Clinical aspects of device-detected arrhythmias

Abstract

INTRODUCTION: Cardiac implantable electronic devices (CIEDs) enable continuous monitoring of the heart rhythm. CIEDs constitute a unique opportunity for detecting arrhythmias, as the duration of cardiac monitoring is of the utmost importance for the detection rate. The CIED population consists mostly of patients from older age categories where risk factors for atrial fibrillation (AF) are common. A dual-chamber device can detect and store episodes with a high atrial rate, i.e. atrial high-rate episodes (AHREs). AHREs confirmed to be AF, atrial flutter or focal atrial tachycardia are termed subclinical AF. Both terms refer to patients with no symptoms attributed to AF, with no previous diagnosis of clinical AF. These episodes of device-detected AF are associated with increased risk of ischaemic stroke, although the risk seems to be lower than in patients with documented clinical AF, and the benefit of oral anticoagulation (OAC) treatment in this population has not been established. Patients presenting with syncope represent a diagnostic challenge. Initial evaluation can provide the underlying mechanism in up to half of the patients. However, the mechanism remains unexplained in many patients, and long-term electrocardiogram (ECG) monitoring with an implantable loop recorder (ILR) enables ECG recording at the time of syncope recurrence, which can reveal the underlying mechanism.

The aim of this thesis is to highlight different aspects of arrhythmias diagnosed with CIEDs, both from a diagnostic and a therapeutic point of view. More specifically, it aims to describe the incidence of subclinical AF/AHREs in a pacemaker population, along with its OAC treatment, and the incidence of ischaemic stroke and vascular dementia. In addition, it will explore the role of the baseline 12-lead ECG in predicting the syncope mechanism during ILR monitoring, and whether age and gender impact the evaluation before the implantation and subsequent diagnostic yield of the ILR. Finally, the thesis will test the hypothesis that patients with incident AF during inpatient care after coronary artery bypass graft (CABG) surgery often experience a relapse of AF within a year, with little chance of detection.

METHODS AND RESULTS: In study I, consecutive patients were enrolled who had been implanted with a dual-chamber device for the indication of sinus node disease or atrioventricular block/ bundle branch block between 2010 and 2014 in Halland County in Sweden. The incidence of subclinical AF/AHREs, ischaemic stroke, or vascular dementia, and the initiation of and/or any change of OAC treatment were recorded during follow-up. At inclusion, 271 patients had clinical/known AF, of which 80% (216/271) were on OAC treatment. Four hundred eleven patients had no history of AF, and of these 30% (125/411) were diagnosed with subclinical AF/AHREs during a mean follow-up of 38 months. 62% of these were prescribed OAC treatment. Patients with congestive heart failure (p= .03) and age >75 years (p= .0002) were more often diagnosed with subclinical AF/AHREs. The annual stroke incidence was 2.1% in patients with clinical/known AF, 1.9% in patients with subclinical AF/AHREs, and 1.4% in patients with no AF. Corresponding values for a diagnosis of vascular dementia was 11.2%, 5.6% (p= .09), and 6.2% (p= .048).

The study population in studies II and III consisted of consecutive patients with unexplained syncope in Halland County in Sweden, who had been selected to be implanted with an ILR after an initial non-diagnostic evaluation between 2007 and 2016. In study II, baseline 12- lead ECG was compared with clinically adjudicated cause of syncope. In study III the role of age and gender in the evaluation before implantation, and in the diagnostic yield of the ILR, was reported. There is a notable difference between the two terms ILR-guided diagnosis (study II) and ECG-based diagnosis (study III). ILR-guided diagnosis refers to all patients where the ILR has informed the clinical diagnosis, i.e. where captured ECG recordings both during syncope recurrence or other times have enabled a clinical diagnosis to be made, while ECG-based diagnosis only includes patients with syncope recurrence. In total, 300 (147 women) patients were included. The mean age was 66±16 years. In study II, 49% (146/300) received an ILR-guided diagnosis. Bifascicular block was the second most common pathological baseline 12-lead ECG finding (n=33). It was most common in patients ≥60 years of age (31/33), and more common in patients who received an ILR-guided diagnosis (bifascicular block: 25/33, 76%; normal baseline 12-lead ECG: 90/205, 44%, p< .001). Among patients with bifascicular block, 96% (24/25) were clinically adjudicated to have an arrhythmia-caused syncope, and of these, 23 had ECG recordings of a bradyarrhythmia. Bifascicular block was a strong predictor of a clinically adjudicated arrhythmia-caused syncope, with an adjusted odds ratio of 5.5 (95%CI (confidence interval) 2.3-13.2), p< .001, and a positive predictive value of 73%. In the total population, bifascicular block predicted a clinically adjudicated arrhythmia-caused syncope due to bradyarrhythmia, with an adjusted odds ratio of 11.4 (95%CI 5.0-26.2), p< .001. In study III, women experienced syncope recurrence and received an ECG-based diagnosis more often than men (women: 56/147, 38%; men: 33/153, 22%; p= .001), mainly because of a higher incidence of non-arrhythmic syncope recurrence, i.e. syncope with a normal ECG recording (women: 27/147, 18%; men: 15/153, 10%; p= .045). Patients ≥60 years of age had the lowest rate of pre-implant tests (<40 years: 6.5±1.2; 40-59 years: 5.75±1.0; and ≥60 years: 5.1±1.9; p= .002) but the highest rate of arrhythmic syncope (<40 years: 3/11, 27%; 41-59 years: 7/18, 39%; and ≥60 years: 37/60, 62%; p= .045). Fifty patients with no recurrent syncope had ECG findings potentially indicative of recurrent syncope.

Study IV was a sub-study of the prospective AFAF study (Atrial Fibrillation AFter CABG and percutaneous coronary intervention). In short, the AFAF study investigates the incidence of AF after percutaneous coronary intervention or CABG surgery by non-invasive handheld ECG recordings. It is investigated three times daily during the first postoperative month, and thereafter for two weeks at three, 12 and 24 months in addition to routine care. This sub-study added continuous ECG monitoring with an ILR. The primary endpoint was the proportion of patients with incident or recurrent AF during the 12-month monitoring period. The secondary endpoints were the proportion of patients who developed persistent AF and calculated AF burden. In total, 27/40 (68%) patients were diagnosed with incident AF, 21 in hospital and six later. Eighteen of these 27 (67%) also experienced AF recurrence, and three patients progressed into persistent AF. The incidence of AF episodes was highest during the first 30 postoperative days, as 17/40 patients had episodes of AF after discharge within this period. The rate of incident and recurrent AF after the first 30 days was low: three patients had incident AF and 10 patients recurrent AF. The CHA2DS2-VASc (Congestive heart failure, Hypertension, Age >75 years (2 points), Diabetes, Stroke (2 points), Vascular disease, Age 65-74 and Sex (female)) score was higher in patients with AF than in patients who remained in sinus rhythm: median 4 (IQR (interquartile range) 1) and median 3 (IQR 2) respectively, p= .006. In patients with paroxysmal AF, the AF burden was low: 0.1% (IQR 0.28). Handheld ECG identified fewer patients with AF after discharge than the ILR (handheld ECG: 9/20, 45%; ILR: 20/20, 100%; p= .001).

CONCLUSIONS: CIEDs are a valuable asset in arrhythmia diagnostics, and can inform clinical decisions. Subclinical AF/AHREs were common, and were associated with older age and congestive heart failure. The stroke incidence was low, but clinical/known AF was associated with an increased risk of vascular dementia. In syncope patients bifascicular block at baseline 12-lead ECG predicted a clinically adjudicated arrhythmia-caused syncope, commonly due to intermittent complete heart block. Women experienced syncope recurrence more often than men, especially for non-arrhythmic reasons. The highest rate of arrhythmic syncope and the lowest rate of pre-implant tests were found in patients ≥60 years of age. In patients treated with CABG surgery, the recurrence rate of AF was high in patients with incident AF during hospitalisation, especially during the first postoperative month. After the first month, the rate of incident and recurrent AF was low. The ILR was more effective in detecting patients with AF than handheld ECG.