Article Text
PDF
Abstract
Background Atrial fibrillation (AF) is the most common arrhythmia among patients with hypertrophic cardiomyopathy (HCM). The relationship between paroxysmal AF and exercise capacity in this population is incompletely understood.
Methods Patients with HCM underwent symptom-limited cardiopulmonary testing with expired gas analysis at Stanford Hospital between October 2006 and October 2012. Baseline demographics, medical histories and resting echocardiograms were obtained for all subjects. Diagnosis of AF was established by review of medical records and baseline ECG. Those with paroxysmal AF were in sinus rhythm at the time of cardiopulmonary testing with expired gas analysis. Exercise intolerance was defined as peak VO2<20 mL/kg/min. We used multivariate logistic regression to evaluate the association between exercise intolerance and paroxysmal AF.
Results Among the 265 patients recruited, 55 had AF (28 paroxysmal and 27 permanent). Compared with those without AF, subjects with paroxysmal AF were older, more likely to use antiarrhythmic and anticoagulant medications, and had larger left atria. Patients with paroxysmal AF achieved lower peak VO2 (21.9±9.2 mL/kg/min vs 26.9±10.8 mL/kg/min, p=0.02) and were more likely to have exercise intolerance (61% vs 28%, p<0.001) compared with those without AF. After adjustment for age, sex and body mass index (BMI) exercise intolerance remained significantly associated with paroxysmal AF (OR 4.65, 95% CI 1.83 to 11.83, p=0.001).
Conclusions Patients with HCM and paroxysmal AF demonstrate exercise intolerance despite being in sinus rhythm at the time of exercise testing.
- Hypertrophic cardiomyopathy
- atrial fibrillation
- exercise capacity
Statistics from Altmetric.com
Introduction
Atrial fibrillation (AF) is the most common arrhythmia in patients with hypertrophic cardiomyopathy (HCM), occurring in approximately 20% of patients.1–3 AF was previously considered to have a benign prognosis in patients with HCM.4 ,5 However, recent studies have found it to be associated with a significantly higher risk of stroke and heart failure-related mortality.1 ,3 ,6 ,7 AF has also been identified as an independent risk factor for functional disability and exercise intolerance in patients with HCM.1 ,8 These studies relied on New York Heart Association (NYHA) classification for estimation of exercise intolerance, which is a subjective measure of exercise capacity and may not reflect the impact of diastolic dysfunction on patients with HCM. Symptom-limited cardiopulmonary testing with expired gas analysis (CPX) is a more objective measure of functional disability in this population.9
The mechanism of association between AF and exercise intolerance in patients with HCM remains unknown. Postulated mechanisms include incomplete ventricular filling resulting from loss of atrial contraction or increase in ventricular rate. Alternatively, AF and exercise intolerance may result from a more advanced or pathological HCM phenotype, rather than the acute haemodynamic effects of this arrhythmia. There is a paucity of data on the significance of paroxysmal AF with regard to exercise capacity. Although patients with HCM in AF during exercise testing would be expected to demonstrate exercise intolerance, we hypothesised that patients with paroxysmal AF would demonstrate exercise intolerance even during sinus rhythm.
Methods
Participants
Between October 2006 and October 2012, 265 consecutive patients with HCM underwent CPX with stress echocardiography at Stanford Hospital and Clinics. Personal and familial medical histories were obtained prior to CPX. The presence of AF was confirmed by review of medical records and ECG. Patients with episodes of AF that only occurred in the setting of a reversible cause, comprising postoperative or postmyocardial infarction states, or brief episodes of atrial arrhythmias noted on loop monitoring or device interrogation alone, were not classified as having AF. Paroxysmal AF was defined as the presence of a clinical diagnosis of AF based on at least one electrocardiographically documented episode of AF identified by chart review. In addition, all subjects with paroxysmal AF were in sinus rhythm at the time of CPX. Permanent AF was defined as the presence of electrocardiographically documented AF identified by chart review, in addition to confirmation of AF on ECG at the time of CPX. Patients with permanent AF were not actively being managed with a rhythm control strategy. The Stanford institutional review board approved this study, and all participants provided written, informed consent.
Measurements
Cardiopulmonary exercise testing
Details regarding CPX have been previously published.9 Medications were not changed prior to testing. All tests began at a speed of 2 miles per hour and a grade of 0%, and subjects underwent CPX using an individualised ramp protocol.10 Protocols were designed to achieve maximal exertion in approximately 8–12 min based on the maximal external work rate estimated from a pretest questionnaire.11 Blood pressure (BP) was measured manually every 2 min during exercise and recovery. Twelve-lead electrocardiographic data was recorded continuously. Heart rate (HR) was recorded from the ECG while standing at rest, at each 20-s interval during the test, and at peak exercise. Subjects were placed in the supine position immediately after exercise.
Exercise capacity in metabolic equivalents (METs) was calculated from treadmill speed and grade.12 An abnormal BP response was defined as a systolic BP that decreased or failed to increase ≥20 mm Hg from the resting value.12 Chronotropic incompetence was defined as an inability to reach 80% of age-predicted maximal HR (220-age).13
Ventilatory expired gas analysis was performed using a metabolic cart (Viasys, Yorba Linda, California, USA). The system was calibrated in a standard fashion using reference gases prior to each test. Minute ventilation (VE), oxygen uptake (VO2), carbon dioxide production (VCO2) and other variables were acquired breath-by-breath and averaged over 30 second intervals. Peak VO2 and peak respiratory exchange ratio (RER) were defined as the highest measured 30-s interval value obtained during the test. Peak VO2 was expressed as an age, sex, protocol and body size predicted value.14 The VE/VCO2 slope was calculated from VE and VCO2 responses throughout exercise, using least squares linear regression.15 The ventilatory threshold was determined by the analysis of ventilatory equivalents.16
Echocardiography
All subjects underwent resting echocardiography within 1 month before exercise testing. HCM was diagnosed based on the presence of a hypertrophied, non-dilated LV (wall thickness >12 mm) in the absence of other primary causes of left ventricular hypertrophy.17 Subjects were classified into five categories based on predominant site of hypertrophy: (1) proximal septal; (2) reverse curvature; (3) apical; (4) concentric groups; and (5) multiple sites of hypertrophy.18 Standard views for M-mode and cross-sectional studies were performed. Dimensions, fractional shortening, pressure gradients, valvular regurgitation and transmitral LV filling velocities were evaluated in accordance with American Society of Echocardiography guidelines.19 Indexed left atrium (LA) volume was calculated as LA volume/body surface area (mL/m2).19 The presence of systolic anterior motion (SAM) of the mitral valve was defined as any contact of the leaflet with the septum during systole.20 Peak instantaneous LV outflow tract (LVOT) flow velocity was measured at rest and during the strain phase of the Valsalva manoeuvre using continuous wave Doppler in apical views. The mitral regurgitation (MR) jet was interrogated with meticulous attention paid to the flow pattern to accurately distinguish it from the subaortic obstruction jet. Tissue Doppler index was performed in the apical view, with a small sample volume placed at the lateral margin of the mitral annulus. Early (E′) and late (A′) diastolic velocities were measured. The lateral E/E′ ratio was calculated. Patients in AF during testing, with a history of mitral valve replacement, myotomy-myectomy or alcohol septal ablation were excluded from diastolic function analyses.
After peak exertion, subjects were immediately placed in the left lateral decubitus position. An experienced technician performed imaging. Peak LVOT gradient was measured first, and degrees of MR were assessed subsequently.
Statistical analysis
Participants were divided into three groups: (1) no AF; (2) paroxysmal AF; and (3) permanent AF. Differences between the no AF and paroxysmal AF groups were compared using the χ2 or unpaired t test. For these comparisons Fisher's exact method was used when any single cell value was less than 5 and the Mann-Whitney U was used when the data showed a strong departure from normality. Exercise intolerance was defined as a peak VO2<20 mL/kg/min based on the Weber classification of functional impairment.21 Logistic regression analyses were performed to assess the associations between peak VO2 or exercise intolerance and AF. Covariates in the primary model included age, sex and BMI, and were selected based on their clinical relevance and their known association with either AF or exercise intolerance. Sensitivity analyses were performed adding to the primary multivariate model several covariates of potential interest that were associated with exercise intolerance: calcium channel blockade, β blockade, chronotropic incompetence, LA size, SAM, E/A ratio, deceleration time and lateral E′.9 Regressions were also performed using an alternative definition of exercise intolerance as peak VO2<60% of age-predicted peak VO2.14 p Values <0.05 were considered statistically significant. All analyses were conducted using STATA (V.12.2, StataCorp, College Station, Texas, USA).
No extramural funding was used to support this work. The authors are solely responsible for the design and conduct of this study, all analyses, the drafting and editing of the paper and its final contents.
Results
Baseline characteristics
The baseline characteristics of the study group stratified by AF category are displayed in table 1. The cohort consisted of 265 subjects with an average age of 52 years; 165 (64%) were white and 103 (39%) were women. Most patients were NYHA Class I (60%), 2 (1%) had mitral valve surgery, 13 (5%) had alcohol septal ablation and 31 (12%) had undergone surgical myomectomy.
Baseline characteristics
AF was present in 55 (21%) subjects; 28 (11%) with paroxysmal AF and 27 (10%) with permanent AF. As expected, subjects with paroxysmal AF were more likely to be on antiarrhythmic (18% vs 1%, p<0.001) and anticoagulant medications (18% vs 1%, p=0.001) compared with participants without AF. Paroxysmal AF subjects also had higher rates of dyslipidaemia (39% vs 19%, p=0.03) but not hypertension or diabetes when compared with the no AF group.
Echocardiographic characteristics
The echocardiographic characteristics of the study group stratified by AF category are displayed in table 2. The mean left ventricular EF of the cohort was 69±10%. The median interventricular septum dimension was 17±5 mm. Rest SAM was present in 130 (49%) subjects, and 12 (5%) had 3+ rest MR. The mean rest LVOT gradient for the entire group was 22±35 mm Hg, which increased to 41±51 mm Hg with exercise.
Echocardiographic rest and exercise data by AF status
Subjects with paroxysmal AF had significantly larger LA diameter (45 mm vs 40 mm, p=0.002) and indexed volume (46 mL/m2 vs 38 mL/m2, p=0.02) than those without AF. The presence of rest SAM was significantly higher in the paroxysmal group compared with the no AF group (79% vs 48%, p=0.002). There was no significant difference between paroxysmal and no AF groups in mean LVOT rest gradient (31 mm Hg vs 23 mm Hg, p=0.30) or exercise gradient (44 mm Hg for both, p=0.78). Nor was there a significant difference in mean E/A ratio between the paroxysmal and the no AF groups (1.19 vs 1.30, p=0.45).
Exercise test responses
Exercise test data stratified by category of AF are displayed in table 3. In the total group, chronotropic incompetence was present in 108 (41%) subjects and an abnormal BP response in 48 (18%) subjects. Occasional premature ventricular contractions were noted in 68 (30%) patients in the total group, with most experiencing <6 premature ventricular contractions per minute. Across all subjects, mean estimated METs was 10.7±4.3, peak VO2 was 25.8±11.0 mL/kg/min and VE/VCO2 slope was 29±6. Maximal RER was 1.12±0.10, suggesting subjects exercised close to maximal intensity.
Exercise test data by AF status
Subjects in the paroxysmal AF group achieved lower average peak HR (124 bpm vs 143 bpm, p=0.002) and had more chronotropic incompetence (64% vs 37%, p=0.006) when compared with the no AF group. Resting HR did not differ between groups. Compared with those with no AF, average METs achieved were lower in the paroxysmal AF group (9.2 vs 11.0, p=0.04). The perceived exertion scores and maximal RERs achieved did not differ significantly between groups.
Peak VO2 by AF category is displayed in figure 1. Compared with those without AF, subjects with paroxysmal AF had lower peak VO2 (21.9 mL/kg/min vs 26.9 mL/kg/min, p=0.02) and higher rates of exercise intolerance (61% vs 28%, p<0.001).
Box and whisker plots of peak VO2 across atrial fibrillation (AF) category. Paroxysmal AF subjects had significantly lower peak VO2 compared with the no AF group (21.9±9.2 mL/kg/min vs 26.9±10.8 mL/kg/min, p=0.02).
Logistic regression analyses demonstrating the associations between peak VO2, exercise intolerance and AF category are displayed in table 4. After adjusting for age, sex and BMI (Model 3), there was a statistically significant association between peak VO2 and paroxysmal AF (OR 0.74 per 5 mL/kg/min increase, 95% CI 0.56 to 0.98, p=0.03). There was also a significant association between exercise intolerance and paroxysmal AF (OR 4.65, 95% CI 1.83 to 11.83, p=0.001) after multivariate adjustment. We performed sensitivity analyses using an alternative definition of exercise intolerance as peak VO2<60% of age-predicted peak VO2.14 The association between paroxysmal AF and exercise intolerance approached significance in this multivariate model (OR 3.3, 95% CI 0.98 to 11.12, p=0.05).
Logistic regression models of peak VO2 and exercise intolerance by AF category
Sensitivity analyses were performed with the inclusion of several variables of interest in the multivariate logistic regressions (Model 3). The association between paroxysmal AF and exercise intolerance remained significant after adjustment for use of calcium channel blockers or β blockers (OR 4.70, 95% CI 1.82 to 12.14, p=0.001), LA size (OR 4.05, 95% CI 1.56 to 10.53, p=0.004), chronotropic incompetence (OR 3.34, 95% CI 1.19 to 9.37, p=0.02), SAM (OR 4.09, 95% CI 1.59 to 10.49, p=0.003), E/A ratio (OR 5.00, 95% CI 1.92 to 13.03, p=0.001), deceleration time (OR 4.56, 95% CI 1.74 to 11.95, p=0.002) and lateral E′ (OR 4.77, 95% CI 1.73 to 13.13, p=0.003).
Discussion
In a cohort of 265 subjects with HCM, we found that paroxysmal AF was significantly associated with exercise intolerance after adjustment for age, sex and BMI. The association with paroxysmal AF existed despite the fact that subjects were in sinus rhythm at the time of exercise testing. The reasons for this association are not clear.
Current hypotheses regarding the mechanism by which AF is associated with exercise intolerance in patients with HCM include decreased cardiac output from loss of atrial systole and diminished stroke volume. However, we found paroxysmal AF to be significantly associated with exercise intolerance in patients who maintained sinus rhythm at time of testing, and were therefore not subject to the active haemodynamic consequences of the arrhythmia. For the purpose of comparison, we also presented the associations between permanent AF and exercise intolerance. Although there was a higher prevalence of chronotropic incompetence and SAM in the paroxysmal AF group, sensitivity analyses demonstrated that these factors did not mediate the association between paroxysmal AF and exercise intolerance. We have previously shown a significant association between diastolic dysfunction and exercise capacity in this cohort9; however diastolic dysfunction did not account for the association between paroxysmal AF and exercise capacity in sensitivity analyses. Previous studies have shown LA enlargement increases risk of AF in patients with HCM.1 ,22 However, the association between paroxysmal AF and exercise intolerance in our study was independent of LA volume. These findings suggest the presence of an uncharacterised pathophysiological process that is responsible for the association between paroxysmal AF and exercise intolerance in patients with HCM.
Given the published heterogeneity in clinical outcomes among patients with HCM with AF, the mechanism of association between AF and functional disability is likely multifactorial. However, there may be a common pathophysiological basis for AF and exercise intolerance in some patients with HCM. Genetic susceptibilities, for example, may concurrently increase the risk of exercise intolerance and AF. Potential mechanisms include the induction of characteristic patterns of ventricular remodelling,23 ,24 atrial enlargement23 and systemic vasodilation or reduction in intravascular volume predisposing patients to hypotension and increased adrenergic tone.25 ,26 Additional studies are necessary to assess which mechanisms that underlie exercise intolerance may also increase the propensity for AF. Our findings suggest that paroxysmal AF in HCM may be a marker for underlying advanced disease or more aggressive cardiomyopathy. Patients who develop AF at younger ages have higher rates of HCM-related mortality, stroke and functional disability,1 and may represent a subgroup of patients with a more pathological HCM phenotype.
Our findings may have several potential clinical implications. Exercise intolerance may be an important predictor of occult AF. Therefore patients with HCM with exercise intolerance may benefit from screening for paroxysmal AF. Identifying occult AF may permit earlier initiation of rhythm control strategies that may improve functional capacity, and anticoagulation to reduce the considerable risk of stroke in this population.27 Olivotto et al1 reported increased risk of composite HCM-related mortality and functional deterioration in patients who progressed from paroxysmal to permanent AF, and therefore suggested interventions aimed at preventing or delaying this transition may improve patient outcomes.
A particular strength of this study is the use of symptom-limited CPX with expired gas analysis to assess functional capacity. Previous studies of exercise tolerance in patients with HCM relied primarily on the NYHA classification scheme, which suffers from poor reproducibility, high subjectivity and insufficient sensitivity to reflect small but important clinical changes.28 Furthermore, the NYHA system was designed to assess symptoms of systolic heart failure and may not be as relevant in the HCM population, which has a higher prevalence of diastolic dysfunction. An alternative measure of exercise capacity is the use of estimated METs, but this is frequently inaccurate and may overestimate exercise capacity particularly if protocols incorporate large incremental stages.10 The use of symptom-limited CPX with expired gas analysis provides a more accurate assessment of exercise capacity in the HCM population. To our knowledge, this study comprises the largest investigation of AF in patients with HCM using CPX, providing a unique opportunity to study the effects of AF on exercise capacity in this population.
Limitations
Exercise testing was carried out on a treadmill with immediate supine echocardiography. Therefore instantaneous changes in haemodynamics at the cessation of exercise cannot be accounted for. Although supine bike exercise would allow echocardiography at immediate cessation of exercise, treadmill testing was chosen because it facilitates more physiological exercise and presumably reflects the degree of outflow obstruction generated during normal activity. There are also limitations to the accuracy of AF classification. There may have been patients with undiagnosed paroxysmal AF who were misclassified into the no AF category. However, the prevalence of AF in our HCM population is comparable with that of previous studies.
Conclusions
We demonstrated that paroxysmal AF is associated with exercise intolerance in subjects with HCM, even in the absence of arrhythmia during testing. The presence of AF may serve as a marker for more advanced or severe underlying cardiomyopathy. Conversely, we have found that patients with HCM with exercise intolerance are more likely to have AF, which is a particularly poor prognostic indicator in this population.
Key messages
-
What is already known about this subject
-
Atrial fibrillation (AF) is the most common arrhythmia in patients with hypertrophic cardiomyopathy (HCM), yet its clinical significance in this population remains controversial. Permanent AF has recently been shown to have a significant association with functional disability and exercise intolerance in patients with HCM. The effects of paroxysmal AF on exercise intolerance in HCM remain unknown.
-
How might this impact on clinical practice
-
Although atrial fibrillation (AF) has been previously thought to cause exercise intolerance through the acute haemodynamic effects of the arrhythmia, our findings suggest an alternative mechanism of association. The presence of AF may serve as a marker for more advanced or severe underlying cardiomyopathy.
-
What does this study add
-
In a retrospective study of 265 patients with hypertrophic cardiomyopathy (HCM), we found a significant association between paroxysmal atrial fibrillation (AF) and exercise intolerance in patients who are in sinus rhythm at the time of cardiopulmonary stress testing. To our knowledge, this study comprises the largest investigation of AF in the patient population with HCM using symptom-limited cardiopulmonary testing with expired gas analysis.
References
Footnotes
-
Contributors We confirm that all listed authors have contributed to the analysis, interpretation of study findings and drafting of the manuscript.
-
Competing interests None.
-
Patient consent Obtained.
-
Ethics approval Stanford Institutional Review Board.
-
Provenance and peer review Not commissioned; externally peer reviewed.