A familiar problem. A problem that sits inside an increasingly convoluted landscape of treatment options. Atrial fibrillation and advanced chronic kidney disease (CKD):

Do we anticoagulate?

Which agent to use?

How much?

We all face these questions frequently in both inpatient and outpatient consultation. Prior to the introduction of non-vitamin-K-dependent oral anticoagulants (NOACs, also “novel oral anticoagulants”), we could comfortably focus on a single available body of literature – the use of warfarin in CKD. While that data was well-trodden and generally favored the use of warfarin in non-valvular atrial fibrillation (NVAF), the risks unique to the CKD population remained poorly defined. Even with that bit of uncertainty, most providers were able to have a working clinical response to “Is warfarin OK for my patient?”

With the introduction of NOACs, though, patients and providers have sought to take advantage of their easier dosing regimens and the absence of monitoring requirements. That pursuit has renewed these questions and the sea of emerging clinical trials leaves us once again adrift in data. A recent article published in AJKD by Shroff et al offers a beacon in that sea; they distill down the considerations relevant to clinicians as we answer these questions in this new era. I, for one, am grateful for their efforts.

In their review, the authors assemble the available trials in a practical fashion so that providers can address patients with both early and late CKD. As they note, the large available randomized controlled trials have included representation of lower stage CKD. In CKD stages 1 to 3, the four available NOACs have all demonstrated noninferiority to warfarin, with dabigatran 150 mg twice daily meeting the superiority endpoints as defined by the RE-LY trial.

Shroff et al emphasize that in practice, consideration must be paid to patient-specific factors as we weigh the risk of thromboembolic event against the risk of bleeding. Table 1 nicely summarizes the relevant CKD patient cohort, as well as the principal outcome (reduction in stroke or systemic emboli) and principal safety measures (major bleeding events).

Summary of Pivotal Randomized Trials of NOACs Compared to Warfarin, With Emphasis on Participants With Moderate CKD. Table 1 from Shroff et al, AJKD © National Kidney Foundation.

To illustrate their conclusions, for a patient with a recent thromboembolic CVA and no known bleeding history, dabigatran 150 mg twice daily may make the most sense given its observed superiority. That contrasts against a similar patient with a history of recurrent colonic bleeding who may be better served by an agent with a better bleeding profile (apixaban or edoxaban, for example, given their lower bleeding rates relative to the warfarin control arms).

Shroff et al further emphasize that things remain nebulous when trying to weigh one NOAC against another, as the trials had varying definitions for “moderate CKD,” at inclusion as well as variation in the mean CHADS_2­ scores of patients ultimately included (among other things). As we all know, it isn’t always apples to apples.

Unfortunately, in more advanced CKD – stages 4 through 5D – there is a lot less direction. The authors have tried to leverage the available observational data to help choose the best option, but, in the end, no good answer yet exists for patients with CLcr <25 mL/min (with ARISTOTLE including 25-29 mL/min in their definition of “moderate CKD”, though we could quibble over whether that constitutes a “good” answer). The confusion around advanced CKD is perhaps best embodied by the differences in regulatory guidance across governmental agencies, with the FDA, the European Medicines Agency, and Health Canada all having different positions on dosing of the various NOACs in advanced CKD:

Dosing Recommendations for NOACs by International Regulatory Bodies in Nonvalvular Atrial Fibrillation for Patients with eCL_cr ≤ 30 mL/min. Table 2 from Shroff et al, AJKD © National Kidney Foundation.

Shroff et al note that a large meta-analysis of patients on dialysis found no association with stroke but may associate with higher all-cause bleeding. This calls into question the role of warfarin for NVAF in our dialysis patients. The authors stop short of suggesting that the NOACs may suffer the same unfavorable risk-benefit profile, emphasizing only that we need more information.

My favorite part of this paper, though, isn’t the data review as much as it is the discussion of relevant factors to consider at the time of drug selection. I don’t want my brief summary to rob anyone of the insights afforded from reading it, so I will say that if you’re faced with one of these patients, you would do well to start with that section first. The authors take great care with the topic of assessing renal function, drug interactions worth considering, age, body size, and relevant comorbidities, as well as patient-oriented factors such as adherence and cost.

A few things from this discussion bear repeating:

1. As all of the NOACs are variably dependent on renal clearance, an accurate assessment of renal function is critical. That said, dosing recommendations are centered – largely because of previous FDA guidance on the topic – on the Cockcroft-Gault equation. Since most of us rely on an eGFR (calculated by MDRD or CKD-EPI), this can lead to discrepancy which they illustrate.
2. Age consideration is valuable and may be underappreciated if focus is applied solely to renal function.
3. Drug interactions vary greatly between agents. This comes as no surprise, but again could go unnoticed if renal function serves as the sole determinant for selection. This section is a wonderful review, but it’s also a potent reminder that no two risk-benefit assessments are the same.

As a beacon in the dark, this review does an excellent job of letting us know where we are. We wait patiently for more data to emerge – pharmacokinetic, observational, randomized controlled trials – we need them all. Until that emerges and gives a clearer sense of what lies ahead, this paper also serves as a written reminder that, in so many ways, we remain adrift.

– Post prepared by Ed Gould, AJKDBlog Contributor. Follow him @paininthekidney.

To view Shroff et al (subscription required), please visit AJKD.org. 

Title: Non–Vitamin K–Dependent Oral Anticoagulants for Nonvalvular Atrial Fibrillation in Patients With CKD: Pragmatic Considerations for the Clinician
Authors: Gautam R. Shroff, Rachel Stoecker, and Allyson Hart
DOI: 10.1053/j.ajkd.2018.02.360