Author + information
- Received April 23, 2018
- Revision received May 31, 2018
- Accepted June 4, 2018
- Published online November 19, 2018.
- Shaun Giancaterino, MD,
- Florentino Lupercio, MD,
- Marin Nishimura, MD and
- Jonathan C. Hsu, MD, MAS∗ ()
- Cardiac Electrophysiology Section, Division of Cardiology, Department of Medicine, University of California, San Diego, La Jolla, California
- ↵∗Address for correspondence:
Dr. Jonathan C. Hsu, Cardiac Electrophysiology Section, Division of Cardiology, Department of Medicine, University of California-San Diego, 9452 Medical Center Drive, 3rd Floor, Room 3E-417, La Jolla, California 92037.
Insertable cardiac monitors (ICMs) are small, subcutaneously implanted devices offering continuous ambulatory electrocardiogram monitoring with a lifespan up to 3 years. ICMs have been studied and proven useful in selected cases of unexplained syncope and palpitations, as well as in atrial fibrillation (AF) management. The use of ICMs has greatly improved our ability to detect subclinical AF after cryptogenic stroke, and application of this technology is growing. Despite this, current stroke and cardiology society guidelines are lacking in recommendations for monitoring of subclinical AF following cryptogenic stroke, including the optimal timing from stroke event, duration, and method of electrocardiogram monitoring. This focused review outlines the current society guidelines, summarizes the latest evidence, and describes current and future use of ICMs with an emphasis on detection of subclinical AF in patients with cryptogenic stroke.
Atrial fibrillation (AF) is the most common cardiac arrhythmia, and it can lead to thromboembolic stroke, most commonly from emboli arising from the left atrial appendage (1). It is estimated that AF is the source of 1 in 4 ischemic strokes (2). Stroke accounts for 1 in every 20 deaths in the United States and is a leading cause of serious long-term disability and health care expenditure (3). Cryptogenic stroke is the current term used to describe a symptomatic cerebral infarct for which no probable cause is identified following an adequate diagnostic evaluation (2,4). Initial work-up typically includes echocardiography, a 12-lead electrocardiogram (ECG), inpatient cardiac telemetry, or 24-h Holter monitoring, laboratory screening for hypercoagulable states (more often performed in patients younger than 55 years of age), and magnetic resonance imaging or computed tomography imaging of the brain and vasculature of the head and neck (2). Cryptogenic strokes are estimated to account for 20% to 30% of all ischemic strokes, equivalent to approximately 300,000 cases annually in North America and Europe (5).
Critics argue that no universally accepted definition exists for cryptogenic stroke, and diagnostic criteria have not been standardized. In response to this, Hart et al. (5,6) have proposed a similar clinical construct termed embolic stroke of undetermined source (ESUS), to identify patients with embolism as the likely stroke mechanism. ESUS is defined as a nonlacunar brain infarct without an identified cardioembolic source or occurring secondary to occlusive atherosclerosis. Diagnostic criteria for ESUS are specific and include the following: 1) nonlacunar brain infarct on imaging; 2) <50% arterial stenosis proximal to the infarct; and 3) no major-risk cardioembolic source (including no permanent or paroxysmal AF diagnosed by ECG). Given the ambiguity in the definition of cryptogenic stroke, the concept of ESUS may become more widely adopted in both future clinical practice and research. For the purpose of this review, however, cryptogenic stroke is the preferred term because it is currently better recognized in the cardiology and stroke community.
Subclinical AF is variably defined as AF detected in asymptomatic individuals without a prior diagnosis by ECG (7). Through improvements in ambulatory ECG (AECG) monitoring, it has become increasingly evident that subclinical AF is strongly associated with cryptogenic stroke (7–9). Given that newly diagnosed AF after stroke carries a higher risk of recurrent stroke, detection of subclinical AF following cryptogenic stroke has the potential to change management. In patients with clinical AF, treatment with an oral anticoagulant agent (OAC) such as warfarin or a direct oral anticoagulant agent (DOAC) has been shown to be superior to aspirin or no therapy in primary and secondary prevention of stroke (1,10,11). A similar benefit for starting OAC for patients with subclinical AF has not yet been demonstrated.
Because of the intermittent nature of paroxysmal AF and often poor correlation between symptoms and episodes, subclinical AF can be difficult to detect and diagnose (12). The development of long-term ECG monitoring through insertable cardiac monitors (ICMs) has greatly improved the ability to detect AF after cryptogenic stroke, evidenced most notably in the landmark CRYSTAL-AF (Cryptogenic Stroke and Underlying Atrial Fibrillation) study (13) (Central Illustration). ICMs have also proven useful in patients with recurrent unexplained syncope and carry Class I and II recommendations for this indication from major society guidelines (14,15).
Despite a growing body of evidence on the incidence of AF detection in patients who have had cryptogenic stroke, current stroke and cardiology society guidelines are lacking in recommendations for monitoring of subclinical AF, including the optimal timing from stroke event, duration, and method of AECG monitoring recommended. This review seeks to outline the current use of ICM for syncope and AF detection, summarize the latest guidelines and evidence, and describe future implications of ICM use specifically for detection of subclinical AF in patients with cryptogenic stroke.
Review of Insertable Cardiac Monitoring Technology
There are multiple modalities now available for AECG monitoring, and many have been studied for use in detection of subclinical AF among patients with cryptogenic stroke and in unexplained syncope (11,16–18). Most of the external devices available, including Holter, event, and mobile cardiac telemetry monitors, may have limited utility in cases of rare yet recurrent events because of inadequate surveillance duration and potential intermittent monitoring patterns. In contemporary practice, small, implantable devices can monitor and record the heart rhythm for several years (19–21). With the capability of long-term continuous monitoring, the ICM has been shown to be a sensitive method of detection for subclinical AF, unexplained syncope, and palpitations.
ICMs, also known as implantable loop recorders, are small devices requiring a minor invasive procedure for implantation in the subcutaneous tissue. Most contemporary models weigh <3 g, are one-third the size of an AAA battery, and offer continuous monitoring with a life span up to 3 years (17,19–21). Devices are inserted near the left 4th intercostal space corresponding to the V2-V3 ECG lead location, and an ECG tracing is measured between 2 electrodes at the ends (Figure 1). The implantable monitors available at the time of this writing are the Reveal XT and Reveal LINQ models (Medtronic, Minneapolis, Minnesota) (Figures 2A and 2B), the SJM Confirm and Confirm Rx models (St. Jude Medical, Saint Paul, Minnesota) (Figures 2C and 2D), and the BioMonitor2 model (Biotronik, Berlin, Germany) (Figures 2E and 2F).
ICMs are capable of recording the cumulative AF burden, as well as storing a fixed number of ECG waveforms. The latest models are now capable of wirelessly transmitting device data and ECG waveforms automatically by cell phone technology to the clinician’s inbox for review. Automated AF detection algorithms have been developed using R-R wave interval variability over 2-min periods (22). Advanced algorithms have added p-wave detection to improve specificity for AF (as in the Reveal LINQ model) (23). Small, industry-funded studies have demonstrated favorable performance in the identification of AF when compared with the gold standard of Holter monitoring with expert adjudication of events (24,25). Published reports on sensitivity, specificity, positive predictive value, and negative predictive value using an AF duration analysis were as follows: 98.1%, 98.5%, 91.9%, and 99.7% (Medtronic Reveal XT) and 83.9%, 99.4%, 97.3%, and 98.5% (St. Jude Confirm), respectively. The BioMonitor2 master study is pending publication; however, preliminary data from the manufacturer (Biotronik) report sensitivity, specificity, positive predictive value, and negative predictive value of 93.6%, 99.2%, 93.4%, and 99.3%, respectively, when using a similar AF duration-based analysis (Master Study of the Insertable Cardiac Monitor BioMonitor 2; NCT02565238).
ICM monitoring does have certain limitations, including a significant cost of the device and procedure (estimated initial cost $4,000), with additional ongoing monitoring costs. The need for a small, invasive procedure, although unlikely to cause significant harm to patients, can result in rare complications. The RIO (Reveal In-Office) and RIO 2 (Reveal LINQ In-Office 2) studies demonstrated a favorable safety profile of ICM device insertion, with the most common (albeit rare) complications being incision site hemorrhage and device dislocation (26,27). RIO 2 investigators reported procedure- or device-related complication rates of <1% for Reveal LINQ ICM devices inserted in either the hospital or office setting, a finding suggesting that this procedure can be safely performed in the clinic (27). Despite good sensitivity, the use of AF detection algorithms has the potential to lead to an elevated number of false-positive results because of motion and myopotential artifacts, as well as misclassification from ectopic beats, necessitating that a physician personally review all possible AF tracings to ensure an accurate diagnosis. Given the large numbers of patient data stored on these devices, there is the potential for data overload from a clinician perspective (28).
Indications for Insertable Cardiac Monitors
Established indications for ICM use in current practice include unexplained syncope, palpitations, and management of AF (22). With regard to AF management, ICMs have been studied for AF monitoring in rhythm control strategies, for AF monitoring following catheter ablation, and for detection of subclinical AF after cryptogenic stroke. The primary focus of this review is detection of subclinical AF after cryptogenic stroke, and this indication is discussed in greatest detail.
Guidelines for Insertable Cardiac Monitors in Syncope
The following organizational guidelines discuss the use of ICMs for evaluation of syncope: 1) 2017 International Society for Noninvasive and Holter Electrocardiology (ISHNE)/Heart Rhythm Society (HRS) expert consensus statement on ambulatory ECG and external cardiac monitoring (29); 2) 2017 American College of Cardiology (ACC)/American Heart Association (AHA)/Heart Rhythm Society (HRS) guideline for the evaluation and management of patients with syncope (14); 3) 2009 European Society of Cardiology (ESC)/European Heart Rhythm Association (EHRA) guideline for the diagnosis and management of syncope (15). These guidelines are summarized in Table 1 (14,15,29).
All guidelines are in general agreement that the decision to use an ICM for syncope is highly dependent on patients’ characteristics, frequency of syncopal events, and pre-test probability of arrhythmic cause. The ACC/AHA/HRS syncope guidelines provide a Class IIa, Level of Evidence: B-R recommendation for ICM use in selected patients with syncope of suspected arrhythmic origin. The joint ESC/EHRA syncope guidelines provide Class I, Level of Evidence: B recommendations for ICM use in both: 1) early phase of evaluation in patients without high-risk features; and 2) high-risk patients after negative findings of a preliminary evaluation.
Evidence for Insertable Cardiac Monitors in Syncope
Unexplained syncope is defined by the ACC as syncope for which a cause is undetermined after initial evaluation, including but not limited to a thorough history, physical examination, and ECG (14). The goals of AECG monitoring in the evaluation of unexplained syncope are to capture and identify any bradyarrhythmias, conduction block, or tachyarrhythmias that correlate with symptoms (29). Following initial diagnostic work-up with 12-lead ECG and inpatient telemetry, ICMs may be used for syncope evaluation in patients with recurrent yet infrequent symptoms when other diagnostic test results have been inconclusive (22). Several randomized controlled trials and observational studies have demonstrated utility of ICMs in the diagnosis of unexplained syncope (14). The RAST (Randomized Assessment of Syncope Trial) was among the first randomized controlled trials to compare ICM versus conventional testing in these patients (30). Investigators randomized 60 patients with unexplained syncope to ICM versus conventional testing (including external loop recorder, tilt testing, and electrophysiological testing) and followed them for a mean of 10.5 months. Diagnosis was made in 55% in the ICM arm compared with 19% in the control (p = 0.0014). The multicenter prospective PICTURE (Place of Reveal in the Care Pathway and Treatment of Patients with Unexplained Recurrent Syncope) study used ICMs in 570 patients with unexplained syncope and followed them for a mean of 10 months (31). ICMs assisted in diagnosis of syncope in 30% of these patients. A 2016 systematic review and meta-analysis including 4 studies and 579 patients found that patients who underwent ICM implantation experienced higher rates of diagnosis (relative risk: 0.61; confidence interval [CI]: 0.54 to 0.68) (18). A cost-analysis study by Krahn et al. (32) showed that an ICM strategy was more expensive per participant, but the cost was lower per diagnosis when compared with conventional testing in cases of recurrent unexplained syncope.
The ISSUE-3 (Third International Study on Syncope of Uncertain Etiology) trial investigators implanted ICMs in 511 patients with more than 3 episodes of unexplained syncope within 2 years (33). Patients with recurrence of syncope and device-detected asystole for longer than 3 s, or asystole for longer than 6 s without syncope, received dual-chamber pacemakers and were randomized to pacing with rate-drop response or to sensing only. Risk of syncope recurrence in patients who had pacing was reduced by 57% (CI: 4 to 81). In the recently presented SPRITELY (Syncope: Pacing or Recording in the Later Years) trial, investigators randomized 115 patients with at least 1 episode of syncope and with bifascicular block to either empirical pacemaker implantation or ICM implantation for further rhythm monitoring (34). After mean follow-up time of 30 months, the empirical pacemaker group experienced lower primary composite outcome rates of death, syncope, symptomatic bradycardia, asymptomatic actionable bradycardia, and device complications (63% event-free survival rate) compared with the ICM group (22% event-free survival rate) with a p value <0.001. In the ICM group, 59% of patients crossed over to receive a pacemaker. These key studies are summarized in Table 2 (18,30–34).
AECG monitoring has an important role in the work-up of palpitations, where studies have demonstrated that initial history, physical examination, and 12-lead ECG are nondiagnostic in roughly two-thirds of patients (29). The 2017 ISHNE/HRS expert consensus statement recommended a range from 24 to 48 h to 2 weeks of AECG monitoring for unexplained palpitations, depending on symptom frequency; therefore, ICMs are not routinely indicated (29). The 2009 EHRA position paper on indications for the use of diagnostic implantable and external ECG loop recorders provided a Class IIA recommendation for ICM use in select cases of severe infrequent palpitations when other ECG monitoring systems fail to document cause (35).
Guidelines for Monitoring for Atrial Fibrillation After Stroke
The following organizational guidelines discuss AECG monitoring for AF after stroke: 1) 2017 ISHNE/HRS expert consensus statement on ambulatory ECG and external cardiac monitoring or telemetry (29); 2) 2016 ESC/EHRA/European Stroke Organization (ESC/EHRA/ESO) guidelines for the management of AF (36); 3) 2014 AHA/American Stroke Association (AHA/ASA) guidelines for the prevention of stroke in patients with stroke and transient ischemic attack (TIA) (10); 4) 2014 Canadian Stroke best practice guidelines for the secondary prevention of stroke (37); and 5) 2014 Canadian Cardiovascular Society focused update of guidelines for the management of AF (38). The 2014 AHA/ACC/HRS guidelines for the management of patients with atrial fibrillation summarize the evidence behind subclinical AF and stroke but do not further discuss ECG monitoring for AF detection in patients with cryptogenic stroke (1). The guideline recommendations are summarized in Table 3 (10,29,36–38).
As the most contemporary guideline, the 2017 ISHNE/HRS expert consensus statement offers 1 of the strongest endorsements for the use of ICMs, by stating that external AECG monitoring may be limited by noncompliance and an [ICM] may be more effective, citing the 6-fold increased detection rate of AF >30 s at 6 months of follow-up as reported in the CRYSTAL-AF study. The official guideline favors “extended” AECG monitoring in patients with cryptogenic stroke and undiagnosed AF; however, it does not specify duration or method. The 2016 ESC/EHRA/ESO guidelines recommend consideration of “long-term” monitoring with either noninvasive ECG monitors or ICMs to document silent AF in stroke patients. These guidelines do not specify duration of monitoring. The 2017 ISHNE/HRS expert consensus and the 2016 ESC/EHRA/ESO guidelines were published after the release of the CRYSTAL-AF study results. The 2014 AHA/ASA guidelines offer the most specific recommendations, suggesting “prolonged, 30-day rhythm monitoring” within 6 months of acute ischemic stroke or TIA with no apparent cause. This guideline does not specify a method of monitoring. The 2014 Canadian stroke best practice recommendations suggest “prolonged” ECG monitoring to detect paroxysmal AF in cases where the initial ECG or 24 to 48 h of ECG monitoring does not show AF but a cardioembolic mechanism is suspected. The Canadian stroke guidelines specify selecting patients who are potential candidates for anticoagulation, but they do not further comment on duration or method of monitoring. The 2014 Canadian Cardiovascular Society guidelines, published after CRYSTAL-AF, also suggest “additional” ambulatory monitoring beyond 24 h for AF detection for “selected older patients” with cryptogenic stroke who would be candidates for OAC therapy. This guideline does not further specify a method of monitoring. Currently, there is lack of expert consensus in North America and Europe regarding the optimal timing, duration, and method of cardiac monitoring for occult AF in stroke patients.
Evidence for Subclinical Atrial Fibrillation and Stroke Risk
Studies have shown that even brief episodes of subclinical AF are associated with ischemic stroke; however, a causal relationship between subclinical AF and the risk of stroke has yet to be established. Much of the evidence for subclinical AF and stroke risk comes from observational studies of atrial tachyarrhythmia detection in patients with existing pacemaker or defibrillator devices without a history of stroke or TIA (7,38,39). A subgroup analysis of patients in the MOST (Atrial Diagnostics Ancillary Study of the Mode Selection) trial showed that patients with asymptomatic pacemaker-detected atrial high rate events (AHRE) lasting at least 5 min had a 6 times higher risk of AF and more than 2 times higher risk of composite endpoint of stroke and death (40). The ASSERT (Asymptomatic Atrial Fibrillation and Stroke Evaluation in Pacemaker Patients and the Atrial Fibrillation Reduction Atrial Pacing Trial) investigators found that device-detected episodes of AHRE >190 beats/min for >6 min were associated with a more than 5 times increased risk of AF and a more than 2-fold increased yearly rate of ischemic stroke or systemic embolism (7).
In a secondary analysis of patients in ASSERT, Van Gelder et al. (41) studied the effect of subclinical AF duration on subsequent stroke risk using time-dependent covariate Cox models. These investigators found that subclinical AF duration longer than 24 h was associated with a significant increase in risk of stroke (hazard ratio: 3.24; 95% CI: 1.5 to 6.95; p = 0.003). Patients with subclinical AF duration <24 h had no statistically significant difference in stroke risk compared with patients with no evidence of subclinical AF.
Mahajan et al. (42) performed a systematic review and meta-analysis including 11 studies, analyzing subclinical device-detected AF and stroke risk. The cutoff point for significant subclinical AF duration that was used to predict stroke risk varied among studies (from >6 min to 1 h to >5.5 h total daily burden). Notable results of this analysis included the following: 1) subclinical AF was strongly associated with an almost 6-fold risk of future clinical AF; 2) patients with subclinical AF duration meeting individual study cutoffs had a 2.4-fold increased risk of stroke, with an overall absolute annual risk of 1.89 per 100 person-years; 3) patients with subclinical AF and a mean CHADS2 (congestive heart failure, hypertension, age, diabetes, stroke) score of 2.1 had an estimated annual stroke rate of 2.76 per 100 person-years; and 4) short episodes of subclinical AF not meeting individual study cutoffs were associated with a lower stroke risk of 0.93 per 100 person-years.
Taken together, these studies suggest that subclinical AF, and especially AF of longer duration, is associated with an increased risk of ischemic stroke, but they do not prove causality. More data are needed to quantify what amount of subclinical AF may be clinically significant and demonstrate that detection and treatment of this condition prevents strokes.
Detection of Subclinical Atrial Fibrillation in Cryptogenic Stroke
Growing numbers of AECG modalities are available and have been studied for detection of subclinical AF in patients with cryptogenic stroke. A discussion and comparison of the various external devices can be found in many of the contemporary review articles of both AECG monitoring and cryptogenic stroke (11,16,17). This section focuses specifically on the evidence for ICMs in the detection of subclinical AF in cryptogenic stroke. Several small, prospective studies evaluated AF detection in patients with cryptogenic stroke with ICMs before the release of CRYSTAL-AF (43–48). These results are summarized in Table 4 (13,43–50).
The landmark CRYSTAL-AF trial was among the first randomized controlled trials in the study of subclinical AF in patients with cryptogenic stroke and the largest study of ICMs at the time. CRYSTAL-AF randomized 441 patients with cryptogenic stroke or TIA in a 1:1 ratio of ICM versus control (standard of care monitoring) for detection of AF (13). Primary and secondary endpoints were time to detection of AF lasting 30 s or longer at 6 and 12 months, respectively. ICMs detected AF in 8.9% of patients by 6 months and in 12.4% of patients by 12 months, compared with 1.4% and 2.0%, respectively, in the control group. A follow-up study of the CRYSTAL-AF ICM cohort found a cumulative AF detection rate of 21.1% at 24 months and 30.0% at 3 years in the ICM arm versus 3.0% and 3.0%, respectively, in the control arm (49). The median time to AF detection in the ICM arm was 84 days in the original study and 8.4 months in the 3-year follow-up study. OAC therapy was prescribed for 94.7%, 96.6%, and 90.5% of ICM-treated patients with detectable AF at 6, 12, and 36 months, respectively. These results are summarized in Table 4.
In response to the CRYSTAL-AF study results, a prospective observational study was initiated by Ziegler et al. (50) to investigate the incidence of subclinical AF in a real-world group of patients who had cryptogenic stroke. Investigators used a database of more than 1,200 patients (Medtronic Reveal LINQ Insertable Cardiac Monitor Registry) who received an ICM for the purpose of AF detection following cryptogenic stroke and were monitored for up to 2 years of follow-up. The primary endpoint was adjudicated AF lasting longer than 2 min. The rate of AF detection at 1, 6, 12, and 24 months was 4.6%, 12.2%, 16.3%, and 21.5%, respectively, with a median time to AF detection of 112 days. These results are summarized in Table 4.
The results of CRYSTAL-AF and Ziegler et al. (50) have shown that subclinical AF in a cryptogenic stroke group is much more prevalent than previously thought. AF detection rates were remarkably similar between these 2 studies. At 24 months after cryptogenic stroke, 1 of every 5 patients monitored with ICM had a diagnosis of AF. The median time to AF detection in both studies (84 days in the original CRYSTAL-AF study and 112 days in the study by Ziegler et al. ) was past the 30 days of monitoring suggested by the current ASA/AHA guidelines. Of patients who ultimately had a diagnosis of AF, 88% of those in CRYSTAL-AF and 79% of patients in the study by Ziegler et al. (50) would have been missed if monitoring had been performed for only up to 30 days. In light of these findings, expert panels have suggested consideration of longer-term monitoring with an ICM if initial monitoring results are negative in the first 30 days (17).
Significant limitations of these data are worth mentioning. First is the lack of proven clinical benefit of higher rates of subclinical AF detection and subsequent OAC use to reduce stroke risk in these studies. Neither CRYSTAL-AF nor Ziegler et al. (50) showed statistically significant reductions in stroke or systemic embolism in the ICM and OAC trial arms. Second is the lack of temporal association between AF detection by ICM and subsequent ischemic stroke. The median time to AF detection after cryptogenic stroke of 84 and 112 days, respectively, in CRYSTAL-AF and the study by Ziegler et al. (50) does not prove causality or delineate a definitive temporal association between the two.
Detection of Subclinical Atrial Fibrillation in Patients at High Risk for Atrial Fibrillation and Stroke
With growing knowledge from studies of subclinical AF in patients with pacemakers or implantable cardioverter-defibrillators and cryptogenic stroke, investigators have now begun studying the incidence of AF in patients at high risk for AF and stroke by using ICMs. The PREDATE-AF (Predicting Determinants of Atrial Fibrillation or Flutter for Therapy Elucidation in Patients at Risk for Thromboembolic Events) study enrolled 245 subjects with no history of AF and a CHA2DS2-VASc (congestive heart failure, hypertension, age, diabetes, stroke, vascular disease, sex) score of 2 or greater to be screened for new onset AF with an ICM (51). The primary endpoint was adjudicated AF lasting longer than 6 min, and subjects were followed for 18 months. Subclinical AF was detected in 22.4% of patients with a mean time to detection of 141 days. Investigators compared this with the range of overall AF prevalence in the established literature (3.0% to 7.5%). A total of 76.4% of patients with newly diagnosed AF were prescribed OAC. In light of these results, the investigators suggest that the use of ICMs for AF screening may be appropriate in subjects with a CHA2DS2-VASc score of 2 or greater.
The ASSERT-II trial implanted ICMs to define the prevalence of subclinical AF in 256 patients >65 years of age without a history of AF, with an average follow-up of 16 months (52). Eligibility also required at least 1 of the following: CHA2DS2-VASc score of 2 or greater, sleep apnea, body mass index >30 kg/m2, left atrial enlargement, or increased N-terminal pro–B-type natriuretic peptide. The primary endpoint was adjudicated AF lasting more than 5 min. In this overall group, subclinical AF was detected in 34.4% per year and in 39.4% per year in patients with a history of stroke, TIA, or systemic embolism. The mean time to AF detection was 5.1 months, and 74% of these patients were prescribed OACs. Investigators concluded that detection of subclinical AF in older patients with cardiovascular risk factors is relatively common after a period of long-term monitoring with an ICM. These investigators suggested the need for randomized clinical trials of anticoagulation in patients with recent stroke who have detectable subclinical AF.
The REVEAL AF trial investigators implanted ICMs to define the incidence of AF in 385 patients with a CHA2DS2-VASc score of 2 or greater with an average follow-up of 22.5 months (53). The primary endpoint was adjudicated AF lasting longer than 6 min. AF detection rates at 1, 6, 12, 24, and 30 months were 6.2%, 20.4%, 27.1%, 33.6%, and 40.0%, respectively. Median time to first AF episode detection was 123 days. OAC therapy was prescribed for 56.3% of patients overall. Of patients who had a diagnosis of AF in this study, 85% would have been missed if they had been monitored for only 30 days. The investigators of this trial stressed the importance of ongoing OAC trials for subclinical AF to direct management of these patients further.
The following 3 recent trials, as summarized in Table 5 (51–53), demonstrate a substantial incidence of undiagnosed subclinical in patients with AF and multiple risk factors for AF and stroke. Much as in patients with cryptogenic stroke who were monitored with ICMs, the average time to AF detection in these high-risk patients was well past the 30-day cutoff, a finding suggesting a future role for ICMs beyond existing traditional external monitoring strategies. Increasing use of cardiac implantable devices such as ICMs for AECG monitoring is likely to increase established knowledge of the epidemiology and natural history of AF. The overall prevalence of any form of AF in the population is likely much higher than previously thought. Higher prevalence of subclinical AF may not correlate with higher stroke rates, and therefore OAC therapy may not always be necessary. Further studies of OAC for subclinical AF are needed.
A proposed flow diagram has been included for ICM device implantation in select patients with cryptogenic stroke on the basis of high-quality evidence discussed in this review and summarized in Tables 3, 4, and 5 (Figure 3).
An association between subclinical AF and cryptogenic stroke exists, but no causality has been proven. Questions that need answering include the following: 1) What cutoff of (ICM or other) device-detected duration of subclinical AF is clinically significant? 2) Can the use of OAC therapy in patients with subclinical AF (detected by ICM) demonstrate a reduction in stroke rates? Further studies are needed to better understand to what extent, if any, subclinical AF is similar to clinical AF.
Many future studies relating to ICMs for subclinical AF are currently enrolling, with future results that will add to current knowledge. Some involve the use of ICMs to determine the prevalence of subclinical AF in different stroke groups. Others are studying OAC therapy for the primary and secondary prevention of stroke in patients with ICM-detected subclinical AF. Of particular relevance to ICMs are the forthcoming ARTESiA (Apixaban for the Reduction of Thrombo-Embolism in Patients With Device-Detected Sub-Clinical Atrial Fibrillation), LOOP (Atrial Fibrillation Detected by Continuous ECG Monitoring; NCT02036450), NOAH (Non-vitamin K antagonist Oral anticoagulants in patients with Atrial High rate episodes, and STROKE AF (Rate of Atrial Fibrillation Through 12 Months in Patients With Recent Ischemic Stroke of Presumed Known Origin; NCT02700945) trials, as summarized in Table 6 (54,55). ARTESiA is studying the use of OAC in the primary prevention of stroke in patients with subclinical AF detected by long-term continuous monitoring with an ICM, pacemaker, or implantable cardioverter-defibrillator (54). LOOP is studying the use of ICMs and subsequent OAC therapy for new AF in the primary prevention of stroke in patients with significant risk factors for stroke. NOAH is comparing edoxaban versus current standard of care in the primary prevention of stroke and systemic embolism in patients without AF but with AHRE and at least 2 stroke risk factors (55). STROKE AF is studying ICMs for the detection of subclinical AF in patients with noncryptogenic acute ischemic strokes. The results of these clinical trials will provide further guidance in the use of ICMs for detection and management of detected subclinical AF in the context of stroke.
Long-term continuous monitoring with ICMs in both clinical trials and real-world patients has significantly increased our ability to detect subclinical AF and aid in the diagnosis of unexplained syncope. Contemporary evidence suggests that a longer duration of monitoring may be needed to detect AF in selected patients after cryptogenic stroke than supported by current guidelines, and future studies should inform the guidelines on use of ICMs in these patients.
Dr. Hsu has received consulting and speaking honoraria from Medtronic, St. Jude Medical, Boston Scientific, and Biotronik; and has received research grants from Biosense Webster and Biotronik. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
All authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.
- Abbreviations and Acronyms
- American College of Cardiology
- ambulatory electrocardiogram
- atrial fibrillation
- American Heart Association
- atrial high rate event
- congestive heart failure, hypertension, age, diabetes, stroke, vascular disease, sex
- confidence interval
- direct oral anticoagulant agent
- European Heart Rhythm Association
- European Society of Cardiology
- European Stroke Organization
- embolic stroke of undetermined source
- Heart Rhythm Society
- insertable cardiac monitor
- International Society for Noninvasive and Holter Electrocardiology
- oral anticoagulant agent
- transient ischemic attack
- Received April 23, 2018.
- Revision received May 31, 2018.
- Accepted June 4, 2018.
- 2018 American College of Cardiology Foundation
- January C.T.,
- Wann L.S.,
- Alpert J.S.,
- et al.
- ↵American Heart Association. Heart and Stroke Statistics. Available at: https://www.heart.org/HEARTORG/General/Heart-and-Stroke-Association-Statistics_UCM_319064_SubHomePage.jsp#. Accessed June 12, 2018.
- Adams H.P.,
- Bendixen B.H.,
- Kappelle L.J.,
- et al.
- Hart R.G.,
- Catanese L.,
- Perera K.S.,
- Ntaios G.,
- Connolly S.J.
- Kishore A.,
- Vail A.,
- Majid A.,
- et al.
- Kernan W.N.,
- Ovbiagele B.,
- Black H.R.,
- et al.
- Garimella R.S.,
- Chung E.H.,
- Mounsey J.P.,
- Schwartz J.D.,
- Pursell I.,
- Gehi A.K.
- Shen W.K.,
- Sheldon R.S.,
- Benditt D.G.,
- et al.
- Andrade J.G.,
- Field T.,
- Khairy P.
- Albers G.W.,
- Bernstein R.A.,
- Brachmann J.,
- et al.
- Solbiati M.,
- Costantino G.,
- Casazza G.,
- et al.
- ↵Medtronic. Reveal LINQ Insertable Cardiac Monitoring System. Available at: http://www.medtronicdiagnostics.com/us/cardiac-monitors/reveal-linq/index.htm. Accessed June 12, 2018.
- ↵St. Jude Medical/Abbott. Implantable Cardiac Monitors. Available at: https://www.sjm.com/en/professionals/featured-products/electrophysiology/recording-and-monitoring/implantable-cardiac-monitor. Accessed June 12, 2018.
- ↵Biotronik. BioMonitor 2. Available at: https://www.biotronik.com/en-us/products/crm/arrhythmia-monitoring/biomonitor-2. Accessed June 12, 2018.
- Ciconte G.,
- Giacopelli D.,
- Pappone C.
- Hindricks G.,
- Pokushalov E.,
- Urban L.,
- et al.
- Pachulski R.,
- Cockrell J.,
- Solomon H.,
- Yang F.,
- Rogers J.
- Rogers J.D.,
- Sanders P.,
- Piorkowski C.,
- et al.
- Zimetbaum P.,
- Goldman A.
- Steinberg J.S.,
- Varma N.,
- Cygankiewicz I.,
- et al.
- Krahn A.D.,
- Klein G.J.,
- Yee R.,
- Skanes A.C.
- Krahn A.D.,
- Klein G.J.,
- Yee R.,
- Hoch J.S.,
- Skanes A.C.
- Brignole M.,
- Menozzi C.,
- Moya A.,
- et al.
- ↵Sheldon RS, Talajic M, Becker G, et al. A randomized pragmatic trial of strategies of permanent pacemaker versus implantable cardiac monitor in older patients with syncope and bifascicular block. Presented at: Heart Rhythm Society Annual Scientific Sessions; May 10, 2018; Boston, MA.
- Glotzer T.V.,
- Daoud E.G.,
- Wyse D.G.,
- et al.
- Glotzer T.V.,
- Hellkamp A.S.,
- Zimmerman J.,
- et al.
- Mahajan R.,
- Perera T.,
- Elliott A.D.,
- et al.
- Ritter M.A.,
- Kochhäuser S.,
- Duning T.,
- et al.
- Etgen T.,
- Hochreiter M.,
- Mundel M.,
- Freudenberger T.
- Brachmann J.,
- Morillo C.A.,
- Sanna T.,
- et al.
- Ziegler P.D.,
- Rogers J.D.,
- Ferreira S.W.,
- et al.
- Healey J.S.,
- Alings M.,
- Ha A.,
- et al.
- Lopes R.D.,
- Alings M.,
- Connolly S.J.,
- et al.
- Kirchhof P.,
- Blank B.F.,
- Calvert M.,
- et al.
- Central Illustration
- Review of Insertable Cardiac Monitoring Technology
- Indications for Insertable Cardiac Monitors
- Guidelines for Insertable Cardiac Monitors in Syncope
- Evidence for Insertable Cardiac Monitors in Syncope
- Guidelines for Monitoring for Atrial Fibrillation After Stroke
- Evidence for Subclinical Atrial Fibrillation and Stroke Risk
- Detection of Subclinical Atrial Fibrillation in Cryptogenic Stroke
- Detection of Subclinical Atrial Fibrillation in Patients at High Risk for Atrial Fibrillation and Stroke
- Future Directions
- Future Studies