Interruption of Rivaroxaban vs Warfarin for AF: A Comparison of Risk

April 9, 2013

Interruption of anticoagulation with rivaroxaban or warfarin should be bridged with an agent such as low-molecular weight heparin.

Physicians often are faced with the need to temporarily discontinue anticoagulation in patients, most commonly in advance of surgical procedures. With warfarin anticoagulation, the risks of discontinuation are familiar and well understood. Discontinuation of the novel anticoagulants however, is less well understood.

ROCKET-AF was a randomized, double-blind trial of rivaroxaban vs. warfarin in 14 624 patients with non-valvular atrial fibrillation (AF) and a high risk of stroke (average CHADS2 was 3.4).1 The study raised concern about increased risk following discontinuation of rivaroxaban because there was a higher rate of the composite endpoint of stroke and systemic embolism in patients who transitioned from rivaroxaban to open-label warfarin-a finding that suggested a “rebound” hypercoagulable phenomenon.

To characterize this relationship, Patel and coworkers2 undertook a post-hoc analysis of interruptions of rivaroxaban in the ROCKET-AF trial, where interruptions were defined as: (1) temporary interruption for 3 or more days; (2) early permanent drug discontinuation; and (3) end-of-study transition to open-label warfarin. There were 2165 patients on rivaroxaban and 2528 patients on warfarin who stopped or interrupted treatment during the trial (average duration of interruption was 5 days). The most common reason for stopping treatment was surgery or an invasive procedure (39.7%). Other reasons for interruption included nonbleeding adverse events (25%), bleeding (13.2%), and medication error (18.1%).

The rates of the primary trial endpoint of stroke and systemic emboli in the rivaroxaban discontinuation group were similar to warfarin discontinuation in the temporary interruption cohort (9 vs.8, p=0.62) and early permanent discontinuation cohort (42 vs. 36, p=0.66).  Notably, however, this was not the case in those who were transitioning to open label therapy, with a stroke rate 3.7 times the rate of stroke with rivaroxaban discontinuation compared with warfarin (22 vs. 6, HR 3.72, p=0.004). 

Based on these data, it is reasonable to conclude that the risk of interrupting rivaroxaban in patients with a high CHADS2 score was due to inadequate anticoagulation during the transition to warfarin and not due to rebound hypercoagulability from the drug itself, as this risk was not observed in those who interrupted rivaroxaban without transitioning to open-label therapy. Indeed, it took longer for the rivaroxaban-to-open-label-therapy group to achieve therapeutic anticoagulation compared with those treated with warfarin throughout.

In general, if there is a need to interrupt anticoagulation for a surgical procedure, regardless of whether it is warfarin or rivaroxaban, physicians can minimize events in high-risk patients by using a bridging agent, such as low molecular weight heparin.  However, it is encouraging that rivaroxaban does not appear to result in rebound hypercoagulability.

1.  Patel MR, Hellkamp AS, Lokhnygina Y, et al. Outcomes of discontinuing rivaroxaban compared with warfarin in patients with nonvalvular atrial fibrllation: analysis from the Rocket-AF Trial (Rivaroxaban Once-Daily, Oral, Direct Factor Xa Inhibition Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation).J Am Coll Cardiol. 2013;61:651-658. (Abstract)

2.  Patel MR, Mahaffey KW, Garg J, et al. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med. 2011;365:883-891.  (Full text)