Serum Concentration of Apixaban in Relation to Renal Function in Older Hospitalized Patients

Abstract

Background and Objective

Apixaban is the most prescribed direct-acting oral anticoagulant drug. According to its product information, clearance is only partially renal. However, little is known about the impact of renal function on apixaban pharmacokinetics in real-world settings. The aim of this study was therefore to investigate serum concentrations of apixaban in relation to renal function in acutely hospitalised, older patients.

Methods

The study was conducted with a prospective, observational design. Apixaban-treated patients ≥ 65 years acutely admitted to Haukeland University Hospital in Bergen, Norway, during a four-month period were included. Serum concentrations of apixaban were measured at hospitalization and assessed in relation to glomerular filtration rate (GFR). Spearman rank test was used to investigate correlation between GFR and dose-adjusted serum concentrations of apixaban. In addition, dose-adjusted serum concentrations were compared between GFR subgroups by Mann–Whitney tests.

Results

In total, 36 patients were included (median age 84.5 years, range 68–96 years). Median GFR at admission was 43 ml/min (range 17–119 ml/min). Dose-adjusted apixaban serum concentrations correlated significantly with GFR (Spearman r = − 0.54, p = 0.0008). Compared with patients with GFR > 90 ml/min, apixaban dose-adjusted serum concentrations were 3.3-fold, 1.8-fold and 2.0-fold higher in patients with GFR < 30 ml/min (p = 0.01), 30–59 ml/min (p = 0.04) and 60–89 ml/min (n.s.), respectively.

Conclusions

The study shows that dose-adjusted serum concentration of apixaban significantly correlates with renal function in older, acute hospitalized patients. These real-life data indicate that apixaban-treated patients with GFR < 30 ml/min may require around 70% lower dose than normal to achieve sufficient antithrombotic effect and prevent risk of bleedings.

Key Points

Dose-adjusted serum concentration of apixaban is strongly correlated to renal function in older patients.

Apixaban-treated patients with severely reduced renal function may require around 70% lower dose than normal to achieve sufficient antithrombotic effect and prevent risk of bleeding.

Introduction

Anticoagulation therapy is of major importance in preventing thromboembolism and cerebrovascular infarction in patients with atrial fibrillation [1–3] as well as venous thrombosis in relation to surgery [4–6]. Warfarin and other vitamin K antagonists have for decades been the only available alternative for oral anticoagulation treatment. During recent years, several direct-acting oral anticoagulants (DOACs) have been registered for clinical use after showing non-inferiority compared with warfarin in preventing thromboembolic events in large clinical trials on patients with atrial fibrillation [7–9]. In these trials, frail older patients were underrepresented, so there is a lack of evidence on the safety of DOAC use in this patient group. Thus, more data on DOAC pharmacokinetics in older patients are necessary to assess safety in a real-life setting.

Apixaban is currently the most prescribed DOAC in Norway. In 2020, two-thirds of DOAC prescriptions were for apixaban. The majority of apixaban-treated-patients are older ≥ 65 years, in 2020 representing 84% of prescriptions [10]. It is therefore of particular relevance to obtain more data concerning safety of apixaban use in older patients. Renal excretion and hepatic CYP3A4 metabolism represent the main elimination pathways of apixaban [11]. According to the product information, 27% of the overall apixaban clearance is renal, while the remaining is mediated by hepatic metabolism. For patients with GFR < 30 ml/min it is recommended to reduce the dose by 50% of normal. For patients with moderate renal failure, i.e., GFR 30–60 ml/min, dose adjustment is recommended for NVAF patients and at least two of the following characteristics: age ≥ 80 years, body weight ≤ 60 kg and serum creatinine ≥ 133 µmol/l [11].

Renal function decreases by age, which affects clearance of drugs eliminated via the kidneys [12–14]. However, knowledge on the impact of renal function on apixaban pharmacokinetics in older patients in real-life settings is limited. Thus, the aim of this study was to investigate dose-adjusted serum concentrations of apixaban in relation to renal function among older, acutely hospitalised patients.

Methods

Patient Inclusion

The study was approved by the Regional Committee for Medical and Health Research Ethics (REK number 2019/1193) and conducted in accordance with the declaration of Helsinki. The project was approved by the medical director on behalf of the hospital prior to study initiation. The study was conducted with a prospective, observational design including patients ≥ 65 years acutely admitted to three hospital wards at Haukeland University Hospital in Bergen, Norway. Patients using apixaban prior to admission were consecutively included in the study during a period of 4 months (January–May 2020), as part of a master thesis in clinical pharmacy.

All participating patients signed the written informed consent approved by the Regional Committee for Medical and Health Research Ethics (REK) form prior to inclusion. Patients terminally ill, with moderate and severe dementia or unable to communicate in Norwegian language, were excluded.

Medication Reconciliation and Blood Sampling

Updated and accurate information about drug use at admission was obtained by medication reconciliation performed by a trained clinical pharmacist using a standard procedure, including an interview guide. Information about potential use of dietary supplements and natural medicines was also identified in this process. The prescribed dose of apixaban and exact time interval between last dose of apixaban and blood sample for concentration measurements were recorded for all patients. Clinical and demographic baseline characteristics of the patients were registered at admission. This included, sex, age, living situation, height, weight, renal function, and haemoglobin level.

Blood samples were drawn according to routine procedure at admission. Serum creatinine concentrations in the samples were applied to estimate glomerular filtration rate (GFR) using the Cockcroft–Gault formula predicting renal function on the basis of creatinine concentration, sex, age and weight. The serum creatinine values used to estimate GFR was from blood samples close in time to apixaban blood sampling.

After hospitalization, an additional drug fasting blood sample was collected for determination of serum concentration of apixaban under controlled conditions. To ensure that time between last dose intake and blood sampling was known, exact apixaban administration time and blood sampling time was recorded for all patients. As recommended apixaban dosage is twice daily, blood samples for determination of serum concentrations were drawn about 12 h after last drug administration, i.e. right before the next dose.

All blood samples for apixaban analysis were collected on glass tubes for isolation of serum. The sample tubes were immediately frozen and stored in the Biobank at Haukeland Hospital. After inclusion of all patients, serum samples were sent for analysis at the therapeutic drug monitoring (TDM) service at Center for Psychopharmacology, Diakonhjemmet Hospital, Norway, where assays for measuring DOAC concentrations had been established [15].

Analytical Assay of Apixaban

Analysis of apixaban serum concentrations was performed by using a certified assay validated according to international standards with intra- and inter-run variability in imprecision and inaccuracy parameters < 10%. Briefly, apixaban was analyzed in serum samples by ultra-performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS). Serum samples were first purified by protein precipitation mixing a 500-µl sample aliquot with 1000-µl acetonitrile-methanol (90/10 vol/vol) solution including ⁴D- or ¹³C-labelled apixaban isotopes as drug-specific internal standard. After centrifugation for 10 min at 1800 g at −2 °C, aliquots of 5-µl purified sample were injected onto an Acquity UPLC system (Waters, Milford, MA, USA), with a Micromass Quattro micro or Quattro premier tandem MS detector (Waters) operated in electrospray ionization-positive mode. Apixaban was detected with the mass transition (m/z) 460 > 199, and chromatographic separation was performed on an Acquity UPLC BEH Shield RP18 column (1.7 µm, 1.0 × 100 mm; Waters). The mobile phase gradient comprised a mixture of acetonitrile and ammonium acetate buffer (pH 4.8). The chromatographic retention time of apixaban was 2.65 min. Lower limit of quantification (LLOQ) of the assay was 4 nmol/l.

Measures and Statistics

The main outcome measure was apixaban dose-adjusted serum concentration in relation to renal function (GFR). Dose-adjusted concentration was determined by dividing concentration by dose (C/D ratio). Non-parametric Spearman test was used to assess the correlation between GFR and apixaban C/D ratio as continuous variables. In addition, C/D ratios were compared between patients with normal (GFR > 90 ml/min; reference group), mild (GFR 60–89 ml (min), moderate (GFR 30–59 ml/min) and severely reduced (GFR < 30 ml/min) renal function using non-parametric Mann–Whitney tests.

In addition to the pharmacokinetic assessment using the C/D ratio as measure, absolute serum concentration of apixaban at a given dosage, which represents the clinically relevant measure, was evaluated in relation to renal function. The target reference concentration range of apixaban has been defined by the European Heart Rhythm Association (EHRA) guidelines to be 34–230 ng/ml (73–500 nmol/l) [16]. Although the concentration boundaries are not strict cut-offs for effect and side effects, they reflect levels generally required to achieve optimal antithrombotic effect with limited risk of bleeding events and were therefore applied as a basis for calculating patient proportions with serum concentrations outside the target range in various GFR subgroups.

GraphPad Prism version 9.4.0 (GraphPad Software, Inc, San Diego, CA, USA), was used for statistical analyses and graphical preparations. A p value < 0.05 was defined as statistically significant.

Results

Patient Characteristics

A total of 36 apixaban patients met the inclusion criteria and accepted the invitation to participate during the 4-month study period. Clinical and demographic characteristics of the patients are presented in Table 1. Median age in the population was 84.5 years (range 68–96 years). At admission, median GFR was 43 ml/min with a substantial interindividual variability (range 17–119 ml/min). The observed GFR was higher in males than in females (Table 1). Median hemoglobin level at admission was within the reference range (11.7–15.3g/dl) for women (12.1 g/dl, range 7.1–14.7). For men median hemoglobin level was 11.1g/dl (range 8.7–14.6) and therefore below the reference range (13.4–17.0 g/dl). The prescribed median daily dose was 10 mg/day for patients with GFR > 60 ml/min, while a median of 5 mg daily was prescribed to patients with GFR < 60 ml/min (Table 2).

Table 1.

Demographic and clinical information of the included patients at hospital admission

Number of patients	36
Median age, years (range)	84.5 (68–96)
Female sex, n (%)	20 (56%)
Body mass index (kg/m2), mean (range)	24.2 (16.5–37.8)
Serum creatinine (µmol/l, median, range)	99 (50–257)
Median GFRa, ml/min (range)
All  Female  Male	43 (17–119) 39 (20–79) 51 (17–119)
Median Hb, mg/l (range), % within reference range
Female  Male	12.1 (7.1–14.7), 55 11.1 (8.7–14.6), 19
Atrial fibrillation, n (%)	34 (94%)
Home residents, n (%)	31 (86%)
Home residents with care service, n (%)	19 (61%)b

Reference range Hb 11.7–15.3 g/dl (female) and 13.4–17.0 g/dl (male)

^aGFR, glomerular filtration rate (measured as creatinine-clearance by Cockcroft-Gault equation)

^bn = total 31 (calculated only for home residents)

Table 2.

Drug use, sampling time, dosing, absolute and dose-adjusted serum concentrations of apixaban in included patients according to renal function

Variable	> 90 ml/min (n = 3); ref.	60–89 ml/min (n = 8)	p	30–59 ml/min (n = 17)	p	< 30 ml/min (n = 8)	p
Number of regular drugs, median (range)	7 (3–12)	9.5 (7–14)	–	8 (3–23)	–	12.5 (6–17)	–
Time since last dose (h), median (range)	9.1 (8.7–11.9)	11.4 (7.6–22.4)	–	10.5 (8.1–21.6)	–	10.6 (8.1–12.0)	–
Median daily dose (mg), (range)	10 (10–10)	10 (5–10)	–	5 (5–10)	0.05	5 (5–5)	0.006
Absolute concentration (nmol/l), median (range)	293 (241–472)	468 (101–981)	–	410 (98–759)	–	483 (370–1027)	–
Dose-adjusted serum-concentration (nmol/l/mg), median (range)	29 (24–47)	58 (10–98)	–	54 (20–110)	0.04	97 (74–205)	0.01
Proportion above reference range (%)	0	50	–	24	–	50	–

p > 0.1 not indicated, reference range for apixaban 73-500 nmol/l

The majority of the patients were living at home (n = 31, 86%), while only five (14%) were residents of nursing homes (Table 1). Median length of hospital stay was 4 days (2–39 years, one patient died during the stay). Non-ventricular atrial fibrillation (NVAF) was the main indication for apixaban treatment (n = 34, 94%) (Table 1), and for 30 patients (83%) apixaban had been administered for more than 6 months. The median number of drugs used regularly by the patients before admission was 8.5 (range 3–23). The five most co-prescribed drug-groups were beta blockers (n = 28), loop diuretics (n = 23), angiotensin receptor blockers/angiotensin-converting enzyme inhibitors (n = 19), statins (n = 13) and proton pump inhibitors (n = 12).

Considering the EHRA guidelines [16] and the apixaban's product information [11], the dose of the drug should be 2.5 mg BID if two out of three fulfilled: weight < 60 kg, age > 80 years and serum creatinine > 133 mmol/l (1.5 mg/dl). The proportion of patients fulfilling criteria for reduced dosing, i.e. 2.5 mg BID, was 44% (16/36). In comparison, 17% (6/36) received lower dose than recommended, and 3% (1/36) received higher dose than recommended in the EHRA guidelines. For one patient who was 96 years old and prescribed 2.5 mg BID, the patients body weight was low (serum creatinine was 109mmol/l), so the appropriateness of the dose could not be assessed.

Absolute Serum Concentrations and Dosing of Apixaban

Time between last apixaban drug administration and blood sampling varied between 7.6 and 22.4 h (median 10.6 h). There was no statistically significant difference in sampling time between patients assigned to the different renal function subgroups (Table 2).

The median serum concentration of apixaban measured during hospital stay was 452 nmol/l and ranged 10-fold (98 to 1027 nmol/l). The absolute median serum concentration of apixaban was approximately 1.6-fold higher in the three subgroups with reduced renal function compared with the reference group with normal GFR > 90 ml/min however, the variabilities were larger in patients with severely reduced renal function (Table 2). The proportions with serum concentrations above the target therapeutic range differed between the subgroups. While 50% of patients with GFR < 30 ml/min had absolute serum concentrations above the upper limit of the target range, the corresponding proportions for patients with GFR 30–59, 60–89 and > 90 ml/min were 24, 50 and 0%, respectively (Table 2).

Apixaban Dose-Adjusted Concentrations in Relation to Renal Function

Dose-adjusted serum concentrations (C/D ratio) of apixaban significantly correlated with the estimated GFR (p 0.0008; r − 0.54) (Fig. 1).

Fig. 1.

Correlation between apixaban dose-adjusted concentration (C/D) ratio and glomerular filtration rate (GFR). Spearman r = − 0.55, p < 0.001

When comparing apixaban C/D ratios according to the five-grade definitions of renal stage function, i.e. normal (1: GFR ≥ 90 ml/min), slightly reduced (2: GFR 60–89 ml/min), moderately reduced (3: GFR 30–59 ml/min) and severely reduced (4: GFR 15–29 ml/min), subgroup 3 and 4 had significantly higher median C/D-ratio than patients with normal renal function (p = 0.04 and 0.01, respectively), (Fig. 2; Table 2). The observed median C/D ratio in subgroup 2 (58 nmol/l/mg) was twice the median C/D-ratio of subgroup 1 (29 nmol/l/mg), but not statistically significant, possibly owing to limited number of patients.

Fig. 2.

Apixaban dose-adjusted concentration (C/D) ratio in subgroups with different renal function/glomerular filtration rate (GFR). Stage 1 = GFR ≥ 90 mL/min (normal renal function), Stage 2 = GFR 60–89 ml/min (slightly reduced renal function), Stage 3 = GFR 30–59 ml/min (reduced renal function) and Stage 4 15–29 ml/min (severely reduced renal function)

Comedication with Pgp- and/or CYP3A4-Inhibitors

Six patients (16.7%) were co-medicated with amiodarone (n = 5) or dronedarone (n = 1), which are known to increase serum levels of apixaban by inhibiting CYP3A4 and/or P-gp. Three of these patients (50%) co-medicated with amidarone or dronedarone had serum concentrations of apixaban above the reference range. Two of the patients with GFR < 30 ml/min used amidarone or dronedarone. Patients co-medicated with amiodarone or dronedarone were randomly distributed between the subgroups.

Discussion

The objective of this study was to investigate serum concentrations of apixaban in older patients with variable renal function in a real-life setting. The findings show that renal function is a highly significant and major predictor of apixaban clearance in older patients and suggest that in patients with GFR < 30 ml/ml, dosing should be reduced by more than 50%, which is recommended in apixaban’s product information. According to our data, patients with GFR < 30 ml/min may require by 65–75% lower dose than normal to avoid an increased risk of dose-dependent side effects, where bleeding is the most relevant. However, it is important to be aware that the subgroup with GFR < 30 ml/min was limited (n = 8), which implies some uncertainties in the estimated dose requirement to achieve sufficient antithrombotic effect without increased risk of bleedings.

DOAC concentrations are related to clinical response. Higher concentrations are reported to be associated with bleeding events [17], while lower concentrations show increased risk of embolic events including strokes [18]. However, clear-cut therapeutic concentration ranges of DOAC are still not implemented in clinical practice [19]. We used a serum concentration reference range based on controlled pharmacokinetic studies of apixaban as an indication of concentrations providing favourable anticoagulation activity with limited risk of bleedings [16]. In the study population, the patient proportions with serum concentrations above this target concentration range increased by decreasing GFR. While none of the included older patients with normal renal function had supratherapeutic concentrations, 50% (4) of all included patients with GFR < 30 ml/min had concentrations above the target range.

Anticoagulation therapy in NVAF patients has for decades been an art of precision, balancing the risk of severe side effects on one side and life-threatening thromboembolic events on the other side. This has been achieved by personalized treatment with the ‘old’ vitamin K antagonists, where dose titration and adjustments were performed by regular monitoring of anticoagulation activity in blood samples [20, 21]. Still, serious side-effects occurred and new direct-acting anticoagulants (DOACs), such as apixaban, were developed with a higher safety margin but with very limited dose flexibility. Under these circumstances apixaban has been shown to be effective (non-inferior to warfarin) and safe during treatment with the two recommended, fixed dosages [9]. However, since only a limited number of older patients > 75 years were included in the ARISTOTLE trial, there has been a lack in knowledge on safety in this population, where reduced renal function is frequent. In the present study, more than 50% of the patients had GFR < 60 ml/min, making the population optimal for investigating the effect of renal function on clearance of apixaban.

Although we did not measure frailty, it is likely that many patients of the study were multimorbid with a generally reduced physiological status. A previous study by Eidam et al. [22] evaluated plasma-concentrations in relation to frailty and found that dose-normalized apixaban plasma trough concentrations were 2.5-fold higher in frail older than in healthy older, which is close to the difference in C/D ratio found between patients with normal GFR versus GFR < 30 ml/min. This shows that alertness is required when prescribing apixaban to older patients with multimorbidity and/or renal failure.

A strong correlation was demonstrated between dose-adjusted serum concentration and GFR. As renal clearance of apixaban is responsible for only 27% of total clearance [11], this would implicate that severely reduced renal function and end stage renal disease would only yield an increase in C/D-ratio of approximately 37%. On the contrary, patients with GFR < 30 ml/min had median C/D-ratio that was 1.83 times higher than the respective C/D-ration of patients with GFR > 30 ml/min. This probably reflects a combination of reduced glomerular filtration and tubular excretion, but it might also relate to chronic kidney disease (CKD) and reduced hepatic CYP3A4 activity. CKD can affect the pharmacokinetics of non-renally eliminated drugs [23]. One theory suggests that accumulation of uremic toxin products because of reduced renal excretion can downregulate hepatic CYP3A4 enzymes. Apixaban is a substrate of CYP3A4, which is estimated to mediate around 25% of the overall clearance [11]. A reduction in hepatic CYP3A4 metabolism has been associated with accumulation of CYP3A4 substrates [24], which is reversible by haemodialysis [25, 26].

In addition, the efflux transporter P-glycoprotein (Pgp) mediates tubular efflux of apixaban. Previous studies have shown that Pgp expression is impaired by age [27, 28], contributing to increased intestinal absorption and reduced renal secretion of Pgp substrates such as apixaban. Therefore, age related Pgp phenotype impairment, and possibly downregulated CYP3A4 activity, may be part of the explanation why many patients have serum concentrations above the reference range.

The findings of the present study should be interpreted with some caution, as the number of included patients was limited, especially in the subgroup with GFR > 90 ml/min. Another scientific limitation is the real-life setting itself, which may affect accuracy of the dose-adjusted concentration estimates. At the same time, real-life data are valuable as they reflect how older patients are treated in clinical practice. Furthermore, a strength of the study was the ability to ensure standardization of the time between dose intake and blood sampling of apixaban.

Conclusions

The study shows that dose-adjusted serum concentration of apixaban significantly correlates with renal function in older, acute hospitalized patients. The data indicate that apixaban-treated patients with GFR < 30 ml/min may require around 70% lower dose than normal to prevent risk of bleeding.

Declarations

Funding

This research did not receive any specific grant from funding agencies in the public, commercial or not for profit sectors.

Conflict of interest

The authors T.C.M. and O.M.S. have no conflicts of interest to report. E.M. received lecturing fees from Lundbeck and Outsuka Pharmaceutical Co. The authors did not receive support from any organization for the submitted work.

Ethics approval

The study was approved by the Regional Committee for Medical and Health Research Ethics (REK number 2019/1193) and conducted in accordance with the declaration of Helsinki. The project was approved by the medical director on behalf of the hospital prior to study initiation.

Data protection, confidentiality, and privacy

The plans for data protection, confidentiality and privacy were reviewed and approved by the Regional Committee for Medical and Health Research Ethics.

Consent to participate

All participating patients signed the written informed consent approved by the Regional Committee for Medical and Health Research Ethics.

Availability of data and material

The datasets generated during and/or analyzed during the current study are not authorized to be publicly available under the current ethics approval. However, any queries regarding the datasets generated and analyzed for this project may be directed to the corresponding author.

Author contributions

All authors contributed to the study concept and design. T.C.M. was responsible for data collection, and all authors contributed to data analysis. The first draft of the manuscript was written by T.C.M., with major edits made by E.M. and O.M.S. All authors reviewed and commented on previous versions of the manuscript and approved the final version. All authors have reviewed and approved the final submitted manuscript and agree to be accountable for its content.