Demographic and Clinical Characteristics to Predict Paroxysmal Atrial Fibrillation: Insights from an Implantable Loop Recorder Population ANTONIO FRONTERA, M.D.,* ALEXANDER CARPENTER, M.B.CH.B.,* NAUMAN AHMED, M.B.B.S.,* MATTEO FASIOLO, PH.D.,† MARTIN NELSON, S.R.,* IHAB DIAB, M.D.,* TIM CRIPPS, M.D.,* GLYN THOMAS, PH.D.,* and EDWARD DUNCAN, PH.D.* *From the Cardiology Department, Bristol Heart Institute, University Hospitals Bristol NHS Foundation Trust, Bristol, UK; and †Mathematical Sciences Department, University of Bath, Bath, UK Background: There is growing interest in detecting paroxysmal atrial fibrillation (PAF) to identify patients at high risk of thromboembolic stroke. The implantable loop recorder (ILR) is emerging as a powerful new tool in the diagnosis of PAF. Widespread implantation has significant cost implications and their use must be targeted at those patients at most risk. Methods: We retrospectively studied a population of 200 adult patients who underwent ILR implantation for the investigation of syncope or palpitations. Clinical data, baseline electrocardiogram (ECG) characteristics, and echocardiographic data were collected. All ECGs and electrograms (EGMs) were scrutinized by two blinded investigators. PAF incidence was defined as episodes lasting >30 seconds on EGMs recorded in ILR memory. Results: Our ILR population consists of 200 patients, 111 (56%) male, with a mean age of 61.4 years (range 19–95). PAF was detected in 42 patients. The following factors were significant predictors of PAF by multivariate logistic regression analysis: cigarette smoking (odds ratio [OR] = 3.73, 95% confidence interval [CI] = 1.40–10.24, P = 0.009) and incomplete right bundle branch block (IRBBB; OR = 9.04, 95% CI = 2.51–34.64, P = 0.00088). Significant differences included incidence of IRBBB (P = 0.012), cigarette smoking (P = 0.026), hypercholesterolemia (P = 0.015), age (P = 0.002), estimated glomerular filtration rate (P = 0.031), left atrial volume (P = 0.019), and PR interval (P = 0.031). The PAF group had significantly higher CHA2 DS2 -VASc scores (P = 0.01). Conclusions: Our study reports predictive factors for PAF in an ILR population. We suggest that cigarette smoking and IRBBB are independently associated with paroxysmal AF in patients presenting with palpitations or syncope. (PACE 2015; 38:1217–1222) paroxysmal atrial fibrillation, implantable loop recorder, cryptogenic stroke, right bundle branch block

Introduction Atrial fibrillation (AF) is the most common sustained arrhythmia, affecting at least 1% of the general population,1 AF is a leading cause of morbidity and mortality worldwide, with complications including thromboembolic stroke,2 cardiac failure,3 cognitive impairment,4 and a

No conflicts of interest or funding sources to disclose. Address for reprints: Antonio Frontera, M.D., EP office, Level 7, Bristol Heart Institute, Bristol BS2 8HW, UK. Fax: +44-01179299737; e-mail: [email protected] Received January 23, 2015; revised June 9, 2015; accepted July 5, 2015.

1.5- to 1.9-fold increased risk of death independent of cardiovascular disease.5 The risk of stroke in paroxysmal AF is equal to that of permanent AF.6,7 Data suggest there may be an unknown population burden of asymptomatic, undiagnosed AF with a substantial stroke risk. Indeed, paroxysmal atrial fibrillation (PAF) may explain a great proportion of strokes labeled as cryptogenic.8 Anticoagulation in PAF patients produces similar benefits to those in persistent AF.9 With up to a third of patients asymptomatic, it can be a challenge to demonstrate the presence of AF in order to inform correct diagnosis and management strategies to minimize stroke risk. The implantable loop recorder (ILR) is emerging as a powerful tool in the detection of asymptomatic PAF, which may not have been

doi: 10.1111/pace.12692

©2015 Wiley Periodicals, Inc. PACE, Vol. 38

October 2015

1217

FRONTERA, ET AL.

identified on routine 12-lead electrocardiogram (ECG) testing or 24-hour Holter monitor.10 There is, however, a paucity of research investigating how we could most effectively use the ILR to identify patients at risk of PAF and stroke. Current studies demonstrate that a significant number need to be implanted to identify each case of PAF.10 There remain little data to define the profile of patients with an increased propensity to develop PAF. The aim of this study is to identify demographic, electrocardiographic, or echocardiographic predictors of paroxysmal AF, which could be combined to produce a risk profile used to identify those at highest risk of AF and subsequent stroke, who could receive early therapeutic intervention to reduce risk of first stroke. Methods Study Population This study is a retrospective analysis of 200 patients who underwent ILR implantation from January 2010 to March 2013. All patients in the study presented to the Bristol Heart Institute arrhythmia service with unexplained syncope or palpitations with a suspected cardiac cause warranting further investigation in accordance with national best practice guidelines.11 Those with known episodes of AF were excluded from the study. ILR devices used were the R Medtronic Reveal XT (Medtronic Inc., Minneapolis, MN, USA) and the SJM ConfirmTM (St. Jude Medical, St. Paul, MN, USA). Clinical Data Demographic data were collected from the medical notes. This included age (analyzed as a continuous variable), gender, comorbidities (including hypertension, ischemic heart disease [IHD], diabetes, high cholesterol—defined as >5 mmol/L total cholesterol or >3 mmol/L LDL cholesterol—and cigarette smoking, obesity— defined as a body mass index [BMI] >30). Biochemical data including estimated glomerular filtration rate (eGFR) were collected. For each patient a CHA2 DS2 -VASc score was calculated. Baseline Analysis Data All baseline ECG and ILR data were scrutinized by two blinded, experienced electrophysiologists using manual caliper measurement. Twelvelead ECGs were recorded at baseline, using a paper speed of 25 mm/s and amplitude of 1 mV/cm. Data collected included heart rate, PR interval, QRS duration, presence of left bundle branch block (LBBB), and complete (RBBB) or incomplete right bundle branch block (IRBBB). LBBB and RBBB were defined in accordance to international

1218

criteria.12 IRBBB was defined as an rsr’, rsR’ or rSR’ pattern in leads V1 or V2 with QRS duration 200 ms. AF during manual detection or detection by the ILR was defined as one or more periods of R–R interval irregularity lasting >30 seconds in keeping with international consensus guidelines13,14 and recent landmark trials in the field.10 Ninety percent of the R loop recorders implanted were Medtronic Reveal  R XT. The Medtronic Reveal XT was programmed for automatic detection at a period of 2 minutes, in keeping with the capabilities of the device, but could record shorter episodes if activated by the patient. SJM ConfirmTM loop recorders were automated to detect AF of duration >30 seconds. ILRs were checked every 3 months in the pacing clinic or whenever patients activated the device. At these visits electrogram (EGM) data were downloaded. Echocardiographic Analysis One hundred and seventy-eight (89%) study patients had echocardiograms performed less than R 1 year before the ILR procedure on EchoPAC  (GE Healthcare, Waukesha, WI, USA), and these were reviewed by two blinded expert echocardiographers collecting data for left ventricular ejection fraction (LVEF %) and left atrial (LA) volume. Data were collected via formal reports or re-measured R offline using the EchoPAC system. Statistical Analysis Multivariate logistic regression was used to analyze the relationship between demographic, ECG and echocardiographic features, and the detection of PAF on ILR downloads. Continuous variables were reported as means ± standard deviation (SD). Continuous variables were compared with Welch’s t-test, while categorical variables were compared with Fisher’s exact test. Ordered categorical data were compared using a MannWhitney U-test. The Strengthening Reporting in Observational Studies (STROBE) guidelines15 were observed in the production of this paper. Results Our ILR population consisted of 200 adult patients, 111 (55.5%) male, with a mean age of 61.4 years (range 19–95) in whom an ILR was implanted for the investigation of syncope (65%), palpitations (17%), or both (18%). Overall group comorbidities included IHD (13.5%), hypertension (53.5%), diabetes (10%), cigarette smoking (20.5%), and hypercholesterolemia (29%). Mean BMI was 27 (SD = 4.3) and eGFR 72.5

October 2015

PACE, Vol. 38

PREDICTORS OF PAROXYSMAL ATRIAL FIBRILLATION

(SD = 15.9). Twenty-five (12.5%) patients were recorded as suffering with structural heart disease or known cardiomyopathy. PAF was detected in 42 patients (21%). Thirty-four (81%) of these patients were diagnosed via automatic ILR detection, the remainder via manual EGM evaluation. Mean age of this group was 69 (range 25–88); 29 (69%) were male. The presence of symptoms during recording did not predict AF (P = 1.00), with 26 (62%) of AF patients experiencing symptoms. Characteristics of the study population and differences between the PAF and non-PAF group are displayed in Table I. The median CHA2 DS2 -VASc score for the PAF group was 3.5 (interquartile range 2). Significant statistical differences between the PAF and non-PAF group were seen in the following factors: cigarette smoking (P = 0.026), high cholesterol (P = 0.015), age (P = 0.002), eGFR (P = 0.032), LA volume (P = 0.02), PR interval (P = 0.03), and the proportion displaying incomplete RBBB (P = 0.012). Multivariate logistic regression analysis revealed the following factors to be positive predictors of paroxysmal AF (odds ratios and confidence intervals are shown in Table II): cigarette smoking (P = 0.0089) and IRBBB (P = 0.00088; see Fig. 1) Discussion Our study investigates predictive factors of AF among a population of patients undergoing ILR implantation for palpitations or syncope. The main findings of this retrospective study were that the electrocardiographic feature of IRBBB and the cardiovascular risk factor of cigarette smoking were independent predictors for PAF, using multivariate analysis. We investigate predictors for paroxysmal AF in a population of patients with ILRs. IRBBB is a common finding within the general population with prevalence estimated at 2–7%, with male gender, increasing age, and low BMI postulated as predictive factors.16,17 It is reported as the most common abnormality found on routine ECG screening of healthy athletes, present in 13% of American college athletes.18 While there have been longstanding suggestions of RBBB as a marker for cardiovascular risk and mortality, IRBBB has long been considered a benign finding.16,17 The suggestion of IRBBB as a predictor of AF has been previously postulated in a paper by Nielsen et al.,19 specifically, lone AF. Their group conducted a case-control study comparing ECG markers in patients with early-onset lone AF to a healthy control population. They found IRBBB was strongly and independently associated with early-onset lone AF, using a single 12-lead ECG recorded at enrollment. We concur with their suggestion as to a possible explanation, that is,

PACE, Vol. 38

that intraventricular conduction block may reflect advanced myocardial disease, known to progressively alter the myocardial structure including the His–Purkinje conduction system. The IRBBB pattern may represent an early sign of fibrosis within the Purkinje system and may thus serve as a surrogate marker for widespread deterioration of the conduction system. This hypothesis does not, however, explain why complete RBBB is not a predictor of AF in our cohort. Indeed, although there is a trend toward increased AF in the complete RBBB cohort (P = 0.08), the odds ratio (OR) is unexpectedly lower (OR 3.64 vs 9.04) than in the incomplete RBBB group. The current study is unfortunately underpowered to provide further information about the role of complete RBBB in predicting AF. Cigarette smoking is thought to increase AF occurrence by promoting atrial fibrosis.20,21 Increasing age and smoking may together converge to produce progressive fibrosis of the heart which, in turn, has been associated with AF.22 The hypothesis of atrial fibrosis as an arrhythmic mechanism is supported by the increased LA volume demonstrated within the AF group, another factor thought to contribute to paroxysmal AF. Our population largely consists of individuals with structurally normal hearts (87.5%), further adding significance to our data as we are reporting the prevalence of AF in a grossly normal population. In our study IRBBB was found in 23.8% of the patients with PAF, suggesting a greater than normal proportion of PAF patients had interventricular block. Looking at the descriptive differences between the two groups within our study, other variables represented to a statistically significant greater extent within the PAF group include hypercholestorelemia, cigarette smoking, increasing age, PR interval, and LA volume as well as reduced eGFR. As reported, IRBBB and cigarette smoking were demonstrated as independent predictors via multivariate analysis. There was no significant difference in AF detection between groups with palpitations and syncope, suggesting a minimal predictive value of presenting symptoms. The current study reports predictors of any AF during follow-up. Unfortunately it is underpowered to predict AF burden and duration. This may prove an area of important future study as the thromboembolic risk conferred by very short episodes of AF diagnosed by loop recorder is unclear and it may be more clinically important to predict longer episodes of PAF. We have also demonstrated a mean CHA2 DS2 VASc score in the PAF group of 3.5, representing an average annual predicted risk of ischemic stroke of between 3.2% and 4.8%23 —a level at which anticoagulation would normally be

October 2015

1219

FRONTERA, ET AL.

Table I. Clinical, Electrocardiographic and Echocardiographic Characteristics of the Whole Study Population, Non-PAF and PAF Subgroups, and P-Value for Subgroup Differences (Welch’s, Fischer’s T-Test, and Mann-Whitney U Test)

Variable Categorical data (n [%]) Male gender IHD Hypertension Diabetes Cigarette smoking High cholesterol LBBB RBBB/IRBBB IRBBB RBBB Continuous data (mean [SD]) Age

Whole Population, n= 200

NonPAF Patients, n= 158

PAF Patients, n= 42

P-Value

111(55.5) 27 (13.5) 107 (53.5) 20 (10) 41 (20.5) 58 (29) 14 (7) 37 (18.5) 21 (10.5) 16 (8)

82 (51.9) 20 (12.7) 81 (51.3) 14 (8.9) 25 (15.8) 36 (22.8) 11 (7) 20 (12.7) 11 (7) 9 (5.7)

29 (69) 7 (16.7) 26 (61.9) 6 (14.3) 16 (38.1) 22 (52.4) 3 (7.5) 17 (40.5) 10 (23.8) 7 (16.7)

0.33 0.62 0.56 0.39 0.026 0.015 1 0.003 0.012 0.06

59.4 (19.3) range 19–95 26.8 (4.4) 73.7 (16.3) 58.1 (8.5) 57.7 (24.7) 71.5 (11.6) 173.3 (25.3) 96.7 (17.9)

69 (16.5) range 25–88 27.8 (4.2) 68.1 (13.8) 55.2 (12.7) 71.1 (32.8) 73 (11.6) 183 (23.5) 101 (21.8)

0.002

2 (2) 2 (2)

2 (2) 3.5 (2)

1 30 Hypertension Cholesterol Smoking Cardiomyopathy Ischemic heart disease Number of syncopes Electrocardiographic data Increasing heart rate (bpm) PR interval (ms) Left bundle branch block Incomplete right bundle branch block Complete right bundle branch block Echocardiographic data Left ventricular ejection fraction (%) Left atrial volume (mL)

BMI = body mass index; BPM = beats per minute; CI = confidence interval.

Figure 1. Forest plot showing odds ratios and 95% confidence intervals for prediction of paroxysmal AF of clinical, electrocardiographic, and echocardiographic factors. AF = atrial fibrillation; BMI = body mass index; HR = heart rate; CRBBB = complete bundle branch block; IHD = ischemic heart disease; IRBBB = incomplete right bundle branch block; LA = left atrial; LBBB = left bundle branch block; LVEF = left ventricular ejection fraction.

can predict episodes of PAF among an unselected general population. We also faced limitations related to the capabilities of the current generation of ILR devices. Although they represent a promising new tool for the clinician, the two devices used in

PACE, Vol. 38

our study must be understood in terms of their R technical specifications: the Medtronic Reveal XT requires 2 minutes of irregular activity to detect AF and has a memory of 27 minutes of detected activity and 22.5 minutes total for up to three episodes of patient activated activity.25 The

October 2015

1221

FRONTERA, ET AL.

SJM ConfirmTM is able to detect irregular activity lasting 30 seconds with a memory for 48 minutes of stored EGMs.26 Our use of two devices with differing detection limits does place limitations on the degree to which we can draw conclusions as to the nature or burden of PAF experienced. It is noteworthy that the vast majority of our cohort R had a Medtronic Reveal XT device. Also, in landmark studies >90% of AF episodes diagnosed lasted 6 minutes or longer and hence the differing diagnostic criteria of the two devices should have had a lesser impact (Ref CRYSTAL AF). Working within the limitations of current technology, we

can still draw valid conclusions regarding the predictive factors for PAF. No doubt as this field advances, further positive developments in device therapy will continue to improve the utility of these devices. Conclusions This study reports that the clinical findings of cigarette smoking and ECG baseline characteristics of IRBBB are strongly and independently associated with paroxysmal episodes of AF in a population of patients investigated by ILR for syncope or palpitations.

References 1. Go AS, Hylek EM, Phillips K a, Chang Y, Henault LE, Selby JV, Singer DE. Prevalence of diagnosed atrial fibrillation in adults. JAMA 2001; 285:2370–2375. 2. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: The Framingham Study. Stroke 1991; 22:983–988. 3. Wang TJ, Larson MG, Levy D, Vasan RS, Leip EP, Wolf Pa, D’Agostino RB, et al. Temporal relations of atrial fibrillation and congestive heart failure and their joint influence on mortality: The Framingham Heart Study. Circulation 2003; 107:2920–2925. 4. Ott A, Breteler MMB, de Bruyne MC, van Harskamp F, Grobbee DE, Hofman A. Atrial fibrillation and dementia in a population-based study: The Rotterdam Study. Stroke 1997; 28:316–321. 5. Benjamin EJ, Wolf Pa, D’Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: The Framingham Heart Study. Circulation 1998; 98:946–952. 6. Hart RG, Pearce La, Rothbart RM, McAnulty JH, Asinger RW, Halperin JL. Stroke with intermittent atrial fibrillation: Incidence and predictors during aspirin therapy. J Am Coll Cardiol 2000; 35:183–187. 7. Healey J, Connoly SJ, Gold MR. Subclinical atrial fibrillation and the risk of stroke. N Engl J Med 2012; 366:120–129. 8. Elijovich L, Josephson SA, Fung GL, Smith WS. Intermittent atrial fibrillation may account for a large proportion of otherwise cryptogenic stroke: A study of 30-day cardiac event monitors. J Stroke Cerebrovasc Dis 2009; 18:185–189. 9. Hohnloser SH, Pajitnev D, Pogue J, Healey JS, Pfeffer MA, Yusuf S, Connolly SJ. Incidence of stroke in paroxysmal versus sustained atrial fibrillation in patients taking oral anticoagulation or combined antiplatelet therapy: An ACTIVE W Substudy. J Am Coll Cardiol 2007; 50:2156–2161. 10. Sanna T, Diener H-C, Passman RS, Di Lazzaro V, Bernstein RA, Morillo CA, Rymer MM, et al. Cryptogenic stroke and underlying atrial fibrillation. N Engl J Med 2014; 370:2478–2486. 11. Rogers G, O’Flynn N. NICE guideline: Transient loss of consciousness in adults and young people. Br J Gen Pract 2011;61:40–42. 12. Willems JL, Robles de Medina EO, Bernard R, Coumel P, Fisch C, Krikler D, Mazur NA, et al. Criteria for intraventricular conduction disturbances and pre-excitation. J Am Coll Cardiol 1985; 5:1261– 1275. 13. January CT, Wann LS, Alpert JS, Calkins H, Cleveland JC, Cigarroa JE, Conti JB, et al. 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the Heart Rhythm Society. J Am Coll Cardiol 2014; 64:e1–76. 14. January CT, Wann LS, Alpert JS, Calkins H, Cigarroa JE, Cleveland JC Jr, Conti JB, et al. ACC/AHA Task Force Members. 2014

1222

15.

16.

17.

18. 19. 20.

21. 22. 23.

24.

25. 26.

AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: Executive summary: A report of the American College of Cardiology/American Heart Association Task Force on practice guidelines and the Heart Rhythm Society. Circulation 2014;130:2071–2104. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. J Clin Epidemiol 2008; 61(4):344– 349. Liao YL, Emidy LA, Dyer A, Hewitt JS, Shekelle RB, Paul O, Prineas R, et al. Characteristics and prognosis of incomplete right bundle branch block: An epidemiologic study. J Am Coll Cardiol 1986; 7:492–499. Bussink BE, Holst AG, Jespersen L, Deckers JW, Jensen GB, Prescott E. Right bundle branch block: Prevalence, risk factors, and outcome in the general population—Results from the Copenhagen City Heart Study. Eur Heart J 2013; 34:138–146. Le V-V, Wheeler MT, Mandic S, Dewey F, Fonda H, Perez M, Sungar G, et al. Addition of the electrocardiogram to the preparticipation examination of college athletes. Clin J Sport Med 2010; 20:98–105. Nielsen JB, Olesen MS, Tango M. Incomplete right bundle branch block: A novel electrocardiographic marker for lone atrial fibrillation. Europace 2011; 13:182–187. Shan H, Zhang Y, Lu Y, Zhang Y, Pan Z, Cai B, Wang N, et al. Downregulation of miR-133 and miR-590 contributes to nicotine-induced atrial remodelling in canines. Cardiovasc Res 2009; 83:465–472. Goette A, Lendeckel U, Kuchenbecker A, Bukowska A, Peters B, Klein HU, Huth C, et al. Cigarette smoking induces atrial fibrosis in humans via nicotine. Heart 2007; 93:1056–1063. Burstein B, Nattel S. Atrial fibrosis: Mechanisms and clinical relevance in atrial fibrillation. J Am Coll Cardiol 2008; 51:802–809. Friberg L, Rosenqvist M, Lip GYH. Evaluation of risk stratification schemes for ischaemic stroke and bleeding in 182 678 patients with atrial fibrillation: The Swedish Atrial Fibrillation cohort study. Eur Heart J 2012; 33:1500–1510. Liao J, Khalid Z, Scallan C, Morillo C, O’Donnell M. Noninvasive cardiac monitoring for detecting paroxysmal atrial fibrillation or flutter after acute ischemic stroke: A systematic review. Stroke 2007; 38:2935–2940. Medtronic. Reveal XT: Insertable Cardiac Monitor (Fact sheet). Available at http://www.medtronic.com/reveallanding/ downloads/reveal-xt.pdf (accessed November 25, 2014). Medical SJ. SJM Confirm ICM DM2100 Spec Sheet. Available at http://professional-intl.sjm.com//media/pro/products/ep/s-z/sjmconfirm-icm/Spec_Confirm_ID_2100_Rev_B_Final.ashx (accessed November 25, 2014).

October 2015

PACE, Vol. 38

Demographic and Clinical Characteristics to Predict Paroxysmal Atrial Fibrillation: Insights from an Implantable Loop Recorder Population.

There is growing interest in detecting paroxysmal atrial fibrillation (PAF) to identify patients at high risk of thromboembolic stroke. The implantabl...
165KB Sizes 0 Downloads 14 Views