Apixaban Concentrations With Lower Than Recommended Dosing in Older Adults With Atrial Fibrillation

Smrithi Sukumar, BA; Markus Gulilat, PhD; Bradley Linton, Pharm D; Steven E. Gryn, MD; George K. Dresser, MD; Jeffrey E. Alfonsi, MD; Ute I. Schwarz, MD, PhD; Richard B. Kim, MD; Janice B. Schwartz, MD


J Am Geriatr Soc. 2019;67(9):1902-1906. 

In This Article

Abstract and Introduction


Objectives: Lower than recommended doses of direct-acting oral anticoagulants are often prescribed to older adults with nonvalvular atrial fibrillation (NVAF). Our goal was to determine the consequences of lower than recommended dosing on plasma apixaban concentrations during the clinical care of older adults with NVAF.

Design: Convenience sample of patients receiving anticoagulation during 2017.

Setting: Academic medical center.

Participants: Stable adults older than 65 years with NVAF receiving apixaban on a chronic basis.

Measurements: Patient age, weight, creatinine, co-medications, and apixaban concentrations.

Results: A total of 110 older adults with NVAF (mean age = 80.4 y; range = 66–100 y with 45% women) were studied. Overall, 48 patients received recommended dosing of 5 mg twice/day, and 42 received lower than recommended dosing. One patient in each category had concentrations below the expected 5% to 95% range at time of peak concentrations. Differences in proportion of apixaban concentrations within or outside expected ranges were not significant between patients receiving lower than recommended doses and those dosed as recommended at 5 mg twice/day (P = .35). However, in patients dosed as recommended with 5 mg twice/day, four had concentrations above the 5% to 95% range for peak levels expected at 3 to 4 hours after dosing; in two, this occurred around the midpoint of the dosing interval. Twenty patients received 2.5 mg twice/day as recommended. One-third had apixaban concentrations higher than expected peak concentrations compared with the clinical trials, and more than two-thirds had levels above the reported median for peak concentrations.

Conclusions: Apixaban concentrations in older adults with NVAF seen clinically were higher than expected based on clinical trial data. The findings raise questions about the optimal dosing of apixaban in older adults with NVAF encountered outside of clinical trials and suggest a role for the monitoring of apixaban concentrations during care of patients that differ from those in randomized trials or when considering dosing outside of published guidelines.


Direct-acting oral anticoagulants (DOACs) are replacing vitamin K antagonists for anticoagulation due to fewer food and medication interactions and simplified dosing and monitoring regimens.[1] Although DOACs were shown to have equivalent or superior efficacy to prevent stroke or systemic emboli in patients with nonvalvular atrial fibrillation (NVAF) with fewer intracranial hemorrhages in randomized trials,[2–4] data are limited on older adults with NVAF during routine clinical care. These patients are often older, more likely to be women, and they have more comorbidities, falls, and higher bleeding risks than those enrolled in clinical trials. Possibly due to these factors, postmarketing analyses of DOAC use in patients with NVAF report prescribed doses often inconsistent with product labeling.[1] Lower than recommended dosing is more common than higher than recommended dosing, especially for apixaban in older patients.[5–8]

The consequences of underdosing are currently uncertain. If underdosing resulted in lower apixaban concentrations, increased stroke rates would be expected, as would lower bleeding rates. Analyses from administrative claims data lacking full assessment of dosing accuracy reported increased stroke rates without increased bleeding in patients with NVAF receiving apixaban classified as "underdosed."[7] Underdosing was also initially reported to result in worse outcomes in patients enrolled in the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation II (ORBIT-AF) registry.[9] However, when the outcomes were adjusted for patient risk characteristics, no significant difference in outcomes between "underdosed" patients compared with patients dosed as recommended was detected in the ORBIT-AF registry.[10] These studies did not determine drug concentrations in relation to dosing or outcomes to investigate potential mechanisms for alterations in responses or outcomes.

Our primary goal was to measure plasma apixaban concentrations during routine clinical care of older adults with NVAF and compare apixaban concentrations between patients receiving recommended vs lower than recommended dosing relative to concentrations reported from the pivotal trial on which marketing approval was granted (ARISTOTLE trial).[11] We found that older adults with NVAF receiving lower than recommended dosing of apixaban had the same proportion of concentrations within the ranges reported from patients receiving recommended doses. Only patients receiving recommended doses had concentrations in excess of those observed in clinical trials.