Anticoagulation On Demand? The Pill-in-the-Pocket Approach to A-fib Treatment

For decades now, stroke prevention in patients with atrial fibrillation has largely consisted of chronic oral anticoagulation, often with no end in sight. But a growing number of experts are asking: what if there were a way to keep the benefits of treatment while drastically minimizing the risk and inconvenience?

Known as intermittent, on demand, or “pill-in-the-pocket” anticoagulation, the strategy has gotten a feasibility boost from recent advancements in both medical therapy and rhythm monitoring technology. The approach involves continuously looking for A-fib with either implanted or wearable devices and then initiating oral anticoagulation for a limited time afterward. The idea is that strokes related to the onset of A-fib will be prevented while only exposing the patient to the inherent risks of bleeding—including the dreaded intracranial hemorrhages—for a short period of time. And an added bonus? Strategies like this, should they pan out in ongoing studies, involve the patient much more actively in his or her own care, which can have trickle-down effects on other preventive behaviors.

Rod Passman, MD (Northwestern University Feinberg School of Medicine, Chicago, IL), who has been heavily involved in researching the pill-in-the-pocket approach, explained that chronic anticoagulation has been used in patients with A-fib because for several decades the treatment of choice for stroke prevention was warfarin, a drug with a long onset/offset of action. Also, there were no good methods for determining when and if a patient went into A-fib, a key requirement for on-demand anticoagulation.

But now, Passman said, the emergence of safe and effective non-vitamin K antagonist oral anticoagulants (NOACs) with more rapid onset/offset of action and a variety of both implantable and wearable devices that can detect A-fib has opened up the possibility of more targeted therapy, especially when many patients can achieve something approaching a cure for their A-fib—or at least several arrhythmia-free years—with antiarrhythmics or catheter ablation. He questioned the logic of giving chronic oral anticoagulation to a 65-year-old patient who undergoes ablation for highly symptomatic A-fib and remains free of recurrences on continuous monitoring after the procedure, for example.

“My point is that the way we practice this aspect of A-fib management is stupid,” Passman said. “We know that only about half the patients who should be receiving an anticoagulant are actually prescribed one and when we do prescribe it probably only half take it at the end of a year or two. So to continue to think that this is something that patients want and can be compliant with I think is really foolish. My opinion is that we need to leverage these advances in pharmacology and technology to really fundamentally ask whether we’ve been treating this disease correctly.”

Other experts interviewed by TCTMD agreed that pill-in-the-pocket anticoagulation is an idea worth pursuing for an A-fib population that is expected to continue to grow in the coming years. In fact, the strategy is already being used in select patients, despite the lack of large randomized trials showing that it’s safe and effective. But before being implemented more broadly, questions around the biological rationale of treating only after detected episodes of A-fib; when to start anticoagulation after an episode and for how long; the capabilities of existing technology; and the best way to ensure patient compliance all need to be worked out.

The Evidence

Guidelines make no mention of pill-in-the-pocket anticoagulation, given the lack of solid evidence, instead stating only that the need for and choice of anticoagulant therapy should be reevaluated at periodic intervals. But the idea of starting anticoagulation when it’s needed, and pressing pause when it’s not, has been addressed in some small, pilot studies.

The 59-patient REACT.COM study, for example, evaluated tailored anticoagulation with a NOAC guided by continuous monitoring with an previously implanted insertable cardiac monitor. Anticoagulation was used for 30 days after any A-fib episode lasting for an hour or more. The total time on a NOAC was reduced by 94% compared with how long patients would have been treated if they were on chronic anticoagulation. No patients had a stroke, but there was one TIA that was not preceded by A-fib at any time in the year leading up to the event. There were two major bleeds in patients not taking NOACs that likely would have been fatal had they occurred on the background of anticoagulation, Passman said.

Another pilot study, TACTIC-AF, evaluated use of intermittent anticoagulation guided by continuous monitoring through dual-chamber pacemakers and implantable cardioverter-defibrillators in 48 patients. Those who remained free of A-fib episodes lasting 6 minutes or more with a total A-fib burden of less than 6 hours a day for 30 consecutive days stopped taking their NOAC. In this study, there was a 75% reduction in anticoagulation time. There were no strokes or TIAs, and only one major bleed in a patient not on oral anticoagulation.

Now [patients are] going to be active participants in their care by taking the responsibility of wearing the watch, keeping the watch charged, looking at the data, responding to the data, and so that’s a little bit of a paradigm shift. Suneet Mittal

These studies suggest that pill-in-the-pocket anticoagulation is feasible, but it still isn’t definitively known whether it is safe—namely, providing stroke protection comparable to chronic anticoagulation—and whether it can be offered to patients who do not have an indication for an implantable device, which could limit the ability to scale up this procedure to a broader swath of the A-fib population.

Jason Andrade, MD (University of British Columbia, Vancouver, Canada), said the state of the evidence highlights the need for caution when considering using intermittent anticoagulation, noting that the only large randomized trial evaluating a tailored anticoagulation approach—IMPACT, in which patients with implanted devices were randomized to anticoagulation guided by remote monitoring or usual, office-based care—was stopped early after showing no difference in a composite of stroke, systemic embolism, or major bleeding.

Pill-in-the-pocket anticoagulation “is a reasonable thing to think about,” Andrade commented to TCTMD. “The problem is that the research to date hasn’t gotten to the point where you’d definitively get an answer.”

Potential Barriers

Aside from the lack of a definitive trial, there are potential obstacles that stand in the way of wider implementation of intermittent anticoagulation guided by continuous rhythm monitoring.

Perhaps most important is lingering controversy about the temporal relationship between episodes of A-fib and strokes. If strokes in patients with atrial fibrillation are not directly related to episodes of a certain duration—and the arrhythmia is simply a marker of other things that are independently associated with stroke, for instance—then the whole concept of pill-in-the-pocket anticoagulation falls apart.

On the other hand, assuming a temporal link between an A-fib episode and stroke, there is then the question about how long an episode needs to be to trigger a stroke and how long the risk of an event lasts after that point. When to initiate oral anticoagulation and how long treatment should last are major unresolved questions, Andrade said.

Another issue that potentially limits use of intermittent anticoagulation is the accuracy of wearable devices designed to detect arrhythmias, which will be integral to bringing the strategy to the large group of patients with A-fib who might be eligible for it. Wearable consumer devices are starting to move into this space and there are now two that have been cleared by the US Food and Drug Administration for sensing A-fib: the Apple Watch Series 4 (Apple) and the KardiaBand (AliveCor), which attaches to earlier versions of the Apple Watch.

Suneet Mittal, MD (Valley Medical Group, New York, NY), speaking with TCTMD, insisted that the algorithms used to detect A-fib in these wearable devices have to get better before they can be trusted to guide treatment.

“I think they need more work. I think that what we know right now is that when patients are at rest, these devices work reasonably well, though they’re still not able to classify all the rhythms properly. When they classify them at rest they do a good job but unfortunately there’s still too many rhythms that the devices are not able to classify. And then, more importantly, when people move around, the accuracy of the system seems to drop off significantly and so both of those issues have to be improved on,” Mittal told TCTMD.

Passman, however, pointed to a recent study by his group showing that among 24 patients with paroxysmal A-fib equipped with both Apple Watches with KardiaBands and insertable cardiac monitors, the sensitivity of the smartwatch for detecting A-fib episodes of an hour or more was 97.5%.

The watches are sensitive, and in Passman’s opinion, the performance of wearables to detect A-fib duration of an hour or longer will be sufficiently high to support pill-in-the-pocket anticoagulation.

For older patients who have other comorbid disease like heart failure, hypertension, diabetes, all of those things which might make them have a stroke anyway, it’s not clear that they would benefit. Jean Connors

But Mittal has other concerns about a strategy of on-demand anticoagulation, including patient compliance. This is a potential “Achilles’ heel,” he said, especially when considering that A-fib can recur months or years after an apparently successful ablation. A patient can be convinced to be compliant with wearing some type of monitor for the first several months, Mittal said, “but let’s say day after day that the patient is free of A-fib. Can we be as sure that the patient will be motivated at a year, 2 and 3 years, to still be checking their rhythm and be compliant with it? We don’t know.”

Passman acknowledged that compliance with wearing a smartwatch, for example, over time remains to be seen. But he pointed out that in an online survey his team recently sent out through the StopAfib.org website to get a sense of where patients stand on pill-in-the-pocket anticoagulation, 85% of the 1,500 respondents said they would wear a watch for 14 hours a day to avoid chronic anticoagulation and 79% would enroll in a randomized trial if the opportunity was offered to them. “People really simply do not want to be on these drugs long-term and if they can minimize their exposure to it without compromising stroke risk, well that’s a win-win,” Passman said.

Looking for Answers

At least some of these issues might be addressed if Passman and his colleagues can get the planned REACT-AF trial off the ground; they’re currently going through the process of applying for grant funding from the US National Institutes of Health.

The investigators plan to enroll 5,000 patients at 100 US sites. They’ll recruit patients with paroxysmal A-fib and CHA₂DS₂-VASc scores of 1 to 4 and randomize them to chronic NOAC therapy or NOAC therapy guided by continuous monitoring using the Apple Watch Series 4, in which any A-fib episode lasting more than an hour will trigger a month of NOAC therapy. The noninferiority trial will have a primary composite endpoint of any stroke, arterial embolism, or cardiovascular death through 3 years of follow-up; a secondary endpoint, major bleeding, will be assessed for superiority.

Jean Connors, MD (Brigham and Women’s Hospital, Boston, MA), medical director of the anticoagulation management service at her center, said REACT-AF is targeting the right patients, as those with higher risk scores are those with multiple risk factors for stroke—like hypertension and diabetes—and would generally need to be on daily oral anticoagulation. She said it would be helpful if the trial includes a subset of patients also equipped with an implantable monitor, so that the accuracy of the smartwatch can be verified.

Whether REACT-AF will provide the type of solid evidence needed to guide clinical decision-making regarding the use of intermittent anticoagulation depends on event rates and power calculations, Andrade pointed out. But, he added, if the trial is able to show that continuous monitoring with the watch allows patients to limit their exposure to oral anticoagulation without increasing their risks of thromboembolism and stroke, it has the potential to lead to big changes in practice.

Before a pill-in-the pocket approach can be adopted more widely, this type of pivotal trial is needed to show that it is as effective for stroke prevention and reduces bleeding risk, cost, and the need to see the doctor, Passman said. “I think that this represents a potential paradigm shift and if we’re going to change the practice of medicine we need the highest levels of evidence to support that this is reasonable.”

They’ve shown that the strategy is feasible and that patients want to try it, Passman said, “and now we need to prove to the world that this approach doesn’t expose patients to an elevated stroke risk but does allow them to get off anticoagulation for vast periods of time, if not entirely.”

Best Candidates for Pill-in-the-Pocket

The population that might benefit from this strategy is potentially huge. In its latest statistics, the American Heart Association estimates that more than 5 million Americans are living with atrial fibrillation, including 700,000 who don’t know they have it, with the prevalence projected to rise to more than 12 million by 2030. Taking a wider view, results from the Global Burden of Disease Study 2016 indicate that around the world more than 46 million people have atrial fibrillation or flutter, most of whom would have indications for chronic oral anticoagulation to prevent stroke.

Not all of those patients would be eligible for intermittent anticoagulation, Passman said, noting that those with persistent or permanent A-fib and a high enough stroke risk would have to keep taking their pills every day. But he suggested it would be appropriate in patients with paroxysmal A-fib and CHA₂DS₂-VASc scores in the lower range, perhaps those who have undergone ablation or who are taking antiarrhythmic drugs. That group, which is now mostly treated according to guidelines with chronic anticoagulation, encompasses about half of the entire A-fib population, he added.

This isn’t for everyone with A-fib. A-fib is a very heterogeneous disease. It requires very individualized treatment. Rod Passman

“And [that proportion is] only going to grow as ablation becomes more and more commonplace and rhythm control strategies and the benefits of rhythm control strategies are realized more and more,” Passman said. “So this isn’t for everyone with A-fib. A-fib is a very heterogeneous disease. It requires very individualized treatment. But this strategy could affect a half if not more of the AF population, we believe.”

Others agreed that it’s the lower-risk patients with indications for oral anticoagulation who would be the best candidates for pill-in-the-pocket anticoagulation, although their predictions for the proportion of the population likely to benefit were not quite as optimistic as Passman’s.

Andrade estimated that intermittent anticoagulation could potentially be applied to 20% to 25% of the A-fib population when considering that physicians might want to limit its use to younger, more active patients who have a risk of bleeding complications and trauma related to sports and other activities. Older patients might be more accepting of being on an anticoagulant long-term, he said.

Even if only 20% of the total A-fib population could avoid chronic anticoagulation using on-demand anticoagulation, that would be worthwhile, according to Connors.

“For the younger, healthier person who has paroxysmal A-fib, this is going to be a great strategy. But for older patients who have other comorbid disease like heart failure, hypertension, diabetes, all of those things which might make them have a stroke anyway, it’s not clear that they would benefit,” she said.

Mittal said he’d feel comfortable offering it to patients with a CHA₂DS₂-VASc score of 1 or 2 after a successful ablation. If you have a 50-year-old person who has hypertension and diabetes plus no evidence of A-fib after ablation, he said, “it is really hard to convince that patient that for their whole life now they’re stuck with blood thinners. And so there are very educated patients who are looking for alternatives to balance risks and benefits for themselves.” Mittal noted that occlusion of the left atrial appendage could also be an alternative to chronic anticoagulation in these types of patients—a strategy being tested in the OPTION trial—and that ultimately the role of pill-in-the-pocket anticoagulation will be determined by forthcoming data.

Getting Patients More Involved

Both Mittal and Andrade highlighted an important feature of pill-in-the-pocket anticoagulation: it gets patients more engaged in their own care.

“For the first time, this strategy is critically dependent on how much engagement we get from our patients in the process, because now they’re going to be active participants in their care by taking the responsibility of wearing the watch, keeping the watch charged, looking at the data, responding to the data, and so that’s a little bit of a paradigm shift,” Mittal said. “Those of us . . . who believe that engaging the patient is important view this as a big positive. For many who think like that’s kind of an unreliable way to go about it, they don’t feel that this is a long-term solution and I think that’s something that also has to be borne out over time.”

This interest in the pill-in-the-pocket approach is driven by patients, Andrade said, “because it allows them to look after their own care and it allows them to dictate their treatments and how things are going.”

The benefits of that increased patient involvement go beyond A-fib management, Andrade indicated. “If you have someone who’s engaged, they’re more likely to be adherent to all aspects of their therapy, so then they’re more likely to take their blood pressure pills, they’re more likely to exercise and be focused on improving their clinical situation, whereas someone who’s not engaged is likely going to fail this type of therapy. And so this wouldn’t be a good approach for them. . . . By increasing patient autonomy by giving them a more active role in their care, I think you’re going to improve outcomes just in general.”