NOACs for Preventing Stroke in AF

Clinical Case

Elaine M. Hylek, MD, MPH; Benjamin A. Steinberg, MD, MHS; Reviewer: Robert P. Giugliano, MD, FACC


January 23, 2019

Editorial Collaboration

Medscape &


An 82-year-old woman presents to the hospital with shortness of breath and tachycardia. She is found to be hypervolemic and in atrial fibrillation (AF) with a rapid ventricular response. She is admitted to the hospital and administered intravenous diuretics and oral beta-blockers. A transthoracic echocardiogram reveals mildly depressed left-ventricular function (ejection fraction of 45%) and a mildly dilated left atrium.

Medical history is significant for paroxysmal AF, hypertension, and obstructive sleep apnea which is treated with continuous positive airway pressure.

Several years ago while taking warfarin and aspirin (ASA), she had a gastrointestinal (GI) bleed requiring transfusion. She is unclear of the source of the hemorrhage but has since been treated with dual antiplatelet therapy (DAPT) with ASA and clopidogrel for stroke prevention, without recurrent bleeding.

Her creatinine level is 1.3 mg/dL and she weighs 75 kg. Thyroid studies are unremarkable, and she had a negative stress test last year (performed by her primary care physician). She describes multiple similar—though of lesser severity—episodes of shortness of breath and tachycardia over the past several months and increasing weight gain and lower-extremity edema.

The cardiovascular medicine service is consulted regarding stroke prevention and rhythm management of AF in the setting of heart failure (HF).


The patient is admitted to the hospital and receives medical therapy for acute HF, including intravenous diuresis and beta-blockers to control the ventricular rate in AF. She remains symptomatic from AF and is transitioned from DAPT to apixaban 5 mg twice daily. Once euvolemic, she undergoes transesophageal echocardiography, revealing no left atrial appendage thrombus, and is then cardioverted to normal sinus rhythm.

She is discharged on apixaban, metoprolol, lisinopril, and furosemide, and initiated on amiodarone for short-term rhythm control in anticipation of possible catheter ablation for AF. Both ASA and clopidogrel were stopped when apixaban was initiated.


Choosing a NOAC

The patient described has a CHA2DS2-VASc score of 5 for her sex (1 point), age ≥ 75 (2 points), hypertension (1 point), and HF (1 point).[1] Thus, she warrants stroke prevention therapy with oral anticoagulation (OAC).[2]

The non–vitamin K antagonist oral anticoagulants (NOACs), as a group, have been demonstrated to be at least as good as warfarin at preventing stroke, with a generally lower risk of bleeding, overall.[3] While dosing of NOACs is often more straightforward than for warfarin and NOACs have fewer interactions, dose adjustments vary by agent.

However, she had a prior GI bleed in the setting of combination therapy with warfarin and ASA. Concomitant antiplatelet therapy is known to increase the risk of bleeding associated with any oral anticoagulation[4,5,6,7] and should be avoided in the absence of a strong indication. This patient does not have a clear indication for antiplatelet therapy and may better tolerate OAC in the absence of antiplatelet therapy.

She was subsequently treated with DAPT for stroke prevention in AF, a strategy primarily tested in the ACTIVE clinical trial program. The ACTIVE W trial was stopped early due to superiority of vitamin K antagonism over ASA plus clopidogrel for thromboembolic prevention.[8] Notably, rates of major bleeding were not significantly lower in the DAPT group. In a more contemporary comparison, the AVERROES investigators[9] compared apixaban with ASA among patients with AF who were intolerant to (or refused) warfarin for stroke prevention. Again, the oral anticoagulant (apixaban) was demonstrated to be superior to antiplatelet therapy for prevention of thromboembolic events. While the trial was not powered to assess bleeding, rates of bleeding on apixaban were not increased compared with ASA. To date, there have not been similar large randomized trials of any of the other NOACs in comparison to ASA.

The patient in this vignette is treated with standard-dose apixaban; while she is ≥ 80 years old, her weight and creatinine level do not support the reduced dose, which requires at least two higher-risk criteria (her weight is > 60 kg and her creatinine is < 1.5 mg/dL).[10] Emerging data suggest that inappropriate dosing of these agents, particularly underdosing of apixaban, may be associated with increased risk for stroke and other cardiovascular events.[11]

Of the four NOACs, apixaban and edoxaban had significantly less major bleeding than warfarin in their respective phase 3 trials versus warfarin, an advantage that was also seen in elderly patients.[3,12,13,14,15] The patient's estimated creatinine clearance, as calculated by the Cockroft-Gault formula, is 39.5 mL/min.[16]

Edoxaban is a once-daily factor Xa inhibitor that is 50% renally cleared, so use in this patient (because her CrCl is < 50 mL/min) would require a dose adjustment from the standard dose of 60 mg daily down to 30 mg daily, based on the ENGAGE AF-TIMI 48 trial.[13,17] It should also be noted that unique, in-depth analyses of the ENGAGE AF-TIMI 48 trial's pharmacokinetic and pharmacodynamic data supported the clinically based dosing strategy for this agent.[18,19]

Dabigatran and rivaroxaban are less preferred because they had similar to higher rates of major bleeding in elderly patients in their respective trials. If the patient were to be treated with rivaroxaban, she should receive a reduced dose of 15 mg daily.[20] In analyses of the ROCKET-AF trial, as well as heterogeneity testing across the trials, rivaroxaban has been associated with an increased risk for GI bleeding compared with warfarin[3,21] and thus may not be the best agent for this patient. If she were treated with dabigatran, her estimated creatinine clearance (> 30 mL/min) would not indicate a dose reduction unless she were taking concomitant dronedarone or ketoconazole (strong P-glycoprotein [P-gp] inhibitors). In the absence of P-gp inhibitors, 150 mg twice daily would be the indicated dose for stroke prevention in a US patient with AF.[22] Current US guidelines recommend close attention to its use in patients with impaired renal function, particularly those of advanced age, and European guidance would support dose modification.[2,23] While a modified dose of 110 mg twice daily was tested in the RE-LY trial,[7] this dose was not approved for use in the United States. The US Food and Drug Administration cited the risk of inappropriate underdosing and possible surfeit of thromboembolic events as the reason for not approving it,[24] and there have been some data to support this risk; however, the reduced dose is available for other indications in the United States.[25]

Rhythm Control

The patient is started on amiodarone for rhythm control, a drug with well-known interactions primarily attributable to its P-gp and cytochrome P-450 3A4 inhibitions. Several post-hoc analyses of the major clinical trials of NOACs have failed to demonstrate a clinically relevant interaction in the form of increased risk for adverse events associated with concomitant amiodarone and approved dosing of NOACs, compared with warfarin.[26,27,28,29] In fact, the trials in general did not require a dose adjustment in that setting, and thus the package inserts do not. However, those post-hoc subgroup analyses may have been underpowered to detect such effects; the most recent US guidelines highlight the uncertainty.[2] Current European guidance supports no NOAC dose adjustment with concomitant amiodarone in the absence of complicating factors.[23]

Should the patient ultimately undergo catheter ablation? Emerging evidence supports more aggressive use of catheter ablation for AF among patients with concomitant HF.[30,31] The majority of observational data, and some clinical trial data, have shown NOACs to be at least as safe and effective as warfarin for periprocedural use around AF ablation,[32,33,34,35,36] during which risks for stroke and bleeding are significantly higher; additional clinical trial data on this topic are forthcoming.[37] Given her CHA2DS2-VASc risk of 5, current guidelines would not support discontinuing OAC following ablation, irrespective of AF burden at that time.[2,38]


For stroke prevention in AF, NOACs have been demonstrated to be safe and effective in a variety of settings and are generally easier to manage than warfarin. However, there remain unique considerations in dosing and management of each of the different NOACs, which may affect clinical decision-making and outcomes.