By Joshua D. Moss, MD
Associate Professor of Clinical Medicine, Cardiac Electrophysiology, Division of Cardiology, University of California, San Francisco
SYNOPSIS: Screening with an implantable loop recorder resulted in dramatically higher rates of atrial fibrillation detection and ensuing anticoagulation, but without a significant decrease in risk of stroke or systemic embolism by six years of follow-up.
SOURCE: Svendsen JH, Diederichsen SZ, Højberg S, et al. Implantable loop recorder detection of atrial fibrillation to prevent stroke (The LOOP Study): A randomised controlled trial. Lancet 2021;398:1507-1516.
Anticoagulation lowers the risk of stroke in patients with clinically diagnosed atrial fibrillation (AF). Further, more-intensive ambulatory monitoring helps clinicians detect more asymptomatic or “subclinical” AF better than conducting periodic ECGs or short-term Holter monitoring. Svendsen et al sought to investigate whether anticoagulation guided by intensive AF screening with an implantable loop recorder (ILR) could prevent strokes.
At four centers in Denmark, 6,004 patients (mean age = 74.7 years; 47.3% women) with at least one risk factor for stroke (other than age) but no previously diagnosed AF or indication for anticoagulation were randomized in a 1:3 ratio to ILR monitoring or usual care. Required risk factors for stroke included hypertension (90.7%), diabetes (28.5%), prior stroke (17.6%), or heart failure (4.4%). Most patients recorded a CHA2DS2-VASc score of 3 or higher. In the ILR group, automated remote transmissions were reviewed daily, and oral anticoagulation was recommended if AF lasting at least six minutes was detected. In the usual care control group, patients were followed per routine with their general practitioner and participated in an annual interview with a study nurse. The primary outcome was the combined endpoint of stroke or systemic arterial embolism, with multiple secondary outcomes, including cardiovascular and all-cause mortality. No patients in the control arm crossed over to the ILR group, and 81 patients in the ILR group did not receive an ILR. The median duration of ILR monitoring was 39.3 months, with 11.7% of ILR patients prematurely discontinuing monitoring (before three years without outcome, AF, or death). Total median follow-up was 64.5 months.
AF was diagnosed at significantly higher rates in the ILR group compared with the control group (31.8% vs. 12.2%), and oral anticoagulation was initiated at similarly higher rates (29.7% vs. 13.1%), most within the first month of AF diagnosis. More ILR patients who started anticoagulation for AF eventually discontinued anticoagulation (5.8% vs. 3.6%).The primary outcome of stroke or systemic embolism occurred less frequently in the ILR group (0.88 events per 100 person-years) than the control group (1.09 events per 100 person-years), but these differences were not statistically significant. Neither cardiovascular death rates nor all-cause mortality rates were significantly different between groups. In subgroup analyses, ILR patients with systolic blood pressure in the highest tertial (≥ 157 mmHg) demonstrated significantly lower rates of stroke or systemic embolism than control patients (3.46% vs. 6.74%; HR, 0.51; 95% CI, 0.31-0.83). Similar differences were not seen for patients with lower blood pressure. Rates of major bleeding, hemorrhagic stroke, and traumatic intracranial hemorrhage were not significantly different between groups. The authors concluded long-term ECG monitoring with an ILR resulted in significantly higher rates of AF detection and ensuing anticoagulation, but no significant decrease in the risk of stroke or systemic embolism by 5.4 years of follow-up.
The relationship between AF and risk of stroke or systemic embolism is well established but complex. Longer episodes of AF (> 23.5 hours) have been shown to be associated with higher risk than shorter episodes, although even shorter episodes may portend significant risk in patients with higher CHA2DS2-VASc scores. Not all cardioembolic strokes occur at or near the time of an AF episode, and not all strokes in patients with AF are caused by cardiac thromboembolism. Anticoagulation to prevent thromboembolism might increase the risk of hemorrhagic stroke. All these factors make it difficult to predict the benefits of intensive arrhythmia monitoring in asymptomatic patients, or when anticoagulation is warranted, questions that may become increasingly important as more patients turn to commercially available wearable monitoring devices like smartwatches. However, the LOOP study offers both a great deal of useful information and new hypotheses to test.
Loop recorders are a safe way to continuously monitor for AF long-term. In 1,420 patients with the device implanted, the ILR was explanted for complications in only nine (eight of whom received a new device). Higher rates of anticoagulation in this patient population were not associated with significantly more bleeding complications over almost six years. That is not to say therapeutic anticoagulation could not have led to some episodes of intracranial bleeding, but that feared event did not occur more often in the ILR group.
More aggressive monitoring, and the ensuing increase in anticoagulation rates, did not result in significantly lower rates of stroke or systemic embolism over six years. However, the event curves did appear to start separating at two to three years follow-up, and they continued to separate thereafter. Only 16.4% of participants were followed for the primary outcome at year 6, necessitating caution in interpretation. But it is not unreasonable to think a larger study or longer follow-up could have led to a different conclusion. Additionally, several factors may have biased outcomes against the ILR group. First, it is possible there was a lower average burden of AF (and lower overall stroke risk) in anticoagulated patients in the ILR group, if longer and/or more frequent episodes were required for AF detection in the control group. It also is possible AF was detected more often with “usual care” than might otherwise be expected, given participation in a clinical trial and higher awareness of AF.
AF is a cause of stroke and a marker of stroke risk. Of the 67 patients in the ILR group who recorded a primary outcome, only 17 experienced an event after AF was first detected, 15 of whom were on anticoagulation. There was no temporal relationship between AF episodes and stroke. More severe hypertension may warrant higher vigilance for AF and more aggressive initiation of anticoagulation, a hypothesis generated from the compelling subgroup data. Anticoagulation for AF will warrant a risk-benefit discussion and shared decision-making with patients. In my own practice, I still recommend anticoagulation for AF episodes longer than six minutes with higher CHA2DS2-VASc scores when a direct-acting oral anticoagulant is an option. In the absence of symptoms, I would not advocate for routine ILR monitoring. However, I may suggest a higher level of vigilance for patients with more severe hypertension.