes of death, hospitalization, and new-onset stroke were significantly lower in patients taking NOACs vs. warfarin [hazard ratio (HR) 0.30, 95% confidence interval (CI) 0.11–0.77]; however, in patients without renal dysfunction there were no significant differences in the composite clinical outcome.17 In summary, Hawkins et al. correctly identify the need for regular monitoring of patients receiving NOAC therapy. Patients with HF and AF are at heightened risk of developing renal dysfunction despite normal parameters at baseline. In fact, some patients with CKD may do worse on warfarin therapy,5 but previous studies have paid less attention to the TTR. NOAC use is contraindicated in severe CKD (creatinine clearance <30 mL/min for dabigatran, and <15 mL/min for others) but, based on the study of Hawkins et al., NOACs only infrequently require termination solely based on dramatic changes in creatinine clearance. Nonetheless, the provision of clear guidance for dose adjustment and renal function monitoring where anticoagulants (whether VKAs or NOACs) are used in AF patients with renal dysfunction should provide a clear pathway for optimizing stroke prevention in these high-risk CKD patients.4 Conflict of interest: G.Y.H.L. is a consultant for Bayer/Janssen, Astellas, Merck, Sanofi, BMS/Pfizer, Biotronik, Medtronic, Portola, Boehringer Ingelheim, Microlife, and Daiichi-Sankyo; and a seaker for Bayer, BMS/Pfizer, Medtronic, Boehringer Ingelheim, Microlife, Roche, and Daiichi-Sankyo. The other authors have no conflicts to declare.