Apixaban Use in Patients with Kidney Impairment: A Review of Pharmacokinetic, Interventional, and Observational Study Data

Abstract

Chronic kidney disease (CKD) remains a significant global health issue and is a leading cause of mortality worldwide. Patients with CKD have an increased risk of developing atrial fibrillation (AF) and venous thromboembolism (VTE). While direct oral anticoagulants (DOACs) have become a standard of care for anticoagulation (AC) in patients with AF and VTE, the appropriate use of these agents in comorbid kidney impairment warrants detailed discussion. This scientific narrative review summarizes the effectiveness and safety of apixaban use in patients with renal dysfunction by assessing the current published pharmacokinetic, interventional, observational, and guideline data. Apixaban is a highly selective, orally active, direct inhibitor of factor Xa, with well-established pharmacokinetics and consistent clinical outcomes across a broad range of patient populations, including those with kidney impairment. Overall, the scientific literature has shown that apixaban has a favorable clinical efficacy and safety profile compared with vitamin K antagonists for patients with AF or VTE and comorbid kidney impairment. These data support the approved label dosing strategy of apixaban in reducing the risk of stroke/systemic embolism in patients with nonvalvular AF and in treating VTE across all ranges of kidney function. Both clinician experience and knowledge of patient-specific factors may be required in the management of comorbid patients with advanced CKD or those requiring dialysis, as data on these patients are limited.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40256-024-00664-2.

Key Points

There is an increased risk of developing atrial fibrillation (AF) and venous thromboembolism (VTE) in patients with chronic kidney disease (CKD).

Direct oral anticoagulants, such as apixaban, are increasingly being used in clinical practice for patients with comorbid CKD, including those with advanced kidney disease.

Published study data support the dosing strategy of apixaban in reducing the risk of stroke/systemic embolism in patients with nonvalvular AF, for the treatment of deep vein thrombosis (DVT) and pulmonary embolism (PE), for the reduction in the risk of recurrent DVT and PE following initial therapy, and in the prophylaxis of DVT/PE after hip or knee replacement surgery across all ranges of kidney function. However, data on patients with end-stage kidney disease, including those requiring dialysis, are limited.

Introduction

Chronic kidney disease (CKD) is a growing global health issue affecting > 13% of the population [1]. CKD is a progressive condition characterized by persistent structural and functional abnormalities in the kidney resulting in impaired renal function [2]. CKD is associated with significant morbidity and mortality, and its global prevalence is expected to increase [2]. Multiple risk factors can contribute to the development of CKD, with diabetes mellitus and poorly controlled hypertension recognized as leading causes [1]. CKD is defined as an estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73 m², albuminuria ≥ 30 mg/24 h, or both, for > 3 months, and is further classified by cause, GFR category, and severity of albuminuria [3, 4]. The stages of CKD based on GFR are summarized in Table 1 [3, 5].

Table 1.

Stages of chronic kidney disease based on GFR categories [3, 5]

GFR category	GFR (mL/min/1.73 m2)	Descriptive terms	Stage of CKD
G1	≥90	Normal or high	Stage 1: GFR normal kidney function but may be some kidney damage or disease
G2	60–89	Mildly decreaseda	Stage 2: Mildly reduced kidney function and may be some kidney damage or disease
G3a	45–59	Mildly to moderately decreased	Stage 3: Moderately reduced kidney function with or without a known kidney disease
G3b	30–44	Moderately to severely decreased
G4	15–29	Severely decreased	Stage 4: Severely reduced kidney function with or without known kidney disease
G5	<15	Kidney failure	Stage 5: Very severely reduced kidney function. Termed end-stage kidney failure or established renal failure

Note estimated GFR is normalized to a body surface area of 1.73 m² and creatinine clearance and eGFR are not interchangeable.

GFR glomerular filtration rate, CKD chronic kidney disease

^aRelative to young adult level. In the absence of evidence of kidney damage, GFR categories G1 and G2 do not fulfill the criteria for CKD

Patients with CKD are at an increased risk of developing a range of cardiovascular comorbidities, such as venous thromboembolism (VTE), which includes deep vein thrombosis (DVT) and pulmonary embolism (PE), both of which warrant initiation of anticoagulation (AC) [6]. In patients with stage 3 or 4 CKD, the adjusted relative risk of developing DVT was reported as 1.71 (95% confidence interval [CI] 1.18–2.49) when compared with patients with normal kidney function [7]. For patients with end-stage kidney disease (ESKD) who require dialysis, the incidence rate of PE is reported to be more than twofold higher when compared with patients with normal kidney function [8, 9]. Additionally, patients with CKD have an increased risk of developing atrial fibrillation (AF) [6]. The risk of AF increases with the progression of CKD; approximately 10–25% of patients with CKD who require dialysis also demonstrate presence of AF [10–12]. In addition to increasing the risk of thromboembolic events secondary to AF or VTE, impaired kidney function is an independent risk factor for bleeding in patients with CKD, with bleeding risk increasing as CKD progresses [10]. Several risk score instruments are used to assess bleeding risk in clinical practice, such as HAS-BLED (Hypertension, Age, Stroke, Bleeding tendency/predisposition, Labile international normalized ratios [INRs], Elderly age/frailty, Drugs such as concomitant aspirin/nonsteroidal anti-inflammatory drugs or alcohol excess), ORBIT (Outcomes Registry for Better Informed Treatment of Atrial Fibrillation), and ATRIA (Anticoagulation and Risk Factors in Atrial Fibrillation), and include kidney function as a parameter required for score calculations [13–16]. Increased bleeding risk in CKD further complicates the use of anticoagulants in these patients [17, 18]. Thus, anticoagulant use in patients with renal impairment deserves special attention [17].

Vitamin K antagonists (VKAs) and/or heparins have historically been used to manage treatment of VTE, as well as thromboprophylaxis in AF or the post-knee/hip replacement surgery setting, in those with comorbid kidney impairment. Trends in prescribing patterns demonstrate that direct oral anticoagulants (DOACs) are now increasingly being used in clinical practice for patients with CKD, including those with advanced kidney disease [18]. Changes in renal function can also complicate the choice of AC therapy; each DOAC exhibits differences in pharmacokinetic (PK) and pharmacodynamic (PD) properties compared with VKAs and heparins, including variations in the proportion of renal clearance (CL_R) to total body clearance. A decrease in CL_R of DOACs increases exposure in patients who may consequently increase the risk of a bleeding event, and thus DOAC dose adjustments may be necessary to optimize the benefit–risk balance in patients with advanced kidney disease.

There are different methods to assess kidney function, and these methods are not superimposable. For readers who are interested in the topic of equations used in renal function assessment, please see the excellent review authored by Parker and Thachil [19]. In our review, the focus is to summarize evidence for apixaban in renal-impaired populations, as reported in existing literature.

Apixaban is a highly selective, orally active, direct inhibitor of factor Xa (FXa), approved for use in adults for the reduction in risk of stroke and systemic embolism (SE) in patients with nonvalvular AF (NVAF), treatment of DVT/PE and reduction in the risk of recurrence following initial therapy, and prophylaxis of DVT/PE after hip or knee replacement surgery [20, 21]. The efficacy and safety of apixaban have been well studied, with apixaban generally demonstrating superiority and/or non-inferiority when compared with either placebo or conventional therapy for the prevention of thromboembolic events in the ARISTOTLE (NCT00412984), AMPLIFY (NCT00643201), and AMPLIFY-EXT (NCT00633893) trials [22–28]. The development program identified practical patient-specific parameters with which to adjust apixaban dose for patients with NVAF at risk of embolic events. Appropriate dose adjustment from 5 mg twice daily to 2.5 mg twice daily is recommended if two or more of the following criteria are met: age ≥ 80 years, weight ≤ 60 kg, or serum creatinine ≥ 1.5 mg/dL (the dosing criteria for apixaban per approved labeling are summarized in Table 2 [20, 21]). Several real-world database analyses have noted a higher proportion of apixaban use versus other DOACs in patients with kidney impairment. A database analysis of 102,504 patients with AF and advanced CKD in the US reported apixaban as having the highest proportion of use among patients with renal impairment (10.4% vs. 9.5% with rivaroxaban, 3.5% with dabigatran, and 0.1% with edoxaban) [29]. Similarly, using US claims data between 2011 and 2017, Reyes et al. noted that in patients with AF and CKD stages 3–5, 29.6% received apixaban, 17.2% rivaroxaban, 6.9% dabigatran, and 46.3% warfarin [18].

Table 2.

Apixaban-approved dosing criteria

	United States [20]	Europe [21]
Approval	December 2012	May 2011
Indication	Reduction in risk of stroke/systemic embolism in patients at high risk for events relating to NVAF, treatment of DVT/PE, and reduction in risk of recurrence following initial therapy, and prophylaxis of DVT that may lead to PE after hip or knee replacement surgery	Reduction in risk of stroke/systemic embolism in patients at high risk for events relating to NVAF, treatment of DVT/PE, and reduction in risk of recurrence following initial therapy, and prophylaxis of DVT that may lead to PE after hip or knee replacement surgery
Recommended dose in patients with:
No renal impairment (> 80 mL/min)	5 mg orally twice daily	5 mg orally twice daily
Mild or moderate renal impairment (> 30–80 mL/min)	No dose adjustment for the prevention of VTE in elective hip or knee replacement surgery No dose adjustment for the treatment of DVT, treatment of PE, or prevention of recurrent DVT and PE Dose reduction to 2.5 mg orally twice daily for the prevention of stroke and systemic embolism in patients with NVAF and serum creatinine ≥ 1.5 mg/dL (133 μmol/L) associated with age ≥ 80 years or body weight ≤ 60 kg	No dose adjustment for the prevention of VTE in elective hip or knee replacement surgery No dose adjustment for the treatment of DVT, treatment of PE, or prevention of recurrent DVT and PE Dose reduction to 2.5 mg orally twice daily for the prevention of stroke and systemic embolism in patients with NVAF and serum creatinine ≥ 1.5 mg/dL (133 μmol/L) associated with age ≥ 80 years or body weight ≤ 60 kg
Severe renal impairment (CrCl 15–29 mL/min)	Dose reduction to 2.5 mg orally twice daily for the prevention of stroke and systemic embolism in patients with NVAF with serum creatinine ≥ 1.5 mg/dL (133 μmol/L) associated with age ≥  80 years or body weight ≤ 60 kg	Dose reduction to 2.5 mg orally twice daily for the prevention of stroke and systemic embolism in patients with NVAF Use with caution for the prevention of VTE in elective hip or knee replacement surgery, the treatment of DVT, treatment of PE, and prevention of recurrent DVT and PE
CrCl < 25 mL/min (including those with ESRD maintained on hemodialysis)	Dose reduction to 2.5 mg orally twice daily for the prevention of stroke and systemic embolism in patients with NVAF with serum creatinine ≥ 1.5 mg/dL (133 μmol/L) associated with age ≥ 80 years or body weight ≤ 60 kg No dose adjustment for prophylaxis of DVT following hip or knee replacement surgery, and treatment of DVT and PE and reduction in the risk of recurrence of DVT and PE	Not recommended for those undergoing hemodialysis
CrCl < 15 mL/mina	Dose reduction to 2.5 mg orally twice daily for the prevention of stroke and systemic embolism in patients with NVAF with serum creatinine ≥ 1.5 mg/dL (133 μmol/L) associated with age ≥ 80 years or body weight ≤ 60 kg No dose adjustment for prophylaxis of DVT following hip or knee replacement surgery, and treatment of DVT and PE and reduction in the risk of recurrence of DVT and PE	Not recommended

anti-FXa anti-factor Xa, CrCl creatinine clearance, DVT deep vein thrombosis, ESRD end-stage renal disease, NVAF nonvalvular atrial fibrillation, PE pulmonary embolism, VTE venous thromboembolism

^aClinical efficacy and safety studies with ELIQUIS did not enroll patients with ESRD on dialysis or patients with a CrCl < 15 mL/min; therefore, dosing recommendations are based on pharmacokinetic and pharmacodynamic (anti-FXa activity) data in subjects with ESRD maintained on dialysis

Objective

The objective of this comprehensive, scientific narrative review is to summarize the effectiveness and safety of apixaban use in patients with kidney impairment by assessing the current published PK, PD, interventional, observational, and guideline data related to apixaban.

Methods

A literature search was performed using MEDLINE for all articles published to 1 January 2023; non-English studies, animal studies, commentaries, non-systematic literature reviews, editorials, and letters were excluded. The search terms (and related synonyms) used to find the relevant articles included ‘apixaban’, ‘renal/kidney impairment’, ‘renal elimination/clearance’, and ‘guidelines’. The full literature search strategy and terms used for article selection for this review are summarized in Online Resource Table S1. Articles were screened for eligibility by title and abstract by two authors (SRM and KS) to ensure relevance.

Literature Search Results

Overall, 554 citations were identified (Online Resource Fig. S1). Articles were excluded if they did not contain at least one outcome variable of interest (e.g., those focused on drug–drug interactions; n = 85), if they focused on prescribing patterns (n = 64) or costs (n = 8), or if they did not assess outcomes based upon clear estimates or categories of kidney function (n = 317). The article selection methodology is further summarized in Online Resource Fig. S1. A total of 47 articles were selected as relevant for this review. Additional key references from personal libraries and other sources were identified and appraised by the authors (n = 62).

Pharmacokinetics and Pharmacodynamics of Apixaban

The PK/PD properties of apixaban have been well characterized by preclinical investigations and subsequently validated in clinical studies. In summary, apixaban has an absolute oral bioavailability of approximately 50%, undergoes rapid absorption with a maximum plasma concentration occurring 2–4 h after oral administration, and has a half-life of approximately 12 h [30–39]. Specific patient factors can influence the exposure to, metabolism, and elimination of apixaban, including age, weight, kidney function, liver function, and concomitant presence of interacting medications. This narrative review focuses on the PK/PD parameters of apixaban in the context of kidney function and renal impairment, including a review of the evidence that determined the extent of apixaban’s dependence on kidney function for total clearance.

Early in the apixaban development program, Raghavan et al. administered [¹⁴C] radiolabeled apixaban as a single 20 mg dose to two groups of healthy male volunteers [31]. Upon analysis of the urine samples, 24.5% of the [¹⁴C] apixaban dose was recovered in the first group and 28.8% in the second group [31]. The results of the analysis were consistent with those from a subsequent single ascending-dose study assessing PK, bioavailability, CL_R, and PD after intravenous and oral administration of apixaban. In this study, CL_R ranged from 17% to 30%, but analyses of pooled intravenous samples demonstrated that CL_R averaged 27% of total clearance [40].

To understand the impact of drug–drug interactions on apixaban PK, Vakkalagadda et al. assessed CL_R in the context of concomitant rifampin, a known hepatic enzyme inducer [39]. The authors reported that CL_R was approximately 34% of total clearance following intravenous apixaban administration. They also noted a 25% decrease in apixaban bioavailability with concomitant rifampin [39]. The totality of these data led to the package insert characterizing the CL_R of apixaban as ‘approximately 27%’.

Multiple studies assessing apixaban doses demonstrated increases in exposure in a dose-proportional manner following examination of both direct drug assays and anti-factor Xa (anti-Xa) levels [30–33, 35, 39–41]. The PK characteristics of apixaban are further summarized in Table 3 [30, 42].

Table 3.

Pharmacokinetics of apixaban [30, 42]

Pharmacokinetic characteristic	Value
Cmax, ng/mLa	128.5
AUC, ng/mL/ha	1051.9
Mean half-life, ha	11.7
Time to Cmax, ha	2.0–4.0
Renal clearance	~0.9 L/h (~27% of total clearance)
Removal by hemodialysis	14% reduction in exposure
Protein binding, %	87
Volume of distribution, L	21

AUC area under the curve, C_max maximum concentration

^a5 mg twice daily

The extent to which kidney impairment affects plasma concentrations and clinical outcomes of patients receiving apixaban is of clinical significance. Multiple publications have sought to characterize the impact of varying degrees of renal impairment on the PK/PD of apixaban. The earliest study, by Chang et al. [43], compared patients with normal renal function (> 80 mL/min), and mild (> 50 to ≤ 80 mL/min), moderate (≥ 30 to ≤ 50 mL/min), and severe (< 30 mL/min) impairment of renal function after a single 10 mg oral apixaban dose. The study also compared different methods of estimating renal function through nomograms and measuring kidney function to address the issue of potential discordance between indirect and direct estimations of kidney function. The four methods used were the Cockcroft–Gault equation (for initial classification of renal function), the Modification of Diet in Renal Disease (MDRD) equation, 24-h urine collection, and iohexol clearance [43]. Chang et al. reported that renal impairment did not impact maximum concentration (C_max) of apixaban, but exposure or area under the curve (AUC) increased as renal function decreased. Compared with patients in the normal renal function group, apixaban AUC was measured as 16%, 29%, and 38% higher in the mild, moderate, and severe renal impairment groups, respectively. The data also revealed that all four methods of assessing renal function showed consistent relationships following a linear slope pattern. The authors conducted a regression analysis with these data and estimated that a patient with a creatinine clearance (CrCl) of 15 mL/min would result in an AUC increase of approximately 44% compared with a patient with normal renal function. Additionally, a direct linear relationship between apixaban plasma levels and anti-Xa activity was reported; the relationship between plasma levels and anti-Xa activity was not affected by renal function. The mean half-life of apixaban was nominally affected by renal function and ranged from 15.1 to 17.6 h. These data are summarized in Table 4. Overall, this PK/PD study concluded that the dose of apixaban should not be adjusted based solely on renal function [43].

Table 4.

Summary statistics for apixaban pharmacokinetics by renal function [43]

Renal function group			Pharmacokinetic parameter
	Cmax, ng/mL Geo mean (% CV)	Tmax, h Median (min–max)	AUC,a ng h/mL Geo mean (% CV)	T½, h Mean (SD)	CL/F, mL/min Geo mean (% CV)	CLR, mL/min Geo mean (% CV)
Normal (n = 8)	224 (25)	2.8 (2–4)	2528 (26)	15.1 (7.6)	65.9 (20)	6.83 (33)
Mild (n = 10)	229 (33)	4.0 (1–6)	3288 (37)	14.6 (7.3)	50.7 (34)	3.81 (53)
Moderate (n = 7)	288 (18)	4.0 (3–4)	4479 (23)	17.6 (6.0)	37.2 (21)	1.94 (89)
Severe (n = 7)	210 (37)	4.0 (3–4)	3221 (49)	17.3 (7.4)	51.7 (69)	1.94 (45)

Table reproduced with permission from Chang M et al. J Clin Pharmacol 2016;56:637–645. Copyright © 2015 The American College of Clinical Pharmacology

AUC_T area under the plasma concentration–time curve from time 0 to the time of the last quantifiable concentration, AUC area under the plasma concentration–time curve from time 0 extrapolated to infinite time, C_max maximum observed plasma concentration, CL/F apparent total body clearance, CL_R renal clearance, CV coefficient of variation, Geo geometric, max maximum, min minimum, SD standard deviation, T_½ plasma terminal half-life, T_max observed time of C_max

^aBecause the percentage of the area that was extrapolated from the last time point to infinity was < 20% of AUC1, AUC_T is not summarized in the table

Tobe et al. also analyzed the relationship between apixaban anti-Xa activity in patients with moderate and severe renal impairment. Their analysis showed that apixaban anti-Xa peak and trough values did not differ between moderate and severe renal impairment, and among the DOACs targeting FXa, apixaban was the only medication to display this pattern [44]. However, in an observational analysis of a Chinese population, Sin et al. noted higher apixaban plasma peak and trough concentrations (C_trough) in patients with stage 3 CKD compared with those with stages 1 and 2. It is worth noting that a limitation of the analysis was that the study population was primarily composed of patients with stage 2 CKD [45].

Population PK studies using data from phase I, II, and III studies demonstrated similar patterns of apixaban exposure with varying renal function. For patients with NVAF and mild, moderate, or severe renal impairment, AUC at steady state approximated 9%, 28%, and 55% higher daily exposure, respectively, compared with patients with normal renal function. It is important to note that these population PK studies estimated the effects of multiple covariates on the PK of apixaban and consistently concluded that the variables assessed (e.g., renal function, age, weight, race, or sex) would not individually require a change in apixaban dose in isolation [36, 37, 46].

Additionally, studies in healthy individuals have shown that the PD effects of apixaban were consistent with its mechanism of action, which is direct reversible inhibition of FXa, and that plasma apixaban concentrations are approximately linear with anti-FXa activity [38, 47]. Renal impairment had no evident effect on the relationship between apixaban plasma concentration and anti-FXa activity [37, 38, 47].

Pharmacokinetics and Pharmacodynamics of Apixaban in Patients on Dialysis

The treatment of patients who require dialysis is of particular interest given the high rate of AF, elevated risk for VTE, and concurrent increased bleed risk. Apixaban is minimally removed during hemodialysis, with different studies reporting a removal range of 4–14% during a 4-h hemodialysis session [42, 48]. This is partially attributed to the high plasma protein binding of apixaban (87%) and to additional non-renal routes of elimination. One condition that represents an area of unmet need is nephrotic syndrome, which can alter protein binding of medications and PD response. There is a paucity of data for apixaban in this population.

Wang et al. analyzed PK/PD in healthy patients with normal renal function versus those with ESKD on a chronic dialysis regimen after a single 5 mg oral apixaban dose. Compared with those with normal renal function, the C_max and AUC were 10% lower and 36% higher, respectively, in patients requiring dialysis [42].

Further data in the dialysis population were published by Mavrakanas et al. comparing two apixaban dosage regimens over a 7-day period in a non-randomized design [48]. These authors compared apixaban C_max, minimum concentration (C_min), and AUC of patients receiving apixaban 2.5 mg twice daily with apixaban 5 mg twice daily in seven patients [48]. The reported exposure levels following apixaban 2.5 mg twice daily in patients on hemodialysis were comparable with the exposure in patients with normal renal function receiving 5 mg twice daily; however, the authors concluded that apixaban 5 mg twice daily should be avoided due to elevated levels after 1 week of administration [48]. An additional study of apixaban exposure by Van den Bosch et al. assessed the effect of both the apixaban dose and the timing of administration, defined as 30 min before or immediately after dialysis, in patients on maintenance hemodialysis [49]. The authors noted that exposure, as measured by C_max and AUC over 48 h (AUC₄₈), was dependent on both drug dose and timing of administration. Greater exposure was reported if apixaban was given immediately after a hemodialysis session as compared with before the hemodialysis session. Regarding differences observed between the 5 and 2.5 mg doses, a greater AUC₄₈ was measured for the 5 mg dose than the 2.5 mg dose, but only if apixaban was given before the dialysis session. The AUC₄₈ for the 5 mg dose was not significantly different from the 2.5 mg dose if given after dialysis. This study also showed that a 2.5 mg dose given after dialysis resulted in a similar AUC₄₈ to a 5 mg dose given before dialysis. The Van den Bosch publication provides valuable information, but it is important to note that there are no formal recommendations for the timing of apixaban administration with regard to a dialysis session [49].

The PK/PD studies summarized here provide important contributions to the literature on characterizing the properties of apixaban. The ultimate purpose of detailed PK/PD studies is to help inform dose selection for clinical trials. Outcomes from randomized controlled trials (RCTs) are the gold standard for assessing the benefit–risk profile of the dose(s) that are included.

Overall, the PK/PD characteristics of apixaban in patients with kidney impairment have been assessed by multiple investigators in preclinical studies. Kidney impairment alone is not a sufficient independent variable to warrant dose adjustment based on PK/PD data. A review of clinical trial and real-world data for this patient population as presented here is necessary to inform clinical treatment decisions.

Apixaban Randomized Clinical Trial Data

The influence of kidney impairment on treatment outcomes in patients with NVAF or VTE receiving apixaban has been described in RCTs and their associated subgroup analyses. However, patients with CrCl < 25 mL/min were excluded from the pivotal phase III apixaban RCTs, and while clinical trial data for patients undergoing hemodialysis are available, they are limited due to relatively small sample sizes. Results from apixaban phase III trials are summarized in Table 5, including the distribution of patients with normal or impaired kidney function (if not excluded) and clinical outcomes [22, 23, 27, 50, 51].

Table 5.

Summary of randomized clinical trials of apixaban treatment for NVAF and VTE, and breakdown of number of patients by kidney impairment

Randomized clinical trials of apixaban	Treatment regimen				Renal function [n (%)]				Outcomes for primary analysis
		Patient population [n]	Age, years [median (IQR)]	Female [n (%)]	Normal (> 80 mL/min)	Mild impairment (> 50–80 mL/min)	Moderate impairment (> 30–50 mL/min)	Severe impairment (≤30 mL/min)	Stroke or systemic embolism, n (%/year) [HR (95% CI); p-value]	Major bleeding, n (%/year) [HR (95% CI); p-value]
ARISTOTLE [22]	Apixaban 5 mg twice daily or reduced dose	9120	70 (63–76)	3234 (36)	3761 (41.2)	3817 (41.9)	1365 (15.0)	137 (1.5)	212 (1.27) [0.79 (0.66–0.95); 0.01]	327 (2.13) [0.69 (0.60–0.80); < 0.001]
	Warfarin (INR 2–3)	9081	70 (63–76)	3182 (35)	3757 (41.4)	3770 (41.5)	1382 (15.2)	133 (1.5)	265 (1.60)	462 (3.09)
AVERROES [23]	Apixaban 5 mg twice daily or reduced dose	2808	70 (9)a	1148 (41)	–	–	–	–	51 (1.6) [0.45 (0.32–0.62); < 0.001]	44 (1.4) [1.13 (0.74–1.75); 0.57]
	Aspirin 81–324 mg daily	2791	70 (10)b	1174 (42)	–	–	–	–	113 (3.7)	39 (1.2)
AUGUSTUS [50]	Apixaban (5 mg twice daily or reduced dose) and aspirin or aspirin placebo	4641	70.7 (64.2–77.2)	1337 (29)	4152 (91.6); SCr < 1.5 mg/dL	380 (8.4); SCr ≥ 1.5 mg/dL			1.61 (37) [1.01 (0.18–1.85)]c	1.22 (28) [ARD 0.8 (0.07–1.53)]
	Warfarin (INR 2–3) and aspirin or aspirin placebo								2.63 (60)c	2.02 (46)
AMPLIFY [27]	Apixaban 5 mg twice daily	2691	57.2 (16)a	1122 (41.7)	1721 (64.0)	549 (20.4)	161 (6.0)	14 (0.5)	59 (2.3) [0.84 (0.60–1.18); < 0.001]d	15 (0.6) [RR 0.31 (0.17–0.55); < 0.001]
	Warfarin INR 2–3	2704	56.7 (16)b	1106 (40.9)	1757 (65.0)	544 (20.1)	148 (5.5)	15 (0.6)	71 (2.7)	49 (1.8)
CARAVAGGIO [51]	Apixaban 5 mg twice daily	573	67.7 (11.1)b	579 (50.7)	201 (35.1)	235 (41.0)	97 (16.9)	40 (7.0)	–	3.6 [0.84 (0.26–2.71); 0.8819]
	Dalteparin 150 IU/kg once daily	569			222 (39.0)	209 (36.7)	84 (14.8)	54 (9.5)	–	4.3 [0.92 (0.47–1.83)]

CI confidence interval, HR hazard ratio, INR international normalized ratio, IQR interquartile range, NVAF nonvalvular atrial fibrillation, RR risk reduction, SCr serum creatinine, SD standard deviation, VTE venous thromboembolism

^aMean (SD)

^bMedian (SD)

^cCardiovascular death, stent thrombosis/myocardial infarction/stroke

^dRecurrent VTE or VTE-related death

Patients with Nonvalvular Atrial Fibrillation (NVAF)

The dosing criteria for apixaban in patients with NVAF are distinctive and differ from the treatment or prevention of VTE. While kidney function is part of the dose adjustment criteria, it is not the sole determinant. The dose of apixaban for patients with NVAF is 5 mg orally twice daily, but is decreased to 2.5 mg if two of the following three criteria are met: ≥ 80 years of age, body weight ≤ 60 kg, or serum creatinine level ≥ 1.5 mg/dL. This dosage regimen was determined after a detailed analysis of PK data and the phase II clinical trials APROPOS and BOTICELLI [52, 53]. A therapeutic utility index (TUI) method, which assesses the balance of efficacy and safety as a function of steady state AUC, was employed to integrate data and estimate the optimal balance by Leil et al. [54]. Twice-daily dosing demonstrated a higher TUI than once-daily dosing, revealing that C_trough was the best predictor of VTE events from phase II data, and the 2.5 and 5 mg twice-daily dosing regimens showed the highest TUI of all doses evaluated. Furthermore, Leil et al. state, “age, CrCl, and body weight did not appear to have significant effects on the exposure–response relationship with regard to efficacy and safety” as independent variables [54]. Thus, at least two of the three criteria need to be met for decreasing the dose to 2.5 mg twice daily.

Two registrational clinical trials evaluated benefit–risk and established the dose of apixaban in patients with NVAF, including those with comorbid kidney impairment: ARISTOTLE (Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation) [22] and AVERROES (Apixaban Versus Acetylsalicylic Acid to Prevent Stroke in Atrial Fibrillation Patients Who Have Failed or Are Unsuitable for Vitamin K Antagonist Treatment) [23].

In the ARISTOTLE trial, apixaban was compared with warfarin (target INR 2.0–3.0) in patients with NVAF, and at least one other risk factor for stroke. Apixaban demonstrated superiority in prevention of stroke or SE, for reduction of major bleeding, and the secondary endpoint of reducing all-cause mortality when compared with warfarin [22, 55]. Among the 18,201 patients enrolled, 7518 (41%) had an eGFR of > 80 mL/min, 7587 (42%) between 50 and 80 mL/min, and 3017 (17%) ≤ 50 mL/min [22, 55]. The initial publication noted directionally consistent results irrespective of renal function [22, 55]. A subanalysis of ARISTOTLE by Hohnloser et al. examined outcomes by renal function grouping, including CrCl 50–80 mL/min and < 50 mL/min. Clinical trial outcomes were reported using multiple methods of estimating renal function, specifically Cockcroft–Gault, the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation, and cystatin C eGFR. Results for both efficacy and safety were consistent regardless of method used (Fig. 1) [55]. Additionally, Hijazi et al. analyzed renal function over time in the ARISTOTLE trial. A total of 2294 (13.6%) patients in ARISTOTLE showed worsening renal function over time, defined as a 20% decrease in renal function in the first 12 months after randomization. This analysis showed that the benefit of apixaban versus warfarin was maintained for all efficacy and safety endpoints regardless of whether renal function declined over time or remained stable [56]. Furthermore, a lower dose of apixaban 2.5 mg twice daily was received by 4.7% of patients randomized to apixaban in ARISTOTLE who met two of three dose reduction criteria [57]. Patients with NVAF who met one of these criteria had a higher risk of stroke or SE and major bleeding, but consistent benefits of apixaban were observed in this patient group compared with patients with NVAF without renal dysfunction [57]. It is also noteworthy that an additional 904 patients in the ARISTOTLE trial met one, but not two, of the dose reduction criteria. In a secondary analysis of those patients, Alexander et al. reported that patients who received the 5 mg twice-daily dose and met the one dose-reduction criterion of serum creatinine ≥ 1.5 mg/dL demonstrated lower rates of major bleeding and maintained efficacy benefits compared with warfarin. This was consistent regardless of whether renal function was assessed by serum creatinine (as in the dose reduction criteria) or CrCl calculated by the Cockcroft–Gault formula [56]. The ARISTOTLE trial is unique among the pivotal phase III DOAC trials for NVAF as patients with renal function CrCl ≥ 25 mL/min were included (as opposed to ≥ 30 mL/min in other DOAC trials). Another subgroup analysis reviewed the 269 patients from ARISTOTLE with NVAF and advanced CKD with CrCl 25–30 mL/min [58]. This analysis showed that the rates of stroke or SE, major bleeding, the combination of major plus clinically relevant non-major bleeding, and intracranial bleeding were all directionally consistent and favored apixaban versus warfarin in patients with CrCl 25–30 mL/min compared with the remainder of the patients enrolled in the trial. This analysis also revealed that the total exposure of apixaban, as measured by AUC at steady state, showed substantial overlap in levels among patients with mildly, moderately, or severely reduced kidney function who received the 5 mg dose (Fig. 2) [58]. For patients who met dose reduction criteria and received the 2.5 mg dose, the total exposure was similar across all ranges of CrCl values (Fig. 3), leading the authors to conclude that the dose reduction criteria for apixaban provide an optimal balance of safety and efficacy in patients with advanced CKD [58].

Fig. 1.

Forest plot of the effect of apixaban versus warfarin for outcomes of stroke or systemic embolism, mortality, and major bleed according to kidney function estimated with the Cockcroft–Gault, CKD-EPI, and cystatin C; insights from the ARISTOTLE study [55]. Interaction p-values are based on categorical estimated glomerular filtration rates. CI confidence interval, CKD-EPI Chronic Kidney Disease Epidemiology Collaboration, eGFR estimated glomerular filtration rate.

Figure reproduced from Hohnloser SH et al. Efficacy of apixaban when compared with warfarin in relation to renal function in patients with atrial fibrillation: insights from the ARISTOTLE trial. Eur Heart J 2012;33:2821–2830 by permission of the European Society of Cardiology

Fig. 2.

Daily AUC at steady state for patients (N = 2805) receiving apixaban 5 mg twice daily in the ARISTOTLE trial by level of estimated CrCl [58]. AUC area under the curve, CrCl creatinine clearance.

Figure reproduced from Stanifer JW et al. Apixaban Versus Warfarin in Patients With Atrial Fibrillation and Advanced Chronic Kidney Disease. Circulation 2020;141:1384–1392 by permission of the American Heart Association

Fig. 3.

Daily AUC at steady state for patients (N = 128) receiving apixaban 2.5 mg twice daily in the ARISTOTLE trial [58]. AUC area under curve, CrCl creatinine clearance.

Figure reproduced from Stanifer JW et al. Apixaban Versus Warfarin in Patients With Atrial Fibrillation and Advanced Chronic Kidney Disease. Circulation 2020;141:1384–1392 by permission of the American Heart Association

Finally, additional subanalyses of ARISTOTLE focusing on other clinically pertinent topics have also included data incorporating kidney function. For example, Halvorsen et al. assessed the impact of renal function stratified by age. Of particular interest was the group of patients in ARISTOTLE aged ≥ 75 years with CrCl ≤ 30 mL/min. The results from this older, renally impaired cohort remained consistent with the overall trial results, which favored apixaban for both efficacy and safety over warfarin [59]. An additional subanalysis with a focus on age and anti-Xa levels was published by Kalaria et al., which reported that CrCl did not affect the anti-Xa activity of apixaban and concluded that dose reduction criteria yield consistent clinical and PD outcomes in NVAF patients [60].

The AVERROES trial investigated the efficacy and safety of apixaban compared with aspirin in patients for whom VKA therapy was unsuitable [23]. The study was stopped early after a median follow-up of 1.1 years due to the superior efficacy of apixaban over aspirin. Apixaban reduced the risk of stroke or SE without significantly increasing the risk of major bleeding or intracranial hemorrhage, and was consistent across subgroups of patients for whom VKA therapy was unsuitable, irrespective of the reason VKA therapy was deemed unsuitable [23, 61]. Additionally, among patients in the AVERROES trial with NVAF and stage 3 CKD (mean eGFR 49 mL/min/1.73 m²), the treatment effect of apixaban compared with aspirin was preserved for both efficacy in stroke prevention and safety in incidence of major bleeding [62]. A further substudy of AVERROES by Bhagirath et al. noted that eGFR was one of multiple variables that could independently affect anti-Xa activity in patients receiving apixaban; however, there was no significant association between the primary efficacy endpoint of stroke or SE and anti-Xa activity. This substudy reported an association between anti-Xa activity with minor, but not major bleeding. Overall, the Bhagirath analysis revealed that, despite considerable variability in anti-Xa activity among patients with NVAF receiving apixaban in AVERROES, low rates of major bleeding and stroke/SE irrespective of anti-Xa activity were observed. Thus, the authors concluded that these clinical trial data did not support routine laboratory monitoring of anti-Xa levels [63].

The post-registrational AUGUSTUS (An Open-Label, 2×2 Factorial, Randomized Controlled, Clinical Trial to Evaluate the Safety of Apixaban Versus Vitamin K Antagonist and Aspirin Versus Aspirin Placebo in Patients With Atrial Fibrillation and Acute Coronary Syndrome or Percutaneous Coronary Intervention) [50] trial compared an apixaban strategy with a VKA strategy with or without aspirin (aspirin or matching placebo) in a 2×2 factorial design in patients with NVAF and acute coronary syndrome (ACS) with and without percutaneous coronary intervention [50]. The AUGUSTUS trial enrolled patients with a CrCl > 30 mL/min calculated by Cockcroft–Gault, and all patients received a P2Y₁₂ inhibitor (primarily clopidogrel) along with either anticoagulant plus aspirin or anticoagulant plus placebo. Significantly less International Society on Thrombosis and Haemostasis (ISTH) major bleeding or clinically relevant non-major bleeding was reported, and there were fewer deaths or hospitalizations in the apixaban group compared with those receiving a VKA-based strategy. In a prespecified subgroup analysis by Hijazi et al., the risk–benefit balance of antithrombotic therapy across kidney function categories was evaluated: 30%, 52%, and 19% of patients in AUGUSTUS had an eGFR of > 80, 50–80, and 30–50 mL/min, respectively, as calculated by CKD-EPI [64]. Compared with VKAs, patients who received apixaban had lower rates of clinically relevant non-major bleeding or major bleeding and ISTH major bleeding across all eGFR categories. Efficacy, as measured by death or hospitalization, also numerically favored apixaban across all kidney function categories [64].

Patients with Venous Thromboembolism (VTE)

The dosing for apixaban for the treatment of DVT and PE differs from that used to reduce the risk of stroke and SE in patients with NVAF; there are no dose adjustments for the treatment of DVT and PE based on age, body weight, or serum creatinine. The main rationale for this approach was based on the clinical importance of not underdosing patients with active clots who may be at high risk of recurrent VTE. Dose reduction to 2.5 mg twice daily is recommended during the extended treatment of VTE after the initial phase of treatment; however, this dose reduction is not related to age, body weight, or kidney function [22, 28].

The two registrational RCTs that have evaluated apixaban use in patients with VTE, or prevention of recurrent VTE, and included patients with varying levels of kidney impairment, are AMPLIFY (Apixaban for the Initial Management of Pulmonary Embolism and Deep-Vein Thrombosis as First-Line Therapy) [27] and AMPLIFY-EXT (Apixaban for Extended Treatment of Venous Thromboembolism) [28]. The AMPLIFY trial compared the efficacy and safety of apixaban with low-molecular-weight heparin (LMWH) followed by warfarin (i.e., a conventional therapy) [27]. In AMPLIFY, apixaban was noninferior to LMWH/warfarin for the primary efficacy endpoint of reducing the risk of recurrent VTE, and showed superiority for the primary safety endpoint of ISTH major bleeding compared with LMWH/warfarin [27]. Results were consistent across the spectrum of kidney function; specifically, the effect of apixaban in the AMPLIFY trial was studied according to the following groups of kidney impairment: normal (> 80 mL/min), mild (< 50 to ≥ 80 mL/min), moderate (< 30 to ≤ 50 mL/min), and severe (≤30 mL/min). The results of this subanalysis demonstrated similar benefits favoring apixaban compared with LMWH/VKA for both efficacy and safety [27]. These data are summarized in Fig. 4. Of note, as with other phase III trials, patients with CrCl < 25 mL/min were excluded from the AMPLIFY trial [28].

Fig. 4.

a Primary efficacy and b safety outcomes according to kidney function, from the AMPLIFY trial [27].

Figure reproduced from Agnelli G et al. Oral Apixaban for the Treatment of Acute Venous Thromboembolism. N Engl J Med 2013;369:799–808. Copyright © 2013 Massachusetts Medical Society with permission from Massachusetts Medical Society

The AMPLIFY-EXT trial studied two doses of apixaban compared with placebo for secondary prevention of a recurrent VTE in patients demonstrating clinical equipoise for continuing extended duration AC. The labeled dosing for this indication is 2.5 mg twice daily [65], but both the 2.5 mg twice-daily and 5 mg twice-daily doses were evaluated with respect to kidney function [28]. The efficacy and safety benefits of apixaban were preserved across all levels of kidney function, including normal, mild, moderate, and severe kidney impairment [28].

The post-registrational study, AVERT (Apixaban for the Prevention of Venous Thromboembolism in High-Risk Ambulatory Cancer Patients) assessed the efficacy and safety of apixaban 2.5 mg twice daily for thromboprophylaxis in 574 ambulatory patients with cancer and an intermediate-to-high risk of VTE (Khorana score ≥ 2) who were starting chemotherapy. Patients who received apixaban had a significantly lower VTE rate than those who received placebo [66]. Overall, 66 patients (12%) with CrCl < 60 mL/min did not have a higher risk of thrombotic or bleeding complications compared with 508 patients (89%) with CrCl ≥ 60 mL/min. There was a low rate of VTE and bleeding events in the 66 patients with kidney impairment, and differences between apixaban and placebo in either the VTE or major bleeding endpoints were unable to be demonstrated (there was one VTE event and one major bleeding event between both study arms) [67].

The CARAVAGGIO (Apixaban for the Treatment of Venous Thromboembolism Associated with Cancer) trial evaluated 1170 patients with cancer who had VTE and received apixaban or LMWH [68]. The main analysis revealed non-inferiority of apixaban versus dalteparin in objectively confirmed recurrent VTE and no differences in major bleeding between the groups. A prespecified subanalysis of 1142 patients from CARAVAGGIO investigated the influence of the degree of renal dysfunction on treatment outcomes. Results showed a similar incidence of major bleeding in patients with moderate renal impairment (CrCl 30–59 mL/min, n = 275) and in those with no or mild renal impairment (60–89 mL/min, n = 444), with a hazard ratio (HR) of 1.06 (95% CI 0.53–2.11). Additionally, treatment with either apixaban or dalteparin in patients with moderate renal impairment demonstrated similar results for recurrent VTE and major bleeding [51].

Patients on Hemodialysis

As noted previously, there are limited apixaban clinical trial data for patients undergoing hemodialysis, and future larger RCTs are needed to determine the optimal regimen in this population. Concern has been expressed regarding long-term apixaban use due to increased bleeding events in patients with ESKD receiving chronic hemodialysis [69]. The RENAL-AF (Renal Hemodialysis Patients Allocated Apixaban Versus Warfarin in Atrial Fibrillation) [70] and AXADIA-AFNET 8 (A Randomized Controlled Study Comparing Apixaban with Phenprocoumon in Patients on Hemodialysis) [71] trials were prospective, randomized, open-label studies conducted specifically in patients with NVAF and ESKD on hemodialysis treated with apixaban versus VKA. The RENAL-AF trial evaluated whether apixaban was non-inferior to warfarin for major or clinically relevant non-major bleeding. However, the trial was stopped prematurely because of enrollment challenges; therefore, it had inadequate power to detect differences in rates of major or clinically relevant non-major bleeding between treatments. Clinically relevant bleeding events were approximately tenfold more frequent than stroke or SE, further highlighting the need for future RCTs in patients with NVAF on hemodialysis [70]. A PK subanalysis of 50 patients enrolled in RENAL-AF characterized PK parameters at days 3 and 28 [70]. The exposure data from RENAL-AF, specifically the AUC over 12 h (AUC₁₂), were compared with data from ARISTOTLE. The median steady-state AUC₁₂ was 2475 ng/mL×h (10th–90th percentiles 1342–3285) and 1269 ng/mL×h (10th–90th percentiles 615–1946) for patients who received 5 and 2.5 mg twice-daily apixaban doses, respectively. The AUC₁₂ for the 5 mg dose in RENAL-AF did not differ from the AUC₁₂ for patients from ARISTOTLE with estimated CrCl 45–59 mL/min, 30–44 mL/min, and 15–29 mL/min. The AUC₁₂ for the 5 mg dose in RENAL-AF was significantly higher than for patients with normal kidney function (CrCl ≥ 90 mL/min) from the ARISTOTLE trial. The AUC for the 2.5 mg dose in RENAL-AF did not differ from the AUC for patients with estimated CrCl ≥ 15 and < 90 mL/min (i.e., those with impaired renal function) from the ARISTOTLE trial. The RENAL-AF investigators conducted independent modeling for eCrCl ≤ 30 mL/min. The authors concluded that in patients with ESKD, apixaban exposures did not differ from those observed in patients with CKD in prior trials [70]. These data are summarized in Fig. 5.

Fig. 5.

Exposure comparison of 12-h AUC for apixaban 5 mg twice daily by different categories of kidney function [70]. AUC₁₂ area under the curve from 0 to 12 h, CKD chronic kidney disease, ESKD end-stage kidney disease, HD hemodialysis.

Figure reproduced from Pokorney SD et al. Apixaban for Patients With Atrial Fibrillation on Hemodialysis: A Multicenter Randomized Controlled Trial. Circulation 2022;146:1735–1745 by permission of the American Heart Association

In AXADIA-AFNET 8, the primary outcome measure was a composite of all-cause death, major bleeding, and clinically relevant non-major bleeding events in accordance with the ISTH consensus. The dosing strategy for apixaban in AXADIA-AFNET 8 was different from the dosing used in the RENAL-AF trial; apixaban 2.5 mg twice daily was used for all trial participants randomized to the apixaban arm. Like RENAL-AF, this trial was unable to recruit the planned number of patients. In total, 48 patients received apixaban and 49 received phenprocoumon. The low enrollment resulted in failure to reach the calculated number of patients and events needed to show statistical significance. Composite primary efficacy outcome events occurred in 10 patients receiving apixaban and 15 patients receiving phenprocoumon. No significant differences were observed in efficacy or safety outcomes between apixaban (2.5 mg twice daily) and phenprocoumon in patients with AF on hemodialysis, and treated patients remained at high risk of cardiovascular events [71]. However, the information may still be informative for the shared decision-making conversation with patients.

In summary, data from the RCTs in patients with NVAF and VTE demonstrate that apixaban has similar or superior efficacy and safety to VKAs and/or LMWH regardless of the levels of kidney function, although RCT data for patients with ESKD on hemodialysis are still limited.

Real-World Data

While RCTs offer comprehensive outcomes that serve as a gold standard for the assessment of therapeutic efficacy and safety, few patients with CKD were included, and patients with kidney failure were excluded from the pivotal phase III clinical trials. Real-world data (RWD) analyses from sources such as registries and databases can provide clinical insights into these populations. Observational RWD studies only show associations, not causality, of the treatment effect and cannot completely rule out confounding factors. However, RWD analyses can offer advantages by including a much larger number of patients in the clinical practice setting and including those who would have been underrepresented or excluded from clinical trials. The RWD for apixaban in patients with kidney impairment and NVAF or VTE are summarized here and in Table 6 [72–83].

Table 6.

Summary of real-world studies of apixaban treatment for NVAF and VTE in patients with kidney impairment

RWD study	Cohort size Apixaban: A Warfarin: W	Data source	Years captured	Kidney function	Outcomes: Apixaban versus warfarin HR (95% CI) or comments
					Stroke/SE or VTE	Bleeding
NVAF
Lee, 2022 [72]	A: 958 W: 2189	Research database	2001–2017	Normal to moderate impairment	0.87 (0.48–1.56)	0.66 (0.31–1.42)
Herndon, 2020 [73]	A: 54 W: 57	US VA database	2013–2019	CKD 4 and 5	No difference	No difference
Elis, 2021 [74]	A: 97 W:155	Israel EMD	2014–2017	CKD 5	No difference	No difference
Fu, 2021 [75]	A:1625 W: 1625	Taiwan EMD	2004–2018	Normal to CKD 5	0.74 (0.57–0.97)	0.78 (0.6–1.00)
Yao, 2020 [76]	A: 10,880 W: 10,680	US claims database and EMD	2010–2017	Normal to CKD 5	0.57 (0.43–0.75)	0.51 (0.44–0.61)
Wetmore, 2020 [77]	A: 6738 W: 10,529	Medicare database	2011–2017	CKD 3–5	0.70 (0.51–0.96)	0.47 (0.37–0.59)
Wetmore, 2022 [78]	A: 4239 W: 12,571	USRDS	2013–2018	ESKD on dialysis	0.89 (0.65–1.21)	0.67 (0.55–0.81)
Siontis, 2018 [79]	A: 2351 W: 23,172	USRDS	2010–2015	ESKD on dialysis	0.88 (0.69–1.12)	0.72 (0.59–0.87)
VTE
Knueppel, 2022 [80]	A: 203 W: 0	UC database	2013–2021	CKD 4–5 or on dialysis	Not evaluated	Not evaluated
Ellenbogen, 2022 [81]	A: 2302 W: 9263	USRDS	2014–2018	ESKD on dialysis	0.83 (0.69–1.002)	0.81 (0.70–0.94)
Wetmore, 2022 [82]	A: 3130 W: 9086	USRDS	2013–2018	ESKD on dialysis	0.58 (0.43– 0.77)	0.78 (0.62–0.98)
Cohen, 2022 [83]	A: 10669 W: 19121	5 US databases	2014–2019	CKD 1–5	0.78 (0.66–0.92)	0.76 (0.65–0.88)

CI confidence interval, CKD chronic kidney disease, EMD electronic medical record data, ESKD end-stage kidney disease, HR hazard ratio, NVAF nonvalvular atrial fibrillation, RWD real-world data, SE systemic embolism, UC University of California, USRDS United States Renal Data System, VA Veterans Affairs, VTE venous thromboembolism

Patients with NVAF

Rhee et al. conducted a network meta-analysis evaluating DOAC studies in patients with AF and kidney impairment, including advanced CKD with CrCl < 30 mL/min [84]. This analysis included 19 studies (5 RCTs and 14 observational RWD studies) and 124,628 patients. After propensity score matching (PSM) to reduce the confounding factors of different treatment groups, the authors noted significantly lower risks of stroke or thromboembolism (HR_pooled 0.78, 95% CI 0.73–0.85) and major bleeding (HR_pooled 0.76, 95% CI 0.64–0.89) with DOACs compared with warfarin, regardless of the severity of kidney impairment [84]. The analysis also showed apixaban and edoxaban had superior effectiveness and safety compared with warfarin. Of all DOACs assessed, apixaban was associated with the lowest risk of major bleeding in patients with kidney impairment (CrCl < 60 mL/min, HR_pooled 0.57, 95% CI 0.46–0.70) and with advanced CKD (CrCl < 30 mL/min, HR_pooled 0.34, 95% CI 0.14–0.83) compared with warfarin [84].

Comparing DOACs or apixaban versus VKAs, retrospective cohort real-world studies involving small numbers of patients with AF and kidney impairment in general failed to show differences in effectiveness (reducing stroke or SE) and safety (major bleeding) [72, 73, 85, 86]. For example, Lee et al. conducted a cohort study consisting of 3525 patients with AF with moderate renal impairment (CrCl < 60 mL/min), 2846 patients with mild renal impairment (CrCl < 90 mL/min), and 1153 patients with normal renal function (CrCl ≥ 90 mL/min) [72]. This study noted no major differences in bleeding or thromboembolic risk for apixaban compared with warfarin across renal impairment range when appropriate DOAC dose reductions were made. Similarly, Herndon et al. conducted a retrospective chart review that included 111 patients from a Veterans Affairs database with stage 4 and 5 CKD or on dialysis that showed no significant difference in major bleeding rate for apixaban compared with warfarin [73]. Furthermore, from a subanalysis of a prospective national registry of patients with AF and renal impairment treated with apixaban (n = 97) or warfarin (n = 155), Elis et al. demonstrated no significant difference in the rates of stroke, SE, or major bleeding [74].

In contrast, in RWD studies that have larger sample sizes, the therapeutic effectiveness and safety of apixaban compared with warfarin emerge in a statistically significant manner. Fu et al. examined the treatment effect of apixaban versus warfarin using electronic medical record (EMR) data collected in Taiwan of 1625 matched patients with AF and kidney impairment, including a subset of patients with eGFR < 30 mL/min/1.73 m². This analysis revealed that compared with warfarin, apixaban was associated with a lower risk of stroke/SE (HR 0.74, 95% CI 0.57–0.97). For major bleeding, apixaban was associated with lower risk with a standard dose (HR 0.66, 95% CI 0.45–0.96) but not with lower doses (HR 0.84, 95% CI 0.63–1.12) [75]. Yao et al. analyzed the US administrative claims database with linked laboratory data of 34,569 patients with AF and eGFR ≥ 15 mL/min/1.73 m². These authors observed that relative to warfarin, the proportion of DOAC use decreased with declining kidney function. Compared with warfarin, apixaban was associated with lower risks of stroke (HR 0.57, 95% CI 0.43–0.75), major bleeding (HR 0.51, 95% CI 0.44–0.61), and mortality (HR 0.68, 95% CI 0.56–0.83). There was no significant interaction between treatment and categories of eGFR [76]. Wetmore et al. studied 22,739 Medicare patients with AF and stage 3, 4, or 5 CKD. In the as-treated analysis for stroke/SE, the HRs, all compared with warfarin, were 0.70 (95% CI 0.51–0.96) for apixaban, 0.80 (95% CI 0.54–1.17) for rivaroxaban, and 1.15 (95% CI 0.69–1.94) for dabigatran. For major bleeding, HRs were 0.47 (95% CI 0.37–0.59) for apixaban, 1.05 (95% CI 0.85–1.30) for rivaroxaban, and 0.95 (95% CI 0.70–1.31) for dabigatran versus warfarin. There was no difference in the risk of all-cause mortality between the DOACs and warfarin [77].

The early data on the use of apixaban in patients with CrCl < 15 mL/min or in those receiving dialysis are based on PK data that showed concentrations of apixaban and PK/PD activity similar to ARISTOTLE, but it is not known whether these concentrations will lead to similar stroke reduction as was seen in the study. As previously mentioned, apixaban registrational clinical trials excluded patients with CrCl < 25 mL/min. Several RWD studies were applied to the assessment of therapeutic efficacy and safety in dialysis patients. Wetmore et al. and Siontis et al. independently conducted two retrospective cohort analyses using the United States Renal Data System (USRDS) database for patients with AF receiving dialysis. These studies comprised 17,156 and 25,523 patients, respectively [78, 79]. Both database analyses reported that apixaban was associated with no significant differences in the risk of stroke/SE, but with a lower risk of major bleeding, compared with warfarin. In addition, Wetmore et al. noted that label-concordant apixaban dosing was associated with a lower mortality compared with warfarin; however, this was not seen with the reduced dose of apixaban [78]. Siontis et al. also noted that the 5 mg twice-daily apixaban dose was associated with significantly lower risk of stroke/SE and mortality compared with the 2.5 mg twice-daily dose in the US dialysis dataset [79].

In a smaller, retrospective evaluation of the efficacy and safety of apixaban (5 mg twice daily [n = 22] vs. 2.5 mg twice daily [n = 73]) in patients with NVAF or VTE with stage 4 or 5 CKD, including those on hemodialysis, Yun et al. demonstrated no significant difference between dose groups in major bleeding events, any bleeding event, ischemic stroke, or VTE [87]. A subgroup analysis demonstrated similar results in patients with a serum creatinine > 2.5 mg/dL or CrCl < 25 mL/min and patients on hemodialysis [87].

As kidney function is known to decline over time, a multicenter, prospective cohort study investigated the decline of eGFR in patients with AF treated with VKAs or DOACs (n = 1667; follow-up ≥ 1 year) and concluded that apixaban and dabigatran were associated with a lower risk of CKD stage progression compared with warfarin [88].

Patients with VTE

For the treatment of VTE, recent RWD publications have reported that compared with VKAs, apixaban was associated with significantly lower risk of recurrent VTE and bleeding, including among patients with late-stage CKD [83, 89, 90]. Cohen et al. evaluated the efficacy and safety of apixaban versus warfarin in patients with VTE at high risk of bleeding, defined as one or more of the following risk factors: age ≥ 75 years; antiplatelet, nonsteroidal anti-inflammatory drug [NSAID]/corticosteroid use; prior gastrointestinal bleeding or gastrointestinal-related conditions; or late-stage CKD. A total of 88,281 patients from the database were included, and apixaban was found to be associated with significantly lower risks of recurrent VTE, major bleeding, and clinically relevant non-major bleeding regardless of bleeding risk factors [89]. Cohen et al. also conducted a retrospective cohort analysis of five US claims databases that included 29,790 patients with VTE and CKD (10,669 apixaban, 19,121 warfarin). The study noted that after inverse probability of treatment weighting to balance patient characteristics between treatment cohorts, the apixaban group was associated with significantly lower risks of recurrent VTE (HR 0.78, 95% CI 0.66–0.92), major bleeding (HR 0.76, 95% CI 0.65– 0.88), and clinically relevant non-major bleeding (HR 0.86, 95% CI 0.80–0.93) compared with the warfarin group. The study also noted that CKD stages did not have a significant impact on treatment effects for recurrent VTE and major bleeding [83]. In a single-center, retrospective cohort analysis using the New York University Langone Health database, 56 patients with VTE and CKD stage 4 or greater were identified. Ahuja et al. noted that DOACs were better tolerated compared with warfarin in these patients [90].

Similarly to NVAF, treatment of VTE patients with ESKD on dialysis is a significant medical concern. Multiple single-center, retrospective cohort analyses using EMR noted that patients treated with apixaban had no difference in bleeding rates compared with warfarin in patients on dialysis [85, 86, 91]. Knueppel et al. conducted a multicenter, retrospective cohort study of patients with VTE and severe or end-stage kidney disease (n = 203, CrCl < 25 mL/min, serum creatinine > 2.5 mg/dL, stage 4 or 5 CKD, or on dialysis) that compared bleeding rates of standard-dose (5 mg twice daily) and reduced-dose (2.5 mg twice daily) apixaban. The study noted that the rate of clinically relevant bleeding was higher in the standard-dose group and rates of VTE recurrence were similar in both groups [80]. The authors concluded that a reduced apixaban dose may be considered in this patient population.

Using the USRDS database, Wetmore et al. and Ellenbogen et al. independently conducted retrospective cohort analyses comparing the safety and effectiveness of apixaban versus warfarin [81, 82]. In the study by Wetmore et al., the authors noted that in the 12,206 patients identified, apixaban was associated with lower risks of both recurrent VTE (HR 0.58, 95% CI 0.43–0.77) and major bleeding (HR 0.78, 95% CI 0.62–0.98) compared with warfarin [82]. Ellenbogen et al. found that among the apixaban- (n = 2302) and warfarin-treated (n = 9263) patients, apixaban was associated with lower risks of major bleeding (HR 0.81, 95% CI 0.70–0.94), intracranial bleeding (HR 0.69, 95% CI 0.48–0.98), and gastrointestinal bleeding (HR 0.82, 95% CI 0.69–0.96), but with a similar risk of recurrent VTE compared with warfarin (HR 0.83, 95% CI 0.69–1.002) [81].

Taken together, in RWD studies involving fewer patients, the data are inconsistent; however, including larger numbers of patients or through meta-analysis, a consistent treatment effect and safety profile emerge when apixaban is compared with warfarin in both patients with NVAF or VTE across various stages of CKD, including those on dialysis. In general, apixaban is associated with equal or significantly lower thrombotic events (stroke/SE for NVAF and recurrence of VTE for VTE) and significantly lower safety events (major bleeding) compared with warfarin, across the spectrum of renal dysfunction.

Guidelines for the Management of NVAF or VTE in Patients with Kidney Impairment

Several clinical practice guidelines provide recommendations and guidance related to apixaban use within the context of kidney impairment. These guideline recommendations are summarized in Table 7 [92–98] and the online resources. Importantly, multiple guidelines currently list VKAs as no longer being preferred over DOACs, including apixaban, in patients with renal impairment for the treatment of NVAF and VTE. In addition to incorporating guidance per approved drug labeling, certain national and international bodies have also assessed available clinical trial data and real-world evidence in the renally impaired population to support guideline recommendations. Upon review of the literature, it is readily apparent that the majority of guidelines on DOAC use in patients with renal impairment focus on NVAF with limited input on VTE. It should be remembered that dosing recommendations are based on the individual guideline’s assessment of the strength of evidence for each recommendation.

Table 7.

Treatment guideline summaries

Organization	Treatment recommendation summary
Chest Physicians Evidence-Based Clinical Practice AF Guidelines [92]	For patients with AF, recommend bleeding risk assessment performed at every patient contact and should initially focus on potentially modifiable bleeding risk factors such as bleeding tendency or predisposition (e.g., treat gastric ulcer, optimize renal or liver function) In ESRD (CrCl < 15 mL/min or dialysis-dependent), suggest that individualized decision making is appropriate  Use well-managed VKA with TTR > 65–70%  DOACs should generally not be used, although in the US, apixaban 5 mg twice daily is the recommended dose in NVAF patients receiving hemodialysis
AHA/ACC/HRS Guideline for the Management of Patients with Atrial Fibrillation [93]	Renal function should be regularly monitored and CrCl calculated at an interval that depends on the individual degree of renal dysfunction and likelihood of fluctuation, and dose adjustments should be made according to US FDA dosing guidelines Use of warfarin or apixaban might be reasonable in patients with AF who have a CHA2DS2-VASc score of ≥ 2 (men) or ≥ 3 (women) who have ESRD or are on dialysis, but further study is warranted
ESC Guidelines [94]	Apixaban dose in patients with eGFR > 30 mL/min 5 mg twice daily (or if dose reduction required, 2.5 mg twice daily) Apixaban dose in patients with eGFR 15–29 mL/min, 2.5 mg twice daily with caution eGFR < 15 mL/min do not use
CHEST VTE Guideline and Expert Panel Report [95]	DOACs and LMWH contraindicated with severe renal impairment Dosing of DOACs with levels of renal impairment differ with the NOAC In patients with renal disease and CrCl < 30 mL/min, a VKA is preferred
AFNET/EHRA Consensus [96]	Knowledge of kidney and liver function is required as all NOACs are eliminated to some extent via the kidneys, and renal function affects NOAC dosing Importantly, kidney function should be assessed using the Cockcroft–Gault formula as it was used in the four pivotal phase III trials In view of NOAC pharmacokinetics, dose-reduction criteria, and available evidence from randomized controlled trials, the use of either apixaban or edoxaban may be preferable in patients with severe CKD (CrCl of 15–29 mL/min) The decision to anti-coagulate and (if so) whether to use an NOAC or VKA in patients with ESRD or on dialysis requires a high degree of individualization
ESC Guidelines for Diagnosis and Management of Acute Pulmonary Embolism [98]	NOACs are not recommended for patients with severe kidney impairment
Canadian Cardiovascular Society/Canadian Heart Rhythm Society Comprehensive Guidelines for the Management of Atrial Fibrillation [97]	For patients with AF who are receiving an OAC, renal function should be assessed at baseline and at least annually (eGFR 30–60 mL/min every 6 months, eGFR 15–30 mL/ min every 3 months). In patients with fluctuating renal function or acute dehydrating illness, renal function should be assessed more frequently For patients with CKD, AC was provided according to the patients’ risk of stroke/systemic embolism and the severity of renal dysfunction In patients with a CrCl < 15 mL/min, the decision to use antithrombotic therapy should be individualized on the basis of physician and patient preference and considering the relative risks of stroke and bleeding

AC anticoagulation, ACC American College of Cardiology, AF atrial fibrillation, AHA American Heart Association, CKD chronic kidney disease, CrCl creatinine clearance, DOAC direct oral anticoagulant, eGFR estimated glomerular filtration rate, EHRA European Heart Rhythm Association, ESC European Society of Cardiology, ESRD end-stage renal disease, FDA Food and Drug Administration, HRS Heart Rhythm Society, LMWH low-molecular-weight heparin, NOAC novel oral anticoagulant, NVAF nonvalvular atrial fibrillation, TTR time in the therapeutic range, US United States, VKA vitamin K antagonist, VTE venous thromboembolism

Discussion

This review provides the healthcare community with a comprehensive resource of available evidence to date of apixaban benefit–risk in comparison with VKA for patients with kidney impairment. The available PK, PD, clinical trial, observational, and guideline data identified and summarized here support the currently approved dosing regimen of apixaban for reducing the risk of stroke/systematic embolism in patients with NVAF, in the treatment of DVT/PE, and in the prophylaxis of DVT after hip or knee replacement surgery across all ranges of kidney function, including the recommended dose reduction to 2.5 mg taken orally twice daily for the prevention of stroke and SE in patients with NVAF with serum creatinine ≥ 1.5 mg/dL (133 μmol/L) if these patients also meet at least one additional criterion: age greater ≥ 80 years and/or body weight ≤ 60 kg [20, 21].

This review summarizes the current published literature following screening of > 550 apixaban articles. Articles excluded from the search and this review include any that were not published in English, animal studies, comment articles, non-systematic literature reviews, editorials, and letter articles, as the data contained within the omitted articles would be less significant to this review or to those represented in the primary data sources.

Further evaluation of the benefit–risk profile of DOACs in patients with advanced CKD and those on dialysis is needed. Several pertinent phase I–IV clinical trials are currently ongoing. A phase I study evaluating apixaban PK/PD in patients with nephrotic syndrome compared with healthy individuals is currently recruiting participants and aims to identify a well tolerated and effective apixaban dose and administration schedule for future RCTs [99]. Another study enrolling patients with NVAF in Thailand with varying CrCl levels aims to evaluate apixaban PK/PD and clinical outcomes further [100]. An investigation of the effect of LMWH and apixaban on the biochemical coagulation profile in patients in Denmark with nephrotic syndrome (CAPTAIN) is also currently recruiting [101].

The benefit–risk evaluation of oral anticoagulants in patients on hemodialysis is complex and some experts believe that studies should compare apixaban with placebo, not only with VKA as in RENAL-AF and AXADIA-AFNET 8, since the benefit–risk profile of VKAs in this population is not certain. The following ongoing trials may help to expand on DOAC treatment options for patients on dialysis. The evaluation of PK/PD and short-term safety of apixaban in patients with NVAF on hemodialysis in Spain (HEMOCIONA) is currently recruiting [102]. A phase II trial has been completed in patients receiving dialysis and includes a pilot trial of AF management with apixaban or warfarin (SAFE-D) [103]. Furthermore, a phase IV evaluation of hemorrhagic and thrombotic risks of oral AC in comparison with no AC in patients with AF on hemodialysis (AVKDIAL) is ongoing [104]. Another phase IV study assessing the overall safety, tolerability, and efficacy of initiating warfarin treatment in patients with ESKD on dialysis and AF (DANWARD) is recruiting [105]. However, a phase III trial evaluating DOACs in patients with renal impairment and VTE (VERDICT) was unfortunately terminated due to recruitment difficulties [106]. The data from these studies, when available, will further help to determine benefit–risk of DOAC treatment in patients with advanced CKD and those on dialysis.

Conclusion

The available clinical and subsequent real-world evidence to date support the approved dosing recommendations of apixaban in patients with kidney impairment. The healthcare community awaits data to clarify further the efficacy, safety, and dosing regimen of apixaban in patients with advanced CKD and those on dialysis. Unfortunately, no large RCTs in patients with CKD requiring dialysis are anticipated due to trial complexity and the history of poor enrollment in these types of trials. Small, controlled cohort analyses may be more forthcoming and can provide further information on the potential benefit–risk in this important patient population.

Supplementary Information

Below is the link to the electronic supplementary material.

Declarations

Funding

This study was sponsored by Pfizer and Bristol Myers Squibb.

Conflicts of interest

Stephen R. Mandt, Noble Thadathil, Christian Klem, and Dong Cheng are employees and shareholders of Bristol Myers Squibb. Cristina Russ, Patricia L. McNamee, and Kevin Stigge are employees and shareholders of Pfizer.

Ethics Approval

Not applicable.

Consent to Participate

Not applicable.

Consent for Publication

Not applicable.

Code Availability

Not applicable.

Data availability Statement

All data generated and/or analyzed during this study are included in this published article (and its Online Resource files).

Authors’ Contributions

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Objectives: SRM. Drafting of manuscript: SRM, NT, CK, DC. Review of manuscript: SRM, NT, CK, DC, KS, PLM, CR.