ROME – Temporary interruption of oral anticoagulation for stroke prevention in patients with atrial fibrillation occurs often and is associated with substantially increased risk of both cardioembolic events and all-cause mortality, according to a new prespecified secondary analysis of the ENGAGE-AF TIMI 48 trial.

The analysis showed that many of these treatment interruptions occur in response to nonserious adverse events such as minor bleeding, planned dental procedures, or simply because of patient wishes. The new ENGAGE-AF TIMI 48 findings should encourage physicians and patients to think twice before interrupting anticoagulant therapy for such reasons, which look pretty flimsy in light of the new evidence of the potentially serious consequences, Ilaria Cavallari, MD, said at the annual congress of the European Society of Cardiology.

“Interruption of oral anticoagulation in patients with AF [atrial fibrillation] should be avoided or as brief as possible and under medical control, especially following nonserious adverse events,” said Dr. Cavallari of Brigham and Women’s Hospital in Boston.

The ENGAGE-AF TIMI 48 study was the pivotal phase III, double-blind, 21,105-patient clinical trial that led to Food and Drug Administration and European approval of the direct oral factor Xa inhibitor edoxaban (Savaysa) for stroke prevention in moderate- to high-risk patients with AF (N Engl J Med. 2013 Nov 28;369[22]:2093-104). The study showed that edoxaban at what later became the approved dose of 60 mg/day, or 30 mg/day in patients with impaired renal function, body weight of 60 kg or less, or on concomitant therapy with a platelet glycoprotein inhibitor, resulted in a 21% reduction in the risk of stroke or systemic embolism and a 20% reduction in major bleeding, compared with warfarin over 2.8 years of follow-up.

Dr. Cavallari presented a prespecified secondary retrospective analysis that focused on treatment interruptions: the reasons and the price paid in terms of thromboembolic events.

One or more treatment interruptions lasting for longer than 3 days occurred in 63% of patients during a median 2.8 years of follow-up. Since these were participants in a clinical trial with relatively close patient contact, it’s likely that the true interruption rate in real-world clinical practice is even higher, she said.

Interruptions were more significantly frequent in patients assigned to warfarin than among the groups assigned to edoxaban. The median duration of treatment interruptions was 9 days.

After excluding patients who were on any other anticoagulant during their interruption – low-molecular-weight heparin being the most common – investigators were left with 9,148 patients.

The endpoints of interest in this analysis were the major adverse events occurring during a time window lasting from 4 days after their last dose of oral anticoagulant until day 34 or when they resumed their study drug. The 30-day incidence of ischemic stroke or systemic embolism was 1.27%. The rate of a composite composed of cardiovascular death, MI, and ischemic stroke or systemic embolism was 4.99%. And the 30-day rate of an endpoint Dr. Cavallari termed primary net clinical outcome – a composite of stroke or systemic embolism, major bleeding, and all-cause mortality – was 7.16%.

These 30-day event rates among treatment interrupters are extremely high, compared with the 1-year rates in patients who didn’t interrupt oral anticoagulant therapy: 0.26% for ischemic stroke or systemic embolism; 0.36% for the composite of cardiovascular death, MI, and ischemic stroke; and 0.56% for the primary net clinical outcome, she continued.

The most common reason for treatment interruptions was adverse events, which accounted for 41% of the interruptions.

Drilling deeper into the types of adverse events that triggered treatment interruption, 1.5% of interrupters did so because of an on-treatment ischemic stroke or systemic embolism, 4.7% did so because of major bleeding, 8% had minor and clinically relevant nonmajor bleeding, another 30% interrupted treatment for other serious or nonserious adverse events.

Interrupting therapy because of an adverse event often had serious consequences, as reflected in an adjusted 3.94-fold increased risk of 30-day all-cause mortality, compared with patients who stopped for other reasons. Patients who stopped treatment because of a stroke, transient ischemic attack, or systemic embolism had a 30-day all-cause mortality of 29.3%. Those who interrupted treatment because of a major bleeding event had an 8.8% 30-day mortality. When minor or clinically relevant nonmajor bleeding was the impetus for a treatment interruption, the associated 30-day mortality was 3.4%.

Almost a third (29%) of treatment interruptions were the result of physician decisions in response to an upcoming invasive procedure, most often dental work.

The 30-day rates of ischemic stroke/systemic embolism and primary net clinical outcome didn’t differ significantly between patients who interrupted warfarin versus edoxaban at the approved dose. Nonetheless, this new secondary analysis from ENGAGE-AF TIMI 48 supports the parent study’s conclusion that edoxaban is preferable to warfarin in patients with AF, according to Dr. Cavallari.

“In light of the increased risk of ischemic events after interruption of oral anticoagulation, NOACs [new oral anticoagulants] represent an attractive alternative to vitamin K antagonists, given their faster onset of action, better adherence rates, safety, and tolerability profiles,” she concluded.

ENGAGE AF-TIMI 48 was funded by Daiichi Sankyo. Dr. Cavallari reported having no financial conflicts of interest regarding her presentation.

bjancin@frontlinemedcom.com

Interrupting oral anticoagulation in AF carries high thromboembolic cost