Subclinical atrial fibrillation detection with a floating atrial sensing dipole in single lead implantable cardioverter‐defibrillator systems: Results of the SENSE trial

Abstract Introduction Subclinical atrial fibrillation (AF), in the form of cardiac implantable device‐detected atrial high rate episodes (AHREs), has been associated with increased thromboembolism. An implantable cardioverter‐defibrillator (ICD) lead with a floating atrial dipole may permit a single lead (DX) ICD system to detect AHREs. We sought to assess the utility of the DX ICD system for subclinical AF detection in patients, with a prospective multicenter, cohort‐controlled trial. Methods and Results One hundred fifty patients without prior history of AF (age 59 ± 13 years; 108 [72%] male) were enrolled into the DX cohort and implanted with a Biotronik DX ICD system at eight centers. Age‐, sex‐, and left ventricular ejection fraction‐matched single‐ and dual‐chamber ICD cohorts were derived from a Cornell database and from the IMPACT trial, respectively. The primary endpoint were AHRE detection at 12 months. During median 12 months follow‐up, AHREs were detected in 19 (13%) patients in the DX, 8 (5.3%) in the single‐chamber, and 19 (13%) in the dual‐chamber cohorts. The rate of AHRE detection was significantly higher in the DX cohort compared to the single‐chamber cohort (P = .026), but not significantly different compared to the dual‐chamber cohort. There were no inappropriate ICD therapies in the DX cohort. At 12 months, only 3.0% of patients in the DX cohort had sensed atrial amplitudes less than 1.0 mV. Conclusion Use of a DX ICD lead allows subclinical AF detection with a single lead DX system that is superior to that of a conventional single‐chamber ICD system.

speakers fees from Zoll and Janssen Pharmaceuticals. Dr Mainigi reports speakers fees from Abbott and Biotronik. Dr Lubitz reports research support from Bristol Myers Squibb/ Pfizer, Bayer HealthCare, and Boehringer Ingelheim, and consulting fees from Abbott, Quest Diagnostics, Bristol Myers Squibb/Pfizer, and is supported by NIH grant 1R01HL139731 and American Heart Association grant 18SFRN34250007. Dr Cheung reports research support from Biotronik, consulting fees from Abbott and Biotronik, and fellowship support from Abbott, Biotronik, Boston Scientific, and Medtronic.

Funding information
Biotronik USA, Lake Oswego, OR Conclusion: Use of a DX ICD lead allows subclinical AF detection with a single lead DX system that is superior to that of a conventional single-chamber ICD system.

K E Y W O R D S
implantable cardioverter-defibrillator, remote monitoring, subclinical atrial fibrillation 1 | INTRODUCTION Implantable cardioverter-defibrillators (ICDs) are a cornerstone therapy for the prevention of sudden cardiac death by treating lifethreatening ventricular arrhythmias. In randomized trials establishing the benefit of ICDs for the sudden cardiac death prevention, the majority of patients received single-chamber ICDs. 1 However, according to data from the National Cardiovascular Data Registry (NCDR) and other international registries, more patients receive dual-chamber ICDs than single-chamber ICDs for primary prevention. [1][2][3] This remains the case despite the fact that 60% of patients who receive dual-chamber ICDs for primary prevention of sudden cardiac death do not have an atrial pacing indication. 4 Multiple studies have shown that implantation of dual-chamber ICDs is associated with higher rates of procedural complications, in-hospital mortality, and need for earlier generator change. 1,3,4 The diagnostic capabilities offered by the presence of atrial intracardiac electrograms (EGMs), however, can provide benefits.
Some studies comparing outcomes of dual-vs single-chamber sensing show reduction of inappropriate ICD therapies with dual-chamber sensing due to supraventricular tachycardia discrimination, although this is not a consistent finding. [5][6][7] Adoption of ICD programming strategies, utilizing higher rates and longer durations for tachycardia detection, have been more beneficial in reducing inappropriate therapies. 8,9 Nonetheless, atrial EGMs allow physicians to accurately diagnose previously undetected atrial and ventricular arrhythmias in patients with ICDs. 10 Specifically, subclinical atrial fibrillation (AF), manifest as devicedetected atrial high rate episodes (AHREs), is associated with increased risk of stroke and mortality. [11][12][13] A single lead ICD system with a floating atrial dipole may allow AHRE detection without the need for implantation of an additional atrial lead. The single lead DX ICD system (Biotronik, Berlin, Germany) utilizes a proprietary ICD lead with a floating atrial dipole 15 to 17 cm from the tip of the ICD lead and an ICD generator with specialized amplification and filtering to allow the atrial signals to be increased four-fold, while minimizing far-field R-wave sensing. In this multicenter prospective trial, we sought to assess the utility of the single lead DX ICD system for identifying AHREs in patients with no prior history of AF and comparing its rates of AHRE detection with patients in singleand dual-chamber ICD cohorts.

| SENSE prospective study population
The SENSE trial (Clinicaltrials.gov identifier NCT02186704) was initiated by study investigators, sponsored by Weill Cornell Medicine and funded by Biotronik, USA. The SENSE trial was a prospective, cohort-controlled, and eight-center study comprised of patients who met standard indications for a primary or secondary prevention ICD.
Enrollment was conducted from 29 July 2014 to 21 March 2017. All prospective study patients underwent implantation of a Biotronik DX ICD system (Biotronik) with a DX ICD generator (Ilesto 7 VR-T DX, Iforia 7 VR-T DX, Itrevia 7 VR-T DX, Inventra 7 VR-T DX, or Iperia 7 VR-T DX) and a DX ICD lead (Linoxsmart S DX 65/15 or Linoxsmart S DX 65/17). Patients were enrolled in the study either before ICD implant or within 30 days after ICD implant. Patients were excluded for: presence of atrial pacing indication, prior history of AF or atrial flutter, AHRE detection before study enrollment, atrial sensing less than 2 mV via the DX ICD system (not the pacing system analyzer), inability to comply with remote monitoring, or life expectancy of less than 1 year. The institutional review board at each center approved the study protocol and written informed consent was obtained from each patient.

| Control population
Two historical control cohorts were utilized for the study: a singlechamber ICD cohort and a dual-chamber ICD cohort. The singlechamber ICD cohort was derived from a retrospective database of patients who underwent single-chamber ICD implantation at Weill Cornell Medicine. Patients with a prior history of AF or atrial flutter or follow-up less than 1 year were excluded. The dual-chamber ICD cohort was derived from a subset of patients of the previously THOMAS ET AL. | 1995 conducted IMPACT trial (Clinicaltrials.gov identifier NCT0559988). 14 The IMPACT trial was a study of 2718 patients with Biotronik dual-chamber ICDs and biventricular ICDs who were randomized to anticoagulation treatment based on remote rhythm monitoring or office-based follow-up for the detection of AHREs. The dual-chamber control group for this study was derived from the subset of patients who had dual-chamber ICDs and had no prior history of AF or atrial flutter. Based on age-, sex-and left ventricular ejection fraction (LVEF)-matching to the prospective SENSE study patient group, 150 out of 184 patients in the single-chamber ICD database and 150 out of 1740 patients in the IMPACT study subgroup were selected to create the final control cohorts. All control group matching was performed with blinding to the presence of AF or AHRE detection. Single-chamber cohort patients had ICD programming parameters made at the discretion of the individual operators.

| Endpoints
The primary endpoint was AHRE detection lasting greater than 30 seconds by 12 months as per intention-to-treat analysis. This study was designed to compare AHRE detection between the DX cohort and the dual-chamber IMPACT control cohort and to compare AHRE detection between the DX cohort and the single-chamber control cohort. Secondary endpoints included: AHRE detection lasting greater than 6 minutes, time to detection of first AHRE, appropriate and inappropriate ICD therapies, adverse device events, and mortality. Additionally, device performance parameters such as atrial sensing, ventricular sensing, and ventricular pacing threshold in the DX cohort were assessed.  Table 1.  The rate of inappropriate AHRE detection in the DX cohort was 13% (9 of 68 total detections). Of these nine, eight were due to electromagnetic interference, and one was due to lead dislodgement.

| DX ICD system performance
In the dual-chamber cohort, the rate of inappropriate AHRE detection was 9% (6 of 68 total detections). Of these six, five were due to sinus tachycardia with far-field ventricular oversensing, and one was due to electromagnetic interference. There was 100% interobserver agreement in AHRE episode adjudication. Rate of inappropriate AHRE F I G U R E 1 Atrial and ventricular arrhythmia detection in a patient with a DX ICD system. A, Newly detected atrial fibrillation is shown with rapid, disorganized atrial activity (AEGM), and rapid ventricular response (VEGM). This episode lasted 8 minutes. B, Ventricular tachycardia is shown here in the same patient which was subsequently terminated with anti-tachycardia pacing. Clear dissociation between the atrial (AEGM) and ventricular (VEGM) electrograms is seen. AEGM, atrial electrogram; FFEGM, far-field electrogram; VEGM, ventricular electrogram THOMAS ET AL.

| ICD device therapies
In the DX cohort, 11 (7.3%) patients received 39 episodes of antitachycardia pacing therapy and 3 (2%) patients received three ICD shocks. There were no inappropriate ICD therapies, and all shocks successfully converted ventricular tachycardia. In the dual-chamber cohort, 15 (10%) patients received 46 ICD shocks at 12 months.
Inappropriate therapy data were not available for the dual-chamber cohort. In the single-chamber cohort, 10 (6.7%) patients received 21 ICD shocks, and 5 (3.3%) patients received 14 inappropriate ICD shocks.

| DISCUSSION
In this multicenter trial, we demonstrate the utility of the DX singlelead ICD system with an atrial sensing dipole for the detection of AHREs in patients with no prior history of AF. The rate of subclinical AF detection in the DX cohort was comparable to that of a dualchamber cohort and higher than clinical AF detection in a singlechamber cohort. Atrial sensing amplitudes in the DX cohort were These findings support a role for the single lead DX system for detection of subclinical AF in ICD patients who do not require atrial pacing.
Subclinical AF is common, as up to 35% of patients with cardiac implantable electronic devices are diagnosed with asymptomatic AF at 30 months. 15 High-risk ICD subgroups, such as those with hypertrophic cardiomyopathy, have been found to have a newly detected subclinical AF rate of 50% during long-term follow-up. 16 Most importantly, AHREs have been associated with increased mortality and thromboembolism. [11][12][13] Furthermore, subclinical AF can cause worsening heart failure, leading to clinical decompensation and hospitalization. 17,18 Prompt detection of AHREs in patients with ICDs with atrial sensing capabilities can facilitate early treatment of subclinical AF with anticoagulation to prevent strokes and with medical optimization to prevent heart failure exacerbation. Currently, there are no randomized clinical trial data demonstrating the benefit of oral anticoagulation treatment in patients with devicedetected AHREs. The NOAH-AFNET and ARTESIA studies are currently underway to examine the utility of anticoagulant therapy guided by the presence of device-detected AHREs. 19,20 Single-chamber pacing systems and VDD pacing systems have been associated with a lower rate of procedural complications. 21 However, the widespread use of early VDD pacing systems has been limited by the reliability of atrial sensing after implant. Only 3% of patients receiving a pacemaker for an atrioventricular block in the United States receive a VDD system and 11% of patients who received a VDD pacing system for AV block had inadequate atrial sensing in long term follow-up. 22 The current DX ICD lead has a widely spaced dipole which permits sensing of a larger 49 mm 2 atrial surface area without the need for direct contact with atrial myocardium. 23 29 Furthermore, in the DX cohort, no patient required an upgrade to a dual lead ICD system to provide atrial pacing.
Therefore, our trial points to an important place for DX ICD system implantations in patients who do not have an indication for atrial pacing but are at risk for the development of subclinical AF.

| Study limitations
The major limitation of this trial is the use of historical-control singlechamber and dual-chamber cohorts. As such, there were significant baseline differences in the clinical characteristics between the three device cohorts, which may have affected rates of subclinical AF occurrence. Of note, the CHA 2 DS 2 -VASc scores of the DX cohort were significantly lower than those of the single and dual-chamber cohorts, which would potentially suggest a lower baseline risk of AF.
Therefore, there was unlikely to be a significant bias towards increased AHRE detection in the DX cohort based on differences in baseline clinical characteristics alone. Furthermore, while patients in the single-chamber cohort had at least 1 year of follow-up, which included routine device clinic follow-up and in-office ECGs, they were not prospectively monitored for AF. Therefore, there may have been patient-to-patient variability in assessment for AF in the singlechamber cohort. In addition, single-chamber ICD systems using R-R interval variability to detect subclinical AF were not available during the study period. Therefore, more contemporary single-chamber devices using these algorithms could have increased AF detection rates. Finally, this trial was not adequately powered to show statistical non-inferiority in AHRE detection between the DX cohort and the dual-chamber cohort.

| CONCLUSION
In this multicenter, prospective cohort-controlled trial, the use of a single lead DX ICD system with an atrial sensing dipole led to rates of subclinical AF detection that were higher than that of a single-chamber ICD system and comparable to that of a dual-chamber ICD system. Atrial sensing characteristics and accuracy of AHRE detections were favorable using the DX system. The DX ICD system may offer significant benefits for AHRE detection in ICD patients who do not have an atrial pacing indication but are at high risk of developing subclinical AF.