Will Riva Be Defeated By It’s (Long-Standing) Nemesis, Apixa?

Author(s)
Amy D. Robertson, PharmD, BCACP
Michelle Balli, PharmD, BCACP

Reviewed By
James C. Lee, Pharm.D., BCACP
Jaini Patel, PharmD, BCACP

Citation
Fralick M, Colacci M, Schneeweiss S, et al. Effectiveness and safety of apixaban compared with rivaroxaban for patients with atrial fibrillation in routine practice: a cohort study. Ann Intern Med. 2020;172:463–473. [Epub ahead of print 10 March 2020]. doi: 10.7326/M19-2522.

The Problem

The use of DOACs for NVAF has rapidly increased due to their ease of use and favorable safety profile.¹ Compared to warfarin, DOACs have a rapid onset and offset, more predictable effects, fixed dosing, no special monitoring, fewer drug interactions, and no interactions with dietary vitamin K.^1,2 AHA/ACC/HRS and CHEST guidelines now recommend DOACs over warfarin for stroke prevention in NVAF, but do not state a preference for one DOAC over another.^3,4 Factor Xa inhibitors, specifically apixaban and rivaroxaban, are the most commonly prescribed DOACs suggesting they are preferred by clinicians in real world practice, but do we have enough evidence to determine which Xa inhibitor is best in class?¹

What’s Known

There have been several trials comparing the safety and efficacy of DOACs and warfarin in NVAF.^5,6 These studies demonstrated a consistently lower risk of hemorrhagic stroke and intracranial hemorrhage (ICH) regardless of DOAC. However, data is conflicting when it comes to other outcomes such as risk of ischemic stroke and gastrointestinal (GI) bleeding. ARISTOTLE revealed a significantly reduced risk of stroke or systemic embolism and major bleeding with apixaban compared to warfarin.⁵ ROCKET-AF showed rivaroxaban was non-inferior to warfarin for reducing the risk of stroke or systemic embolism and had no difference in preventing major bleeding.⁶ While rivaroxaban showed a significantly decreased risk of ICH it was also associated with a higher risk for GI bleeding. Unfortunately, the findings observed in clinical trials often do not reflect the outcomes from the use of new drugs in real-world, everyday practice.² Several observational studies have compared DOACs to warfarin in NVAF. One retrospective cohort analysis comparing DOACs to warfarin demonstrated better effectiveness and safety with apixaban and similar effectiveness and safety with rivaroxaban.²

Currently there are no randomized controlled trials directly comparing apixaban and rivaroxaban. One meta-analysis indirectly compared the efficacy and safety of DOACs in patients with NVAF using apixaban as the reference.⁷ The results (n=15 studies) revealed no significant difference in the risk of stroke between apixaban and other DOACs. There was more variability in major bleeding risk between DOACs. Overall, apixaban was associated with a lower risk of major bleeding in most studies when compared to rivaroxaban. There is a need for additional data to guide clinical decision making with DOACs in practice, and a randomized, prospective trial is currently underway.⁸

What’s New

While we await the results of prospectively conducted comparative efficacy studies, well-designed observational studies may provide us with some insights about a preferred DOAC.  In a retrospective cohort study, investigators compared the safety and efficacy of apixaban versus rivaroxaban in new users with NVAF. Patients ≥ 18 years old were included with an ICD-9 or ICD-10 diagnosis of atrial fibrillation or atrial flutter and a new prescription for apixaban 5 mg or rivaroxaban 20 mg filled in the preceding 180 days between December 28, 2012 and January 1, 2019. Data were obtained from a nationwide commercial insurance claims database and included demographics, health care utilization, diagnoses, laboratory values, and dispensing data from pharmacy claims. Exclusion criteria included less than 180 days of available data or a diagnosis of cancer, valvular heart disease, venous thromboembolism, or stage 5 chronic kidney disease with dialysis. Propensity score matching was used in a 1:1 fashion with consideration of baseline characteristics and to reduce confounders. Laboratory data were not included in the analysis as they were available in only one-third of included patients.

Efficacy was assessed by using a composite outcome of ischemic stroke or systemic embolism.  The primary safety outcome was a composite of ICH or GI bleeding.  The efficacy and safety outcomes were determined by the primary diagnosis codes associated with hospitalizations. Additional safety outcomes collected included incidence of hepatitis, vasculitis, heart failure, or other bleeding events. Patients were followed until the end of the study period or the development of a study outcome, medication change or discontinuation, insurance plan disenrollment, or death.

The final analysis with propensity score matching included 39,351 patients on apixaban and 39,351 patients on rivaroxaban. Baseline characteristics were similar between groups after score matching (Table 1). Most patients were male with an average age of 69 years and a CHADS₂ score of at least 1. Apixaban patients were followed for an average of 288 days, while the average follow-up period with rivaroxaban was 291 days. There was a significantly lower rate of the primary composite efficacy outcome with apixaban versus rivaroxaban (HR 0.82, 95% CI 0.68-0.98). The rate of the composite safety outcome of major bleeding was lower with apixaban versus rivaroxaban (HR 0.58, 95% CI 0.52-0.68). Bleeding event rate differences were driven by lower GI bleeding rates between groups. See Table 2. A subgroup analysis of patients over 70 years of age noted similar findings for both the composite efficacy (165 versus 132 events, HR 0.79, 95% CI 0.63-0.99) and safety outcomes (473 versus 272 events, HR 0.57, 95% CI 0.49-0.66). No significant differences were noted in the incidence of hepatitis, vasculitis, or heart failure hospitalizations.

Table 1: Selected Baseline Characteristics After Propensity Score Matching

Characteristic	Rivaroxaban (n = 39, 351)	Apixaban (n = 39, 351)
Female, n (%)	15,577 (39.6)	15,603 (39.7)
Mean age (SD), y	69.3 (10.6)	69.4 (10.5)
CHADS2 score = 1	16,944 (43.1)	16,672 (42.4)
CHADS2 score ≥ 2	17,503 (44.5)	17,664 (44.9)
Hypertension	31,479 (80)	31,492 (80)
Heart failure	8,518 (21.6)	8,482 (21.6)
Ischemic stroke or TIA	3,627 (9.2)	3,678 (9.3)
Ischemic heart disease	15,282 (38.8)	15,299 (38.9)
Obesity or overweight	11,219 (28.5)	11,222 (28.5)
Prescription antiplatelets, n (%)	3,777 (9.6)	3,806 (9.7)
Mean creatinine level (SD)	0.98 (0.26)	1.01 (0.30)

Table 2: Primary Outcome Results After Propensity Score Matching

Outcome	Rivaroxaban (n = 39, 351)	Apixaban (n = 39,351)	Hazard Ratio (95% CI)
Stroke
Events, n	232	198	0.85 (0.70-1.03)
Rate per 1000 PY	7.41	6.38
Systemic Embolism
Events, n	19	9	0.47 (0.21-1.04)
Rate per 1000 PY	0.60	0.29
GI Bleeding
Events, n	561	290	0.52 (0.45-0.59)
Rate per 1000 PY	17.93	9.35
ICH
Events, n	124	113	0.91 (0.71-1.18)
Rate per 1000 PY	3.95	3.64

*PY = person-years

Our Critical Appraisal

This study included a large patient population with data extracted from a large commercial insurance database. The primary outcomes of this study are similar to those presented in previous studies comparing DOACs.^7,9 The authors also included additional safety outcomes including specific bleeding events and occurrence of hepatitis or vasculitis.

Although the database used in this study represents a national sample, it has several limitations (as do all retrospective studies that rely on existing medical records). Missing data that could have influenced the results include information on over-the-counter medications (e.g. aspirin), body mass index, race, socioeconomic status, or HAS-BLED bleeding risk scores. Almost 30% of the study population was classified as overweight or obese, but without each patient’s BMI it is not possible to evaluate the appropriateness of DOAC usage. The investigators only used primary diagnosis codes associated with hospitalizations to determine event outcome rates, possibly excluding patients with a study outcome listed as a secondary diagnosis. The follow-up was, on average, less than one year. Lastly, the investigators used the CHADS₂ score for risk assessment instead of the currently recommended CHA₂DS₂-VASc score.^3.4

Previous observational studies did not find a significant difference in the rate of stroke or systemic embolism between DOACs, but some have found a lower risk of major bleeding with apixaban versus rivaroxaban and dabigatran.^9,10 One other large observational study reported both a lower risk of stroke or systemic embolism and major bleeding with apixaban versus rivaroxaban.¹¹ An increased risk of major bleeding and ICH has been reported with rivaroxaban when compared to other DOACs.^10,12

While this study provides further favorable efficacy and safety data supporting the use of apixaban over rivaroxaban in NVAF, patients on lower doses of apixaban or rivaroxaban were not included.  Thus, the results of this study should not be extrapolated to populations where dose adjustments are needed (e.g. obese patients or those with renal insufficiency).

When reflecting on the results of this trial, it is also prudent to compare drug-related characteristics. Apixaban is dosed twice daily and rivaroxaban is dosed once daily. For NVAF, rivaroxaban must be taken with food. More consistent anti-factor Xa activity has been reported with apixaban, which may explain the improved efficacy and lower rates of major bleeding.¹³

The Bottom Line

The results of this study reaffirm the safety of apixaban and provide additional evidence to support its preferential use over rivaroxaban. Based on this study and other observational data we currently have, apixaban should be recommended as the preferred DOAC for stroke prevention in patients with NVAF.  None-the-less, insurance coverage and medication adherence concerns must also be factored when making prescribing decisions for individual patients.

The Key Points

• The 2019 AHA/ACC/HRS and 2018 CHEST atrial fibrillation guidelines recommend DOACs as first-line anticoagulation for stroke prevention in patients with non-valvular atrial fibrillation (NVAF)
• There are no head-to-head randomized controlled studies comparing safety and efficacy of DOACs in NVAF
• Apixaban demonstrated a significantly lower risk of stroke/systemic embolism and bleeding complications when compared to rivaroxaban in a large, well-designed retrospective cohort study

FINAL NOTE:  This program will be available for recertification credit through the American Pharmacists Association (APhA) Ambulatory Care Review and Recertification Program.  To learn more, visit APhA BCACP Recertification – Evidence-Based Practice Series .