Direct-Acting Oral Anticoagulant Use at Extremes of Body Weight

applyparastyle “fig//caption/p[1]” parastyle “FigCapt”

Clinical Review

Direct-acting oral anticoagulant use at extremes of body weight: Literature review and recommendations Downloaded from https://academic.oup.com/ajhp/article-abstract/77/11/865/5840388 by Biblioteca Nacional de Salud y Seguridad social user on 09 July 2020

Kelly Covert, PharmD, BCPS, Bill Gatton College of Pharmacy, East Purpose. To review the literature on treatment of venous thromboembo- Tennessee State University, Mountain Home, TN lism (VTE) and prevention of cardioembolic stroke with direct-acting oral anticoagulants (DOACs) in low- and high-body-weight patients and to Donald L. Branam, PharmD, BCPS, FASHP, South College School of make recommendations regarding agent selection and dosing in these pa- Pharmacy, Knoxville, TN tient populations.

Summary. The selection and optimal dosing of DOACs in low- and high- body-weight patients has not yet been fully elucidated by clinical trials; however, evidence suggests that issues of both safety and efficacy in patients at the extremes of body weight may warrant careful consideration when selecting a DOAC for such patients. This review provides a thorough discussion of the use of DOACs in the treatment of VTE and prevention of cardioembolic stroke in patients at the extremes of body weight and provides guidance regarding agent selection.

Conclusion. While the published evidence on use of DOACs in patients at extremes of body weight is sparse, apixaban and rivaroxaban appear to have the most favorable safety and efficacy profiles. Edoxaban and dabigatran should be avoided.

Keywords: anticoagulants, atrial fibrillation, factor Xa inhibitors, obesity, thromboembolism

Am J Health-Syst Pharm. 2020;77:865-876

ince Food and Drug Administration however, neither guidance document S(FDA) approval of dabigatran in discusses nuances of DOAC use in pa- 2010, the direct-acting oral anticoagu- tients at the extremes of body weight.10,11 lants (DOACs) have become an in- The Scientific and Standardization creasingly utilized medication class in Committee of the International Society the treatment of venous thromboem- on Thrombosis and Haemostasis has bolism (VTE) and for the prevention of recommended avoidance of DOACs in cardioembolic stroke in patients with patients with a body mass index (BMI) nonvalvular atrial fibrillation (NVAF). of >40 kg/m2 or a weight of >120 kg due Each DOAC has been compared to war- to a lack of supportive clinical data, farin and was found to be noninferior and the committee has advocated for for treatment of VTE and prevention of drug-specific peak and trough levels cardioembolic stroke, and DOACs’ ease if these agents are used in such pa- of administration, fixed dosing, and lack tients.12 Guidance detailing monitoring of monitoring requirements make them of DOACs suggests that direct thrombin appealing alternatives.1-9 In the majority inhibitors may be monitored using ac- of the landmark DOAC trials, patients tivated partial thromboplastin time Address correspondence to Dr. Covert with low (<60 kg) or high (>120 kg) (APTT) or thrombin time (TT) and that ([email protected]). body weight were underrepresented. factor Xa inhibitors may be monitored The most recent American College of using anti–factor Xa activity.13 However, © American Society of Health-System Chest Physicians guidelines clearly use of factor Xa monitoring requires Pharmacists 2020. All rights reserved. For permissions, please e-mail: journals. recommend DOACs over vitamin K drug plasma calibrants, which are not [email protected]. antagonist (VKA) therapy for VTE and widely available. As such, there is cur- DOI 10.1093/ajhp/zxaa059 cardioembolic stroke prophylaxis; rently no reliable way to monitor DOAC

AM J HEALTH-SYST PHARM | VOLUME 77 | NUMBER 11 | JUNE 1, 2020 865 Clinical Review DIRECT-ACTING ORAL ANTICOAGULANTS effectiveness that is widely available. Results KEY POINTS Furthermore, conflicting pharmaco- Use of DOACs for VTE and kinetic and pharmacodynamic ana- • The utility of direct-acting oral NVAF in overweight adults. Tittl lyses have detailed the impact of body anticoagulants (DOACS) in and colleagues21 used data from a Downloaded from https://academic.oup.com/ajhp/article-abstract/77/11/865/5840388 by Biblioteca Nacional de Salud y Seguridad social user on 09 July 2020 14,15 weight on DOAC efficacy. There under- or overweight patients DOAC registry to evaluate the impact are currently no guidance documents continues to be a topic of clin- of BMI on DOAC effectiveness in 3,432 pertaining to DOAC use in low-body- ical debate. patients, 731 of whom had a BMI of 30.0 weight patients. • If a DOAC is to be used in to 34.9 kg/m2 and 346 of whom had a The obesity epidemic has been an overweight patient, use BMI of ≥35 kg/m2. All DOACs were in- well documented in recent years, and of apixaban or rivaroxaban is cluded, and indications for both VTE according to the Centers for Disease supported by the most robust and NVAF were evaluated. In total, Control and Prevention, 39.8% of US data, and dabigatran and 2,104 patients (61.3%) were prescribed 16 adults were obese in 2015-2016. edoxaban should be avoided. rivaroxaban, 685 (20%) were prescribed Obesity is a well-established risk factor apixaban, 348 (10.1%) were prescribed • A thorough evaluation of em- for deep vein thrombosis (DVT), pul- dabigatran, and 295 (8.6%) were pre- bolic and bleeding risk, as well monary embolism (PE), and heart scribed edoxaban. Ultimately, the au- 17 as a discussion with the patient, failure and NVAF. Furthermore, obe- thors found no significant differences should be undertaken prior to sity causes significant pharmacoki- in DOAC safety or effectiveness be- initiation of a DOAC in these pa- netic changes via alterations in volume tween patients with a BMI of <30 kg/ tient populations. of distribution, protein binding, and m2 and those with a BMI of ≥30 kg/m2. clearance. Since elimination of all This study provided evidence to sup- DOACs relies at least partially on renal port the overall effectiveness of DOACs clearance, alterations in renal function in patients with a BMI of >30 kg/m2 but may significantly impact drug removal. did little to define the role in therapy of Obese patients may have increased individual DOACs. Methods kidney size or increased cardiac output, Piran and colleagues15 completed each of which may increase renal clear- A PubMed search covering the pe- a pharmacokinetic study in 38 obese ance; however, obese patients also have riod 2010 to January 2019 was con- patients (weight of >120 kg) taking a a higher incidence of kidney disease ducted using the following keywords or DOAC for any indication to evaluate secondary to type 2 diabetes mellitus, search terms: dabigatran, rivaroxaban, peak DOAC plasma concentrations. which reduces DOAC clearance.18,19 apixaban, edoxaban, body weight, obe- Twenty-one patients (55%) were pre- While a significantly smaller pro- sity AND venous thromboembolism, scribed rivaroxaban, 10 (26%) were portion of the population is under- and obesity AND atrial fibrillation. prescribed dabigatran, and 7 (18%) weight (1.5% in 2015-2016), this Additional references were identified were prescribed apixaban. Within the population also presents significant from a review of citations. dabigatran cohort, 20% of patients had challenges in anticoagulation manage- Retrieved articles and references peak plasma concentrations below ment.20 Underweight patients typically were evaluated for content, and articles the 10th percentile and 2 patients had have reduced adipose tissue, which that evaluated the use of DOACs for VTE peak plasma concentrations below may alter volume of distribution as well treatment or cardioembolic stroke pre- the median trough level from previous as renal clearance. Given the paucity of vention in patients of low or high body pharmacokinetic studies. Within the data and potential for significant phar- weight were included. High body weight rivaroxaban group, 6 patients (29%) macokinetic changes, dosing guidance was defined as a BMI of ≥30 kg/m2 had peak plasma concentrations for DOAC use in patients at the ex- or a body weight of >120 kg, and low below the fifth percentile. Patients tremes of body weight is needed. body weight was defined as a body with a lower peak plasma concentra- The purpose of this review is to pro- weight of <60 kg. Articles evaluating tion tended to be younger (mean age, vide an overview of the available litera- triple therapy (dual antiplatelet therapy 55 years vs 66 years; P = 0.06); however, ture concerning DOAC use in patients plus an anticoagulant) or the utility there was no significant difference in with high or low body weights and to of DOACs for VTE prophylaxis in body weight in patients with low peak make recommendations for drug se- medically ill or bariatric surgery pa- concentrations vs patients with ap- lection in these populations. A critical tients were excluded, as were studies propriate peak concentrations. There assessment of the thromboembolic and involving pediatric patients (<18 years were no patients in the apixaban group bleeding risk of each patient should of age) or pregnant patients. Given the who had low peak drug levels. While also be considered when selecting a paucity of data, all study designs were this study shed light on key pharma- DOAC for treatment of NVAF or VTE. considered. cokinetic differences between DOACs,

866 aM J HEALTH-SYST PHARM | VOLUME 77 | NUMBER 11 | JUNE 1, 2020 DIRECT-ACTING ORAL ANTICOAGULANTS Clinical Review there is currently no clinical correlation Dabigatran. After the RE-COVER dabigatran group who weighed 100 kg between drug levels and effectiveness and RE-COVER II trials, dabigatran or more. These findings lent further cre- or safety in the populations of interest. was approved by FDA for the treat- dence to the notion that dabigatran is

Kushnir and colleagues22 published ment of VTE disease in 2014. The likely as effective as warfarin in patients Downloaded from https://academic.oup.com/ajhp/article-abstract/77/11/865/5840388 by Biblioteca Nacional de Salud y Seguridad social user on 09 July 2020 a retrospective analysis of outcomes in RE-COVER trial was a randomized, weighing more than 100 kg, but given 795 patients with a BMI of ≥40 kg/m2 re- double-blind, noninferiority trial com- the lack of stratification by BMI, the re- ceiving anticoagulation with apixaban, paring dabigatran to dose-adjusted war- sults are less applicable to patients with rivaroxaban, or warfarin for NVAF or farin for preventing 6-month recurrence a BMI of >40 kg/m2. VTE. In total, 366 patients were pre- of symptomatic VTE or related death.1 The safety and efficacy data on scribed anticoagulation for VTE treat- Dabigatran was noninferior to warfarin dabigatran use in obese patients ment. In this group, 152 patients (41.5%) for VTE prophylaxis (rate of VTE re- is sparse. While the RE-COVER, were anticoagulated with rivaroxaban, currence and/or death, 2.7% vs 2.5%; RE-COVER II, and RE-MEDY trials in- 47 (12.8%) with apixaban, and 167 hazard ratio [HR], 1.05, with a 95% concluded more obese patients than later (45.7%) with warfarin. Ultimately, fidence interval [CI] of 0.65-1.70]), with landmark trials, they were still under- there was no difference in rates of re- similar rates of major bleeding in the powered to truly evaluate the safety and current VTE and cardioembolic stroke dabigratran and warfarin groups (1.6% efficacy of dabigatran in this popula- between patients anticoagulated with and 1.9%, respectively; HR, 0.82 [95% tion. None of the studies reported out- rivaroxaban, apixaban, or warfarin (2% CI, 0.45-1.48]). The mean (SD) BMI and comes in patients with a BMI of ≥40 kg/ vs 2.1% vs 1.2% [P = 0.74] and 2.3% vs weight in the dabigatran group were 28.9 m2; therefore, warfarin remains the an- 1% vs 1.3% [P = 0.71], respectively). (5.7) kg/m2 and 85.5 (19.2) kg, respec- ticoagulant of choice in these patients. A subgroup analysis of all patients with tively. In total, 502 study patients (20%) Rivaroxaban. Based on the results of a BMI of ≥50 kg/m2 was also completed, weighed more than 100 kg, 538 (21%) the EINSTEIN and EINSTEIN-PE trials, and there were no VTEs reported in the had a BMI of 30.0-34.9 kg/m2, and 306 rivaroxaban was approved by FDA for DOAC group. Finally, rates of bleeding (12%) had a BMI of ≥35 kg/m2. The study the treatment of acute DVT and PE in were not significantly different between was underpowered to detect differences 2012.3,5 The first phase of the EINSTEIN groups; however, in the NVAF cohort between BMI groups. The RE-COVER II trial compared rivaroxaban to war- DOAC use was associated with signifi- trial aimed to confirm the results of the farin or low-molecular-weight heparin cantly lower rate of major bleeds than RE-COVER trial and (given the low event (LMWH) in an open-label, random- warfarin use (2.9% vs 7.9%, P = 0.02). rate in the first trial) perform a pooled ized, event-driven, noninferiority trial While the sample size in this study analysis.24 No significant safety or effi- involving patients with acute VTE. The was small, the data are encouraging in cacy differences were found between the continuation phase of the EINSTEIN that they indicate that rivaroxaban and dabigatran and warfarin groups. Within trial included patients who completed apixaban are as effective as warfarin for the dabigatran analysis, 438 patients 6 to 12 months of treatment for acute VTE and cardioembolic stroke preven- (34.2%) weighed 100 kg or more, 544 VTE. In these 2 studies, 245 of 1,731 tion in an obese patient population and (43%) had a BMI of 30.0-34.9 kg/m2, and (14.2%) and 85 of 602 (14.1%) patients have a more favorable safety profile.22 302 (23.6%) had a BMI of ≥35 kg/m2. The in the respective rivaroxaban groups Finally, Talamo and colleagues23 RE-COVER and RE-COVER II trials in- weighed greater than 100 kg.3 The in- published a retrospective analysis of cluded more obese patients than many vestigators did not complete any ana- patients of low (<60 kg, n = 27), normal of the landmark DOAC studies, but they lyses evaluating the safety and efficacy of (61-120 kg, n = 26), or high body were not powered to detect outcome dif- rivaroxaban in patients weighing more weight (>120 kg, n = 26) taking either ferences in high-body-weight patients. than 100 kg. Similarly, the EINSTEIN-PE rivaroxaban or apixaban for any indi- The RE-MEDY trial was designed to trial was a randomized, open-label, cation and found no significant dif- test the efficacy of dabigatran vs war- event-driven, noninferiority trial that ference in the rate of major bleeding farin for long-term VTE prophylaxis.25 compared rivaroxaban against standard between reference-weight and high- This trial confirmed the noninferiority therapy in patients with symptomatic body-weight patients (11.54% vs 0%, of dabigatran to warfarin and found PE.5 In the rivaroxaban group, 345 of P = 0.0744), although no bleeding that dabigatran had a better safety pro- 2,419 patients (14.3%) weighed more events were noted in the reference- file. A subgroup analysis conducted in than 100 kg, but no analyses of high- weight population.23 patients weighing 100 kg or more found body-weight patients were conducted. DOAC use for VTE treatment no significant between-group differ- Ultimately, no conclusions can be drawn in overweight adults. Table 1 dis- ence in the rate of recurrent VTE (2.6% from the EINSTEIN or EINSTEIN-PE plays comparative data from clinical vs 1%, P = 0.55); however, the study was trials regarding the utility of rivaroxaban trials that evaluated the effectiveness of underpowered to detect a difference in obese patients. DOAC use for the treatment of VTE in between the weight subgroups, as there Kubitza and colleagues26 com- high-body-weight patients. were only 299 patients (21%) in the pleted a single-dose pharmacokinetic

AM J HEALTH-SYST PHARM | VOLUME 77 | NUMBER 11 | JUNE 1, 2020 867 Clinical Review DIRECT-ACTING ORAL ANTICOAGULANTS

2 Downloaded from https://academic.oup.com/ajhp/article-abstract/77/11/865/5840388 by Biblioteca Nacional de Salud y Seguridad social user on 09 July 2020 (2.9% (rate of 2 (3.5% 2 2 (2.6% vs 1.8%, 2 Results = 0.48 for interaction) = 0.48 for interaction), and patients with BMI of ≥35 kg/m 2 2 = 0.99 for interaction), patients with BMI of 30.0-34.9 kg/m (3.3% vs 2.2%, P Findings in obese patient population: no significant difference in rate Findings in obese patient population: no significant difference (2.6% vs 1%, P = 0.555 for of VTE between groups of recurrence interaction) No difference in rate of VTE recurrence (2% vs 1.2%, 2.1%, in rate of VTE recurrence No difference in rate of major bleeding (1.3% vs 2.4% P = 0.45); no difference of VTE in patients with BMI 2.1%, P = 0.77); no recurrence >50 kg/m (2.1% vs 1.2% 2%, in rate of VTE recurrence No difference in rate of major bleeding (2.1% vs 2.4% P = 0.45); no difference of VTE in patients with BMI 1.3%, P = 0.77); no recurrence >50 kg/m Findings in obese patient population: no significant difference in rate Findings in obese patient population: no significant difference of primary outcome in patients ≥100 kg (4.4% vs 3%, P = 0.76 for interaction), patients with BMI of 30.0-34.9 kg/m Findings in obese patient population: patients with weight ≥100 kg, VTE, (rate of recurrent apixaban noninferior for acute VTE treatment 2.2% vs 3.5%; P = 0.43) and patients with BMI of >35 kg/m (1.7% vs 1.1%, P = 0.76); no in rate of VTE recurrence No difference in overall bleeding rate (8.3% vs 12%, P = 0.23) difference Less major bleeding with apixaban (0.6% vs 4.3%, P = 0.02) Findings in obese patient population: no significant difference in rate Findings in obese patient population: no significant difference of primary outcome in patients with BMI ≥100 kg (4.1% vs 3.6%, P recurrent VTE, 2% vs 3.56%; P = 0.06) recurrent vs 2.8%, P P = 0.89 for interaction), and patients with BMI of ≥35 kg/m vs 2.3%, P = 0.89 for interaction)

• • • • • • • • Intervention VKA vs apixaban rivaroxaban plus VKA Dabigatran vs VKA Rivaroxaban vs Rivaroxaban Apixaban vs VKA Dabigatran vs VKA Apixaban vs LMWH Apixaban vs VKA Dabigatran vs VKA Studies of Apixaban Study of Rivaroxaban Studies of Dabigatran 299 (21) 47 (100) 502 (20) 538 (21) 306 (12) 58 (100) 438 (34) 544 (43) 302 (24) 30 (19.7) 10 (21.3) 152 (100) 349 (13) 522 (19.4) No. (% of Total) Obese Patients, 2 2

2 2 2 2 2 2 2 2 Weight Weight Categories 30.0-34.9 kg/m 30.0-34.9 kg/m ≥100 kg BMI of ≥100 kg BMI of ≥100 kg BMI of ≥40 kg/m ≥100 kg BMI of >35 kg/m BMI of ≥30 kg/m BMI of ≥40 kg/m BMI of ≥50 kg/m BMI of ≥50 kg/m BMI of ≥35 kg/m BMI of ≥35 kg/m double-blind, noninferiority double-blind, noninferiority randomized cohort double-blind, prospective cohort cohort Study Design Randomized, Randomized, Double-blind, Retrospective, Randomized, Retrospective, Retrospective,

1 25 24

22 22

28 a a a a a (2018) (2018) (2017) (2014) (2013) (2009) (2013) Study was underpowered to detect a difference in patients at weight extremes. to detect a difference Study was underpowered Schulman et al Choi et al Kushnir et al Summary of Clinical Trials of DOAC Use for VTE Treatment in Overweight Patients of DOAC Use for VTE Treatment 1. Summary of Clinical Trials Table Reference heparin; VKA, vitamin K antagonist; VTE, venous thromboembolism. oral coagulant; LMWH, low-molecular-weight nonmajor; DOAC, direct-acting CRNM, clinically relevant Abbreviations: a Kushnir et al Agnelli et al Schulman et al Schulman et al

868 aM J HEALTH-SYST PHARM | VOLUME 77 | NUMBER 11 | JUNE 1, 2020 DIRECT-ACTING ORAL ANTICOAGULANTS Clinical Review crossover study in 48 patients, 12 of The investigators completed a subgroup with a BMI of ≥40 kg/m2 who were whom weighed more than 120 kg, analysis of patients by weight and BMI, anticoagulated with warfarin or apixaban with an average BMI of 43.5 kg/m2. and the noninferior effectiveness of for NVAF or VTE. In total, 58 and 88 pa-

Rivaroxaban was dosed at 10 mg daily, apixaban was maintained in these sub- tients with VTE were anticoagulated with Downloaded from https://academic.oup.com/ajhp/article-abstract/77/11/865/5840388 by Biblioteca Nacional de Salud y Seguridad social user on 09 July 2020 and plasma and urine drug concentra- groups; however, the study was under- apixaban and warfarin, respectively. The tions were used to calculate area under powered for this analysis. investigators found no significant differ- the curve (AUC). Pharmacodynamic Upreti and colleagues14 published ences in rates of VTE recurrence (1.7% vs efficacy was measured by measuring results of an open-label, single-dose, 1.1%, P = 0.76) or overall bleeding (8.3% anti–factor Xa activity. The researchers parallel-group pharmacokinetic study vs 12%, P = 0.23); however, the study was found no significant difference in to assess the impact of body weight underpowered to detect these differ- rivaroxaban maximum concentration extremes on the pharmacokinetics ences given the estimated rates of embo-

(Cmax) in patients weighing more than and pharmacodynamics of apixaban. lism and bleeding. The researchers noted 120 kg compared to patients weighing Nineteen patients were enrolled in a significant reduction in major bleeding 70 to 80 kg (149 µg/L vs 143.4 µg/L); the high-body-weight group (weight in the apixaban group vs the warfarin however, the maximum anti–factor Xa of ≥120 kg and BMI of ≥30 kg/m2). group (0.6% vs 4.3%, P = 0.02), which activity was slightly, although not sig- The researchers found 31% (90% CI, suggests that the favorable safety profile nificantly, reduced in the high-body- 18%-41%) and 23% (90% CI, 9%-35%) of apixaban found in the AMPLIFY trial 26 weight group. This study had several lower mean Cmax and AUC values, re- may extend to obese patients. limitations worth noting. First, the dose spectively, in the high-body-weight High-quality evidence endorsing of rivaroxaban used was subtherapeutic group relative to the reference group. the efficacy of apixaban for the acute for the treatment of VTE, and the clinical Furthermore, the half-life of apixaban treatment of VTE in an obese patient utility of anti–factor Xa monitoring to was approximately 3 hours shorter population is sparse. Apixaban should determine the effectiveness of DOACs in the high-body-weight group (8.8 be considered in this population only is poorly understood. Additionally, the hours vs 12.0 hours), and the mean if patients cannot tolerate warfarin. single-dose nature of this study could volume of distribution was approxi- Apixaban appears to maintain a fa- not accurately account for the long- mately 24% higher (75.6 L vs 61.0 L). vorable safety profile and may be an term influence of renal clearance on Pharmacodynamically, the authors appealing anticoagulant in patients at rivaroxaban levels. measured anti–factor Xa activity to as- higher bleeding risk.7,28 The efficacy data for rivaroxaban in sess apixaban effectiveness and found Edoxaban. In 2013, Büller and col- the treatment of VTE in obese patients a trend toward lower anti–factor Xa leagues9 published results of the HOKUSAI is sparse. Given the heterogeneity of the activity in the high-body-weight group VTE noninferiority trial, in which 4,921 studies that included rivaroxaban, as 3 hours after dosing (mean activity, patients were randomly assigned to re- well as a lack of head-to-head studies 1.85 IU/mL vs 2.79 IU/mL) but no sig- ceive either the standard of care (war- of rivaroxaban vs warfarin in an obese nificant between-group difference 12 farin) or edoxaban (30-60 mg daily). Only patient population, it is difficult to hours after dosing (0.70 IU/mL vs 0.77 611 patients (14.8%) in the edoxaban argue that a recommendation to use IU/mL). Ultimately, the researchers group weighed more than 100 kg, and the rivaroxaban for anticoagulation in this concluded that the pharmacokinetics investigators did not evaluate the impact patient population is evidence based. and pharmacodynamics of apixaban of body weight on safety or efficacy out- Furthermore, the available pharmaco- are modestly impacted by high body comes.9 There have been no pharmacokinetic data indicate mixed results.26,27 weight but that this impact is not clin- kinetic studies evaluating edoxaban in an With regard to safety, the available data ically significant. There were, however, obese population. Given the lack of data, suggest that rivaroxaban is comparable several limitations of this study. First, edoxaban should be avoided for VTE to warfarin in terms of bleeding events. the study’s single-dose nature did not treatment in obese patients. Rivaroxaban may be considered in pa- fully realize the impact of long-term DOACs for cardioembolic tients who have a BMI of ≥30 kg/m2 or a alterations in renal clearance that are stroke prophylaxis in overweight contraindication to warfarin. common in obese patients. Second, the adults with NAVF. Table 2 displays Apixaban. Apixaban was approved for relationship between anti–factor Xa ac- comparative data from clinical trials the treatment of acute VTE in 2012 after tivity and apixaban efficacy is not well that evaluated the efficacy of each the results of the AMPLIFY trial, which established. Finally, the patient pop- DOAC in the treatment of NVAF in demonstrated the noninferior efficacy ulation represented in this study was high-body-weight patients. of apixaban relative to LMWH plus war- younger, with a mean age of 29 years; Studies evaluating multiple DOACs. farin as well as its more favorable safety therefore, these results should not be Kido and colleagues29 published results profile.7 Only 19.4% of patients in the extrapolated to an older population. of a retrospective, single-center cohort apixaban group weighed 100 kg or more, Choi and colleagues28 completed study comparing the efficacy of DOACS and only 13% had a BMI of >30 kg/m2. a retrospective analysis of 390 patients with warfarin in obese patients with NVAF.

AM J HEALTH-SYST PHARM | VOLUME 77 | NUMBER 11 | JUNE 1, 2020 869 Clinical Review DIRECT-ACTING ORAL ANTICOAGULANTS Downloaded from https://academic.oup.com/ajhp/article-abstract/77/11/865/5840388 by Biblioteca Nacional de Salud y Seguridad social user on 09 July 2020

2 Results P = 0.31) P = 0.02) 1.24%, P = 0.66) P = 0.11) (4.03% vs 1.07% and apixaban groups vs rivaroxaban 0%) P = 0.71) and apixaban vs VKA (2.9% 2.9% 7.9%, rivaroxaban P = 0.06) P = 0.71) vs VKA (2.9% 2.9% apixaban and with rivaroxaban 7.9%, P = 0.06) (0.8% vs 0.87% 0.94%, P = 0.48 for dabigatran 110 mg, P = 0.42 for dabigatran 150 mg) lism (0.97% vs 1.28%, 95% CI, 0.55-1.05) systemic embolism, MI, and death outcome of stroke, (3.78% vs 4.51%; HR, 0.84 [95% CI, 0.71-0.99]) 2.51%; HR, 0.84 [95% CI, 0.67-1.07]) No difference in incidence of stroke (0.8% vs 2.4%, in incidence of stroke No difference in rate of bleeding (8.3% vs 12%, P = 0.23) No difference Lower rate of major bleeding with apixaban (0.6% vs 4.3%, No difference in incidence of stroke or TIA (1.75% vs in incidence of stroke No difference in rate of major bleeding (2.18% vs 4.97%, No difference Numerically higher incidence of TIA in the dabigatran group (2.3% vs 1% 1.3%, in rate of stroke No difference in rate of major bleeding with reduction toward Trend No difference in rate of stroke (1% vs 1.3% 2.3%, in rate of stroke No difference in rate of major bleeding with reduction toward Trend No difference in incidence of stroke or systemic embolism in incidence of stroke No difference No significant difference in rate of stroke or systemic embo - in rate of stroke No significant difference of composite reduced occurrence Apixaban significantly in rate of major bleeding (2.21% vs No significant difference Results in patients with BMI of ≥30 kg/m • • • • • • • • • • Findings in obese population: • • • • Intervention rivaroxaban, n = 25; rivaroxaban, apixaban, n = 19) vs VKA apixaban ( n = 103) vs VKA ( n = 152) = 152) vs rivaroxaban ( n = 152) vs rivaroxaban ( n = 174) dabigatran 150 mg vs VKA DOACs (dabigatran, n = 20; ( n = 174), Rivaroxaban Apixaban vs VKA Apixaban vs VKA Apixaban ( n = 103) vs VKA Dabigatran 110 mg vs Study of Dabigatran Studies of Apixaban Study of Rivaroxaban 37 (21) 19 (100) 174 (100) 103 (100) 128 (100) 124 (100) 7,159 (40) 3,099 (17) No. (% of Total) Obese Patients, Study Evaluating Multiple DOACs or

2 2 2 2 2 2 2 Weight Weight Categories weight >120 kg BMI of ≥30 kg/m BMI of ≥40 kg/m BMI of ≥40 kg/m BMI of ≥50 kg/m BMI of ≥50 kg/m BMI of >40 kg/m ≥100 kg BMI of ≥30 kg/m Study Design randomized, double-blind trial single-center cohort noninferiority Retrospective, cohort Retrospective, Post hoc analysis of cohort Retrospective, Retrospective, Retrospective, Randomized, cohort Retrospective, a a a (2009) (2016) (2018) (2018) 2 31 22 22 (2017) (2019) 28 29 Study was underpowered to detect a difference in patients at weight extremes. to detect a difference Study was underpowered Kido et al Summary of Studies of DOAC Use for Cardioembolic Stroke Prevention in Overweight Patients With Nonvalvular Atrial Fibrillation Prevention Stroke 2. Summary of Studies DOAC Use for Cardioembolic Table Reference TIA, transient ischemic attack; VKA, vitamin K antagonist. infarction; ratio; MI, myocardial oral coagulant; HR, hazard CI, confidence interval; DOAC, direct-acting Abbreviations: a Sandhu et al Kushnir et al Connolly et al Kushnir et al Choi et al

870 aM J HEALTH-SYST PHARM | VOLUME 77 | NUMBER 11 | JUNE 1, 2020 DIRECT-ACTING ORAL ANTICOAGULANTS Clinical Review

In total, 128 patients were included, 64 P = 0.42 for comparison of dabigatran treatment of NVAF in an obese patient (50%) of whom were anticoagulated with 150 mg with warfarin]); however, the population given low cardioembolic warfarin, 25 (39.1%) with rivaroxaban, 20 study was not powered to detect a signif- stroke rates and a tendency toward

(31.1 %) with dabigatran, and 19 (29.7%) icant difference in this subgroup. analysis of combined event rates in Downloaded from https://academic.oup.com/ajhp/article-abstract/77/11/865/5840388 by Biblioteca Nacional de Salud y Seguridad social user on 09 July 2020 with apixaban. All patients had a BMI of Reilly and colleagues30 completed DOAC trials. Additionally, the data for >40 kg/m2 or weighed more than 120 kg. a pharmacokinetic analysis of the the use of rivaroxaban in patients with There were no significant differences RE-LY trial to evaluate the association a BMI of >30 kg/m2 is even more sparse. in rates of stroke and/or transient is- between dabigatran serum concen- Rivaroxaban may be considered in pa- chemic attack (TIA) and major bleeding trations and safety and efficacy out- tients with a BMI of 30.0 to 34.9 kg/m2 between the DOAC and warfarin groups comes. The investigators compared who are poor candidates for warfarin (1.75% vs 2.07% [P = 0.8] and 2.18% vs drug levels in patients weighing less therapy. 4.97% [P = 0.11], respectively); however, than 50 kg, 50 to 99.9 kg, or 100 kg or Apixaban. Results of the ARISTOTLE dabigatran-treated patients had a nu- more and found peak concentrations trial prompted FDA approval of merically higher annual rate of TIA than in these weight groups of 2.63, 1.94, and apixaban for NVAF in 2013 based on su- patients who received rivaroxaban or 1.56 ng/mL/mg, respectively. Median perior efficacy results and a significant apixaban (4.03% vs 1.07% vs 0%). Of note, trough concentrations were 1.01, 0.84, reduction in bleeding events.6 A 2015 this study was underpowered to detect and 0.66 ng/mL/mg, respectively. The post hoc analysis of the trial, which insignificant between-group differences utility of drug level monitoring and cluded 7,159 patients (40%) with a BMI due to small sample size. Globally, this correlation of drug levels to clinical of ≥30 kg/m2, aimed to explore the im- study outlined the utility of DOACs as a outcomes of DOAC use is unclear; how- pact of adiposity on stroke incidence in reasonable treatment modality for NVAF; ever, there is certainly a trend of lower patients with NVAF.31 When comparing however, it did little to differentiate be- concentrations in high-body-weight apixaban to warfarin among patients tween DOACs. patients. Reilly et al noted an inverse re- with a BMI of ≥30 kg/m2, the authors Dabigatran. Based on the results of lationship between dabigatran trough found no significant difference in rates the RE-LY trial, dabigatran was approved levels and the probability of a sympto- of stroke or systemic embolism (0.97% by FDA for stroke prevention in NVAF in matic embolic event (C statistic, 0.657; vs 1.28%; HR, 0.76 [95% CI, 0.55-1.05]) 2010. The RE-LY trial was a randomized 95% CI, 0.61-0.71). The results of this but did find a significant difference in noninferiority trial comparing blinded pharmacokinetic analysis raise ques- the composite outcome of stroke, sys- dabigatran use (110 or 150 mg by mouth tions regarding the impact of weight on temic embolism, MI, and death (3.78% twice daily) vs open-label warfarin use, dabigatran serum levels and the drug’s vs 4.51%; HR, 0.84 [95% CI, 0.71-0.99]). and the primary outcome of interest effectiveness. There was no significant difference in was the rate of stroke or systemic em- Given the unfavorable pharmaco- rates of major bleeding between groups bolism.2 Both doses of dabigatran were kinetic profile of dabigatran in obese (2.12% vs 2.51%; HR, 0.84 [95% CI, 0.67- found to be noninferior to warfarin for patients as well as a lack of strong clin- 1.07]). When evaluating the impact of the primary outcome (1.53% vs 1.11% vs ical data, dabigatran should likely be high waist circumference (>102 cm for 1.69%, P < 0.001), with the 150-mg dose avoided in this patient population. men, >88 cm for women) on outcomes, found superior to warfarin (relative risk Rivaroxaban. In 2011, based on re- the authors found no efficacy difference [RR], 0.66; 95% CI, 0.53-0.82; P < 0.001). sults of the ROCKET-AF trial, rivaroxaban between the 2 drugs (rate of recurrent Low-dose dabigatran was demon- received FDA approval for prevention of VTE, 1.29% vs 1.22%; 95% CI, 0.82-1.36), strated to be associated with a lower cardioembolic stroke in patients with but patients taking apixaban had a sig- rate of major bleeding (2.71% vs 3.36%, NVAF.4 This trial was a double-blind, ran- nificantly lower rate of major bleeding P = 0.003); however, there was no signif- domized noninferiority trial comparing (2.14% vs 2.96%; 95% CI, 0.6-0.87). This icant difference in bleeding rates with rivaroxaban to warfarin in over 14,000 pa- study, in addition to the study by Choi use of high-dose dabigatran vs either tients. Safety and efficacy outcomes were and colleagues, suggested that apixaban low-dose dabigatran (3.11% vs 2.71%, reported for body weight and BMI sub- is at least as effective as warfarin at P = 0.052) or warfarin (3.11% vs 3.36%, groups. Only 2.46% and 2.88% of patients preventing stroke and systemic embo- P = 0.31). The RE-LY trial included 3,099 in the rivaroxaban group weighed more lism and has a favorable safety profile.28,31 patients (17%) who weighed 100 kg or than 90 kg and had a BMI of >35 kg/m2, As with the data for high-body- more. There were no significant differ- respectively. There were no differences weight patients with VTE, the evidence ences in rates of stroke or systemic em- noted in safety or thromboembolic out- for apixaban for NVAF in high-body- bolism with use of low-dose dabigatran, comes in obese patients; however, the weight patients is sparse and heter- high-dose dabigatran, or warfarin in the study was underpowered to detect a dif- ogeneous. However, it appears that higher-body-weight population (0.8% ference within these subgroups. apixaban maintains a favorable safety vs 0.87% vs 0.94% [P = 0.49 for compar- It is difficult to draw strong conclu- and efficacy profile when compared ison of dabigatran 110 mg with warfarin; sions on the use of rivaroxaban in the to warfarin. Like the other DOACs,

AM J HEALTH-SYST PHARM | VOLUME 77 | NUMBER 11 | JUNE 1, 2020 871 Clinical Review DIRECT-ACTING ORAL ANTICOAGULANTS the evidence supporting apixaban use in patients with a BMI of >35 kg/m2 is sparse, and warfarin should remain the preferred anticoagulant in that Downloaded from https://academic.oup.com/ajhp/article-abstract/77/11/865/5840388 by Biblioteca Nacional de Salud y Seguridad social user on 09 July 2020 population. Edoxaban. In 2015, FDA approved edoxaban for the prevention of stroke and systemic embolization in patients with NVAF based on the results of the ENGAGE-AF TIMI 48 trial.8 Edoxaban was noninferior to warfarin for the prevention of stroke or systemic em- Results bolism; however, no analysis in overweight patients was conducted, and outcomes were not stratified by body weight. The majority of studies that compared DOACs to warfarin for NVAF either did not include patients taking edoxaban or had such small Findings in fragile patients: lower rate of major bleeding with (1.3% vs 4.5%; HR, 0.27 [95% CI, 0.13-0.54]) rivaroxaban numbers of edoxaban-prescribed pa- Findings in underweight population: no VTE events occurred in either group Findings in underweight population: 1 VTE event warfarin group Findings in underweight population: no VTE events recorded in either group Findings in underweight population: patients with weight ≤60 kg, apixaban noninferior to VKA for acute VTE treatment VTE, 2.7 % vs 4.3%; P = 0.43) (rate of recurrent • tients that little can be inferred from • • • • the results.21,23,29 Given the paucity of data regarding the safety and efficacy of edoxaban in obese patients with NVAF, an alternative DOAC should be vs LMWH plus VKA vs VKA vs VKA vs VKA vs LMWH plus VKA Intervention Rivaroxaban Rivaroxaban selected in this population. Dabigatran Dabigatran Dabigatran Apixaban DOACS for VTE treatment in underweight adults. Table 3 displays comparative data from clinical trials that evaluated the efficacy of

dabigatran for the treatment of VTE in 18 (0.7) 57 (2.2) 10 (0.7) 107 (7) 225 (9) No. (% of Total) low-body-weight patients. Obese Patients, Study of Apixaban Study of Rivaroxaban Dabigatran. The RE-COVER, Studies of Dabigatran RE-COVER II, and RE-MEDY studies included very few low-body-weight patients and did not provide significant in- ≤50 kg ≤60 kg <50 kg <50 kg <50 kg sight into the utility of dabigatran for the Weight treatment of acute VTE in underweight Categories patients.1,24,25 Additionally, there are currently no pharmacokinetic studies evaluating dabigatran use in low-body weight patients to guide clinician de- cision making. Dabigatran should be

avoided in low-body-weight patients Study Design with VTE disease unless significant from EINSTEIN-DVT and from EINSTEIN-PE trials contraindications to warfarin exist. noninferiority noninferiority prospective Pooled analysis of data Rivaroxaban. Both the EINSTEIN Randomized, double-blind, Randomized, double-blind, Double-blind, randomized Randomized, double-blind, a a and EINSTEIN-PE trials included low- a

body-weight patients; however, the per- a (2014) (2013) centage of underweight patients (≤50 kg) (2009) 1 24 25

included was low in both trials (1.6% and (2013) 7 (2013) 2.1%, respectively).3,5 Safety and efficacy 27 outcomes were not evaluated in relation to weight in these studies, so minimal Study was underpowered to detect a difference in patients at weight extremes. to detect a difference Study was underpowered Agnelli et al Summary of Studies of DOAC Use for VTE Treatment in Underweight Patients 3. Summary of Studies DOAC Use for VTE Treatment Table Reference heparin; VKA, vitamin K antagonist; VTE, venous thromboembolism. ratio; LMWH, low-molecular-weight oral coagulant; HR, hazard CI, confidence interval; DOAC, direct-acting Abbreviations: a Schulman et al Schulman et al Prins et al conclusions can be drawn. Schulman et al

872 aM J HEALTH-SYST PHARM | VOLUME 77 | NUMBER 11 | JUNE 1, 2020 DIRECT-ACTING ORAL ANTICOAGULANTS Clinical Review

Prins and colleagues27 performed a underweight patients had a 14% lower who weighed less than 50 kg but was pooled data analysis of the EINSTEIN volume of distribution than heavier pa- not powered to detect safety or efficacy trials that included fragile patients tients (52.7 L vs 61 L), and the apixaban differences within a low-body-weight

(age of >75 years, creatinine clearance half-life was 4 hours longer (15.8 hour subgroup.2 The previously discussed Downloaded from https://academic.oup.com/ajhp/article-abstract/77/11/865/5840388 by Biblioteca Nacional de Salud y Seguridad social user on 09 July 2020 of <50 mL/min, or weight of ≤50 kg). vs 12 hours). These pharmacokinetic pharmacokinetic analysis by Reilly and In total, 1,573 patients (19%) were differences did not translate to differ- colleagues30 noted 21% and 53% higher classified as fragile; however, the ma- ences in 12-hour postdose anti–factor geometric mean drug concentrations in jority (1,279, or 81%) were deemed Xa activity, but they do beg the question patients weighing less than 50 kg com- fragile secondary to advanced age. as to whether underweight patients are pared to patients weighing 50 to 100 kg There was no difference in rates of re- overanticoagulated with the standard and patients weighing 100 kg or more, current VTE in fragile patients treated 10-mg starting dose of apixaban. respectively. Those investigators also with rivaroxaban vs a LMWH plus Apixaban appears to be an effica- noted a positive correlation between warfarin, but there was a significantly cious option for anticoagulation in a drug concentrations and bleeding lower incidence of major bleeding with low-body-weight population; however, events. As such, dabigatran should be rivaroxaban (1.3% vs 4.5%; HR, 0.27 the safety profile is less well estab- used with caution in patients with a low [95% CI, 0.13–0.54]). These results are lished. There are studies demonstrating body weight given their higher risk of difficult to apply to patients of low body a higher risk of bleeding with DOAC bleeding relative to that of their normal- weight but suggest that rivaroxaban use in low-body-weight patients, and weight counterparts. may be a safer option in older patients. apixaban concentrations appear to be Rivaroxaban. The ROCKET-AF trial A study by Kubitza et al26 found higher in a low-body-weight popula- included 93 patients (4.62% of the total significantly higher peak rivaroxaban tion. It is, however, well established population) and 72 patients (4.25% of plasma concentrations in low- that apixaban is safer than either the total) with a weight of ≤70 kg or a body-weight (≤50 kg) patients when rivaroxaban or dabigatran in a normal- BMI of ≤25 kg/m2, respectively; however, compared to standard-body-weight pa- body-weight population.14,33 Apixaban no specific analyses focused on low- tients (1.24 µg/L vs 1.00 µg/L, P = 0.04); use may be considered in low-body- body-weight patients were conducted.4 however, there was no significant dif- weight patients who cannot tolerate Additionally, pharmacokinetic data sug- ference in AUC, half-life, or anti–factor warfarin. gest insignificant changes in rivaroxaban Xa activity between groups. The clinical Edoxaban. The HOKUSAI VTE trial pharmacokinetics and pharmacody- implications of peak plasma concen- enrolled 524 patients (12.7% of the total namics in low-body-weight patients.26 trations of rivaroxaban are currently population) who weighed less than Rivaroxaban may be considered for unknown, making these data difficult 60 kg, but a safety or efficacy analysis NVAF in low-body-weight patients. to utilize in a clinical setting. within weight categories was not com- Apixaban. The apixaban package Ultimately, there is evidence sug- pleted.8 The dose of edoxaban for pa- insert provides a recommendation for gesting that rivaroxaban may be a safer tients with a weight of <60 kg was halved a 2.5-mg reduction of the dose (given alternative to warfarin in the treatment in the study secondary to prior know- by mouth twice daily) in patients with of VTE in low-body-weight patients; ledge of higher bleeding risk in patients 2 or more of the following risk factors: however, there are currently no com- with a low body weight.34 This reduc- body weight of ≤60 kg, age of ≥80 years, parative data on the safety and efficacy tion in edoxaban dose is reflected in the and serum creatinine concentration of rivaroxaban and other DOACs in this drug’s package insert, as 30 mg daily is of ≥1.5 mg/dL. This recommendation patient population. recommended for patients with a weight is based on documented pharmacoki- Apixaban. While there is a partially of ≤60 kg (after 5 days of parenteral anti- netic changes as well as an increased weight-based recommendation for coagulant therapy).35 Based solely on the risk of bleeding in such patients; how- apixaban in NVAF, there is less guidance HOKUSAI VTE trial results, it is difficult ever, the impact of body weight alone on on use of the drug in VTE treatment, and to draw strong conclusions on the utility apixaban safety is unclear.32,33 the available evidence on this topic is of edoxaban for VTE treatment in low- In the ARISTOTLE trial, 1,985 of sparse.32 body-weight patients. patients (11%) in the apixaban group The AMPLIFY trial included 225 DOAC use for cardioembolic weighed 60 kg or less, and improved patients (9% of the total) with a body stroke prophylaxis in under- safety and efficacy outcomes observed weight of ≤60 kg and was not powered weight adults. Table 4 displays with use of apixaban vs VKA persisted to detect safety or efficacy differences comparative data from clinical trials in this group.6 However, it is unclear in this population.7 The study by Upreti evaluating the efficacy of DOACs for the how many of these patients were on and colleagues14 demonstrated 27% and treatment of NVAF in low-body-weight reduced-dose apixaban. Alexander 36 20% higher mean apixaban Cmax and patients. and colleagues published a secondary AUC values, respectively, in patients Dabigatran. The RE-LY trial included analysis of the ARISTOTLE trial to eval- weighing 50 kg or less. Additionally, 376 patients (2% of the total population) uate the impact of each dose-reduction

AM J HEALTH-SYST PHARM | VOLUME 77 | NUMBER 11 | JUNE 1, 2020 873 Clinical Review DIRECT-ACTING ORAL ANTICOAGULANTS

criteria on bleeding risk. Of the 13,356 patients in the ARISTOTLE trial who met only 1 dose-reduction criteria,

1,426 (36%) weighed 60 kg or less. There Downloaded from https://academic.oup.com/ajhp/article-abstract/77/11/865/5840388 by Biblioteca Nacional de Salud y Seguridad social user on 09 July 2020 was a significantly lower rate of major bleeding events in the apixaban group compared to the warfarin group (2.3% vs 4.0%; HR, 0.69 [95% CI, 0.4-0.9]) in patients with a weight of ≤60 kg. The researchers concluded that the 5-mg dose of apixaban maintained the ef-

Results ficacy and safety benefit in patients with isolated advanced age, low body weight, or renal dysfunction. While the safety data on use of apixaban in low-body-weight patients appears promising, several studies have found an increased risk of bleeding with DOAC use in low-body- Findings in underweight population: no difference in incidence Findings in underweight population: no difference or systemic embolism (2.6% vs 2.2% 5.0%, of stroke P = 0.48 for dabigatran 110 mg, = 0.42 150 mg) or systemic embolism in incidence of stroke No difference (2.0% vs 3.2%, P = 0.26) Lower rate of major bleeding with apixaban (2.3% vs 4.0%; HR, 0.69 [0.4-0.9]) • Findings in underweight population: • • weight patients compared to standard- weight patients.23,34,37 Notably, these studies did not compare individual DOACs to warfarin but rather combined all DOACs for analysis. As such, it is difficult to form any conclusions

Intervention regarding apixaban specifically from these studies. Ultimately, apixaban ap- dabigatran 150 mg vs VKA Dabigatran 110 mg vs Apixaban vs VKA Apixaban vs VKA pears to be both safe and efficacious for patients with low body weights. A dose reduction based on low body weight alone is not warranted. Study of Dabigatran Studies of Apixaban Edoxaban. The ENGAGE-AF TIMI 376 (2) 1,985 (11) 1,426 (36) 48 trial included 2,083 patients who No. (% of Total) Obese Patients, weighed 60 kg or less. Of these patients, 684 (9.7%) received edoxaban 60 mg daily and 698 (9.9%) received edoxaban 30 mg daily.8 Edoxaban was deemed ≤60 kg ≤60 kg <50 kg Weight Weight noninferior to warfarin for the preven- Category tion of stroke or systemic embolism; however, the lower edoxaban dose was associated with a significantly higher incidence of ischemic stroke (HR for comparison with 60-mg dose, 1.41; 95% CI, 1.19-1.67; P < 0.001), which is reflected 35

Study Design in the edoxaban labeling. No analyses noninferiority double-blind of a randomized, double-blind trial focused specifically on patients with low Randomized, Randomized, Secondary analysis body weight were conducted; therefore, if edoxaban is selected for stroke pre- a a vention in NVAF in a low-body-weight (2016)

36 patient, a dosage of 60 mg daily should (2009) (2011) 2 6 be used.

Discussion A lack of high-quality, random- Study was underpowered to detect a difference in patients at weight extremes. to detect a difference Study was underpowered Alexander et al Granger et al Connolly et al Summary of Studies of DOAC Use for Cardioembolic Stroke Prevention in Underweight Patients With Nonvalvular Atrial Fibrillation Prevention Stroke 4. Summary of Studies DOAC Use for Cardioembolic Table Reference Abbreviations: DOAC, direct-acting oral coagulant; HR, hazard ratio; VKA, vitamin K antagonist. oral coagulant; HR, hazard DOAC, direct-acting Abbreviations: a ized trials evaluating optimal DOAC

874 aM J HEALTH-SYST PHARM | VOLUME 77 | NUMBER 11 | JUNE 1, 2020 DIRECT-ACTING ORAL ANTICOAGULANTS Clinical Review selection and dosing in high- and low- 8. Giugliano RP, Ruff CT, Braunwald E, underweight_adult_15_16.htm. body-weight patients makes optimal et al. Edoxaban versus warfarin in pa- Accessed January 2, 2019. drug selection difficult. Dabigatran tients with atrial fibrillation. N Engl J 21. Tittl L, Endig S, Marten S, et al. Impact Med. 2013;369:2093-2104. of BMI on clinical outcomes of NOAC and edoxaban should likely be avoided 9. Büller HR, Decousus H, Grosso MA, therapy in daily care—results of the Downloaded from https://academic.oup.com/ajhp/article-abstract/77/11/865/5840388 by Biblioteca Nacional de Salud y Seguridad social user on 09 July 2020 on the basis of negative pharmacoki- et al. Edoxaban versus warfarin for prospective Dresden NOAC Registry. Int netic data and minimal efficacy data. the treatment of symptomatic venous J Cardiol. 2018;262:85-91. In the context of VTE and NVAF, safety thromboembolism. N Engl J Med. 22. Kushnir M, Choi Y, Eisenberg R, et al. and efficacy data support for use of 2013;369:1406-1415. Efficacy and safety of direct oral factor 10. Kearon C, Akl EA, Ornelas J, et al. Xa inhibitors in 795 morbidly obese pa- rivaroxaban is comparable to that for Antithrombotic therapy for VTE disease. tients. Blood. 2018;132:423. warfarin, while apixaban has a more fa- Chest. 2016;149(2):315-352. 23. Talamo L, Maitland HS, Palkimas S, vorable safety profile. Ultimately, if war- 11. Lip GY, Banerjee A, Boriani G, et al. et al. Safety of DOACS in patients of ex- farin cannot be used, both rivaroxaban Antithrombotic therapy for atrial fibril- treme weight. Blood. 2016;128:1450. and apixaban appear to be reasonable lation. Chest. 2015;154(5):1121-1201. 24. Schulman S, Kakkar AK, Goldhaber SZ, 12. Martin K, Beyer-Westendorf J, et al. Treatment of acute venous throm- options for high-body-weight patients, Davidson BL, et al. Use of the direct boembolism with dabigatran or war- and apixaban may be the most ap- oral anticoagulants in obese patients: farin and pooled analysis. Circulation. propriate DOAC for low-body-weight guidance from the SSC of the ISTH. J 2014;129:764-772. patients. Thromb Haemost. 2016;14:1308-1313. 25. Schulman S, Kearon C, Kakkar AK, et al. 13. Baglin T, Hillarp A, Tripod A, et al. Extended use of dabigatran, warfarin, or Conclusion Measuring oral direct inhibitors of placebo in venous thromboembolism. thrombin and factor Xa: a recom- N Engl J Med. 2013;368:709-718. While the evidence on DOAC use mendation from the Subcommittee 26. Kubitza D, Becka M, Zuehlsdorf M, in patients at extremes of body weight on Control of Anticoagulation of Mueck W. Body weight has limited is sparse, apixaban and rivaroxaban ap- the Scientifc and Standardization influence on the safety, tolerability, pear to have the most favorable safety Committee of the International Society pharmacokinetics, or pharmacody- on Thrombosis and Haemostasis. J namics of rivaroxaban (BSY 59–7939) and efficacy profiles. Edoxaban and Thromb Haemost. 2013;11:756-760. in healthy subjects. J Clin Pharmacol. dabigatran should be avoided. 14. Upreti VV, Wang J, Barrett YC, et al. 2007;47:218-226. Effect of extremes of body weight on the 27. Prins MH, Lensing AW, Bauersachs R, Disclosures pharmacokinetics, pharmacodynamics, et al. Oral rivaroxaban versus standard safety, and tolerability of apixaban in therapy for the treatment of venous The authors have declared no potential con- healthy subjects. Br J Clin Pharmacol. thromboembolism: a pooled analysis of flicts of interest. 2013;76(6):908-916. the EINSTEIN-DVT and PE randomized 15. Piran S, Traquair H, Chan N, et al. studies. Thromb J. 2013;11:21. References Peak plasma concentration of direct 28. Choi Y, Kushnir M, Billett HH. Apixaban 1. Schulman S, Kearon C, Kakkar AK, et al. oral anticoagulants in obese patients is safe and effective in morbidly obese Dabigatran versus warfarin in the treat- weighing over 120 kilograms: a ret- patients: a retrospective analysis of 390 pa- ment of acute venous thromboembo- rospective study. Res Pract Thromb tients with BMI ≥ 40. Blood. 2017;130:1105. lism. N Engl J Med. 2009;361:2342-2352. Haemost. 2018;2:684-688. 29. Kido K, Ngorsuraches S. Comparing 2. Connolly SJ, Ezekoqitz MD, Yusuf S, 16. Centers for Disease Control and the efficacy and safety of direct oral et al. Dabigatran versus warfarin in Prevention. Adult obesity facts. https:// anticoagulants with warfarin in the patients with atrial fibrillation. N Engl J www.cdc.gov/obesity/data/adult.html. morbidly obese population with Med. 2009;361:1139-1151. Accessed January 2, 2019. atrial fibrillation. Ann Pharmacother. 3. Bauersachs R, Berkowitz SD, Brenner B, 17. Lavie CJ, Milani RV, Ventura HO. 2019;53(2):165-170. et al. Oral rivaroxaban for symptomatic Obesity and cardiovascular disease: 30. Reilly PA, Lehn T, Haertter S, et al. The venous thromboembolism. N Engl J risk factor, paradox, and impact effect of dabigatran plasma concentra- Med. 2010;363:2499-2510. of weight loss. J Am Coll Cardiol. tions and patient characteristics on the 4. Patel MR, Mahaffey KW, Garg J, 2009;53(21):1925-1932. frequency of ischemic stroke and major et al. Rivaroxaban versus warfarin in 18. Hanley MJ, Abernethy DR, bleeding in atrial fibrillation patients. J nonvalvular atrial fibrillation. N Engl J Greenblatt DJ. Effect of obesity on the Am Coll Cardiol. 2014;63:321-328. Med. 2011;365:883-891. pharmacokinetics of drugs in humans. 31. Sandhu RK, Ezekoqitz J, Andersson U, 5. Buller HR, Prins MH, Lensin AW, et al. Clin Pharmacokinet. 2010;49(2):71-87. et al. The ‘obesity paradox’ in atrial Oral rivaroxaban for the treatment of 19. Burnett AE, Mahan CE, Vasquez SR, fibrillation: observations from the symptomatic pulmonary embolism. N et al. Guidance for the practical man- ARISTOTLE (Apixaban for Reduction Engl J Med. 2012:366(14):1287-1297. agement of the direct oral anticoagu- in Stroke and Other Thromboembolic 6. Granger CB, Alexander JH, McMurray JJ, lants (DOACs) in VTE treatment. J Events in Atrial Fibrillation) trial. Eur et al. Apixaban versus warfarin in pa- Thromb Thrombolysis. 2016;41:206-232. Heart J. 2016;37(38):2869-2878. tients with atrial fibrillation. N Engl J 20. Centers for Disease Control and 32. Eliquis [package insert]. Princeton, NJ: Med. 2011;365:981-992. Prevention. Prevalence of underweight Bristol-Myers Squibb Company; 2014. 7. Agnelli G, Buller HR, Cohen A, et al. among adults aged 20 and over: United 33. Cohen AT, Hamilton M, Byrd A, et al. Oral apixaban for the treatment of acute States, 1960–1962 through 2015–2016. Comparison of the non-VKA oral anti- venous thromboembolism. N Engl J https://www.cdc.gov/nchs/data/ coagulants apixaban, dabigatran, and Med. 2013;369:799-808. hestat/underweight_adult_15_16/ rivaroxaban in the extended treatment

AM J HEALTH-SYST PHARM | VOLUME 77 | NUMBER 11 | JUNE 1, 2020 875 Clinical Review DIRECT-ACTING ORAL ANTICOAGULANTS

and prevention of venous thrombo- non-valvular atrial fibrillation. Circ J. high creatinine: a secondary analysis embolism: systematic review and 2012;76:1840-1847. of a randomized clinical trial. JAMA network meta-analysis. PLoS ONE. 35. Savaysa [package insert]. Parsippany, Cardiol. 2016;1(6):673-681. 2016;11(8):e0130064. doi:10.1371/ NJ: Daiichi Sankyo, Inc.; 2015. 37. Park CS, Choi E, Kim H, et al. Increased journal.pone.0160064. 36. Alexander JH, Andersson UA, risk of major bleeding in underweight Downloaded from https://academic.oup.com/ajhp/article-abstract/77/11/865/5840388 by Biblioteca Nacional de Salud y Seguridad social user on 09 July 2020 34. Yamashita T, Koretsune Y, Yasaka M, Lopes RA, et al. Apixaban 5 mg patients with atrial fibrillation who were et al. Randomized, multicenter, twice daily and clinical outcomes in prescribed non-vitamin K antagonist warfarin-controlled phase II study of patients with atrial fibrillation and oral anticoagulants. Heart Rhythm. edoxaban in Japanese patients with advanced age, low body weight, or 2017;14:501-507.

876 aM J HEALTH-SYST PHARM | VOLUME 77 | NUMBER 11 | JUNE 1, 2020