Dabigatran Discontinuation Rate Highest Early After Treatment Begins

If patients newly diagnosed with A-fib stick with dabigatran (Pradaxa; Boehringer Ingelheim) therapy for at least 6 months, there’s a very good chance they’ll still be taking the oral anticoagulant several months later, new data out of the GLORIA-AF registry show.

The persistence rate was lowest in the first 6 months of treatment (83.7%) and steadily increased for each successive 6-month period up to 2 years, reaching 96.3% between 18 and 24 months, lead author Miney Paquette, MSc (Boehringer Ingelheim, Burlington, Canada), and colleagues report.

“Closer follow-up in the first 6-month period following diagnosis and treatment initiation is key to improving persistence and optimizing patient outcomes,” Paquette said in emailed comments. “Any interventions to improve persistence should be focused in this period and target patient groups at higher risk to discontinue.”

Of note, most of the discontinuations (65%) were for reasons other than adverse events, such as cost, the start of bridging therapy, drug or alcohol abuse, and dementia. “As the results demonstrate that most of the discontinuations are not due to adverse events, there may be an opportunity to further educate patients and clinicians on the importance of remaining on guideline-adherent anticoagulation if they have at least one stroke risk factor,” Paquette said.

Matthew Reynolds, MD (Lahey Hospital & Medical Center, Burlington, MA), who was not involved in the study, made a similar point.

“One of the keys clinically is just seeing the patients and making sure that they have appropriate follow-up and reinforcement of the need for [oral anticoagulation] and a clear understanding of the purpose of it,” he told TCTMD, pointing out that some patients will drop it at the first sign of so-called nuisance bleeding.

“I do think part of our jobs as clinicians is to reinforce how important these treatments are because we know how bad strokes are,” Reynolds said. “That reinforcement is a big part of it, helping people get through that nuisance bleeding, which is not completely avoidable.”

The findings were published online recently ahead of print in the American Journal of Cardiology.

Discontinuation and Clinical Outcomes

Historically, there have been high rates of discontinuation of oral anticoagulation, an area that was dominated by warfarin for decades. There’s some evidence indicating that adherence might be better with the newer non-vitamin K antagonist oral anticoagulants (NOACs), but, Paquette said, there are few prospective studies exploring persistence with NOAC therapy in clinical practice.

To explore the issue, the investigators turned to phase 2 of the GLORIA-AF registry, which included patients newly diagnosed with A-fib in 44 countries spanning five global regions (North America, Latin America, Europe, Asia, and Africa/Middle East). All patients had at least one additional stroke risk factor. The current analysis included 4,859 patients (mean age 70.2 years; 55.7% men) who took at least one dose of dabigatran.

Through the 2-year study period, the overall persistence rate was 70.9%, which the authors note is higher than previously seen with warfarin and in claims-based studies of the NOACs. The longer patients took dabigatran, the greater the likelihood was that they would continue with treatment moving forward—persistence at 1 year was associated with a probability of remaining on therapy at 2 years that exceeded 90%.

Certain characteristics were associated with persistence. Discontinuation was higher in patients with symptomatic A-fib, prior bleeding, and use of proton pump inhibitors and lower in those with increasing body mass index. In addition, compared with patients in Europe, those in North America and Asia had higher discontinuation rates and those in Latin America and Africa/Middle East were more likely to stick with their regimens.

Identifying features associated with discontinuation “could support targeted interventions in these patients,” the authors say.

People who do well with it in the early going are likely to continue to do well with it for a good while later. Matthew Reynolds

Of the 26.9% of patients who discontinued dabigatran, roughly half switched to another oral anticoagulant and half did not. Adverse events like bleeding, dyspepsia, hypersensitivity to the drug, or a severe concomitant medication interaction were listed as the reason for stopping dabigatran in 35% of cases.

Discontinuation was associated with increased rates of stroke, major bleeding, vascular death, and all-cause death compared with remaining on treatment. Switching to another oral anticoagulant mitigated those differences, however. For example, the standardized stroke rate (per 100 patient-years) was 1.76 among those who did not start taking another anticoagulant and 1.02 among those who made a switch. A similar difference was seen for all-cause mortality (7.25 vs 3.12 per 100 patient-years).

“Further research aimed at understanding the reasons why patients discontinue treatment is needed and prospective studies which assess these reasons in greater detail may be helpful to develop effective interventions to improve patient adherence and persistence to anticoagulants,” Paquette said.

‘Reasonably Good’ Persistence

Reynolds said the study provides useful information about persistence with a NOAC—specifically dabigatran—in a population of patients that is less selected than a cohort from a randomized trial.

“The big take-away is that if people get through the first 6 months on that drug, then there is something like a 90% chance that they’ll still be on it after 2 years,” he said. “We know that some patients have side effects and we know that some patients just aren’t good anticoagulant candidates in general, so there’s always some dropout.”

Reynolds said overall persistence in this study was “reasonably good—maybe not quite as high as you would see in a clinical trial, but better than a lot of historical data for oral anticoagulants in general.”

He pointed out that it’s hard to compare this study with prior ones because it’s based on a large, well-organized, prospective, postmarket registry and not a claims database or other cohort of real-world patients.

Nevertheless, the study provides “very reassuring data about this specific drug [in] that the people who do well with it in the early going are likely to continue to do well with it for a good while later,” Reynolds concluded.