Rush to DOACs Spells Trouble for Frail Warfarin Users

AMSTERDAM -- Best let frail older people who are stable on international normalized ratio (INR)-guided vitamin K antagonist (VKA) management stay on it instead of switching to a direct oral anticoagulant (DOAC or NOAC), the open-label FRAIL-AF trial found.

Contrary to expectations, participants randomized to the DOAC switch ended up at elevated risk for a major or clinically relevant non-major bleeding complication over the next 12 months (15.3% vs 9.4%; HR 1.69, 95% CI 1.23-2.32) with no beneficial reduction in thromboembolic events (2.4% vs 2.0%; HR 1.26, 95% CI 0.60-2.61).

This unique trial exclusively comprising frail elderly people with nonvalvular atrial fibrillation (Afib or AF) was stopped early for futility in seeking superiority for the DOAC strategy.

"Switching from a VKA to a NOAC should not be considered without a clear indication in frail older patients with Afib," urged study author Geert-Jan Geersing, MD, PhD, of the University Medical Center Utrecht in the Netherlands, at the European Society of Cardiology (ESC) Congress. The FRAIL-AF trial was simultaneously published in Circulation.

DOACs are proven to confer lower bleeding risks in non-frail people with no tradeoff in efficacy, and are now preferred as first-line therapy for new Afib patients and people not managed well on a VKA. Even so, many frail, older people still rely on decades-old VKA blood thinners, such as warfarin, for stroke prevention.

"I think if someone is on a VKA and [has] great TTR [time in therapeutic range] without bleeding events and does not mind the dietary restrictions nor getting their INR checked, then staying on a VKA antagonist is reasonable," commented Rod Passman, MD, of Northwestern University Feinberg School of Medicine in Chicago.

"The question remains whether we start NOAC vs VKA in a frail population. The other thing to remember is that left atrial appendage occlusion would be an option for many of these patients," he told MedPage Today.

Frailty in aging may be recognized subjectively as having multiple diseases, needing lots of medications, depending on others for daily living, and having a reduced capacity to resist stressors. Various scales exist to quantify the risk of frailty.

In a press conference, Geersing said that Afib is common in the frail, reaching approximately one in four or even more.

He and his FRAIL-AF collaborators warned that they were unable to definitively prove which type of anticoagulation was most effective for this population.

"Event rates for thromboembolic events, major bleeding in isolation, hemorrhagic stroke, or the composite of hemorrhagic and ischemic stroke were low in both treatment arms, withholding us from drawing firm conclusions on these also clinically relevant outcomes," they wrote.

Moreover, the study really compared the VKA vs DOAC management strategies, not specific molecules.

For their open-label trial, Geersing and colleagues included Afib patients age 75 or older scoring 3 or higher on the Groningen Frailty Indicator (GFI), excluding those with a glomerular filtration rate <30 mL/min/1.73 m² or with valvular Afib. Participants were randomized to switch from INR-guided VKA treatment to a DOAC or to continued VKA treatment with INR monitoring at eight Dutch study sites.

The 1,330 people randomized had a mean age of 83 and fewer than 40% were women. Median GFI and CHA₂DS₂-VASc scores were both 4. Over 20% of the cohort had heart failure and another 20% had diabetes.

The switching arm of the study went on the DOACs rivaroxaban (Xarelto; 50.2%), apixaban (Eliquis; 17.4%), edoxaban (Savaysa; 16.5%), or dabigatran (Pradaxa; 8.6%). Off-label DOAC dose reductions were given to 6.6% at the discretion of treating physicians.

Bleeding rates were similar between rivaroxaban (HR 1.95, 95% CI 1.36-2.79) and apixaban (HR 2.17, 95% CI 1.28-3.68) and notably lower notably for edoxaban (HR 1.10, 95% CI 0.57 to 2.13).

"Nevertheless, these analyses should be interpreted with caution as they were post hoc and non-randomized," the authors wrote.

They reported 89% adherence in the intervention arm versus 92% adherence in the control arm.