Sleep Apnea and Left Atrial Phasic Function in Heart Failure With Reduced Ejection Fraction

Canadian Journal of Cardiology

-

(2016) 1e9

Clinical Research

Sleep Apnea and Left Atrial Phasic Function in Heart Failure With Reduced Ejection Fraction Nobuhiko Haruki, MD, PhD,a,b Wendy Tsang, MD,a,b Paaladinesh Thavendiranathan, MD,a,b Anna Woo, MD,a,b George Tomlinson, PhD,a Alexander G. Logan, MD,a T. Douglas Bradley, MD,a,c and John S. Floras, MD, DPhil;a,b for the ADVENT-HF Investigators* a

University Health Network and Mount Sinai Hospital Department of Medicine, University of Toronto, Toronto, Ontario, Canada b

c

Peter Munk Cardiac Centre, University Health Network, Toronto, Ontario, Canada

Sleep Research Laboratory of the University Health Network Toronto Rehabilitation Institute and Toronto General Hospital, Toronto, Ontario, Canada

ABSTRACT

  RESUM E

Background: The study aim was to determine whether phasic left atrial (LA) function of patients with heart failure with reduced ejection fraction differs between those with obstructive sleep apnea (OSA) and central sleep apnea (CSA). Methods: Participation in the Adaptive Servo Ventilation for Therapy of Sleep Apnea in Heart Failure (ADVENT-HF) trial requires 2-dimensional echocardiographic documentation of left ventricular ejection fraction
45% and a polysomnographic apnea hypopnea index (AHI)  15 events per hour. Of initial enrollees, we identified 132 patients in sinus rhythm (82 with predominantly OSA and 50 with CSA). To determine LA reservoir (expansion index; EI), conduit (passive emptying index; PEI), and booster function (active emptying index), we blindly quantified maximum and minimum LA volume and LA volume before atrial contraction.

tude avait pour objectif de de terminer si, chez les Introduction : L’e duction de la fraction patients atteints d’insuffisance cardiaque avec re jection, la fonction auriculaire gauche phasique diffe rait entre les d’e e du sommeil obstructive (ASO) et ceux patients souffrant d’apne e centrale du sommeil (ACS). souffrant d’apne thodes : Pour participer à l’e tude « Adaptive Servo Ventilation for Me Therapy of Sleep Apnea in Heart Failure » (ADVENT-HF), les patients jection ventriculaire gauche
45 % devaient avoir une fraction d’e e par e chocardiographie bidimensionnelle et un indice documente e et d’hypopne e (IAH)  15 e  ve nements par heure selon les d’apne sultats polysomnographiques. Parmi l’ensemble des patients inire tude, nous avons recense  132 cas de rythme tialement inscrits à l’e dominait et 50 chez qui l’ACS sinusal (82 patients chez qui l’ASO pre dominait). Pour de terminer les indices des fonctions de re servoir pre

Heart failure (HF) often is accompanied by either obstructive sleep apnea (OSA) or central sleep apnea (CSA).1,2 OSA results from upper airway collapse, whereas CSA arises from respiratory control instability secondary to the effects of pulmonary congestion and augmented chemoreceptor responsiveness. Despite their different pathogeneses, OSA and CSA play an adverse role in the development of cardiac remodelling and dysfunction by generating recurrent cycles of hypoxia and reoxygenation, by evoking excess sympathetic nervous system

activity, and in the case of OSA by increasing left ventricular (LV) and left atrial (LA) wall stress secondary to acute reductions in intrathoracic pressure.1-4 Positioned between the left ventricle and the pulmonary circulation, the left atrium acts not only to modulate LV filling but also as a pressure-buffering chamber.5 LA structural and functional remodelling caused by pressure or volume overload is common in HF,6 reflects its chronicity and severity, and provides important independent prognostic information.7-9 Sleep apnea (SA) also is an independent risk factor for LA enlargement in patients with congestive HF.10 Whereas upper airway obstruction imposes greater acute increases in LA transmural pressure than central apnea,4 pulmonary capillary wedge pressure during wakefulness is in general higher in those with predominantly CSA.11 Should impaired LA function exacerbate pulmonary congestion, the resulting stimulation of pulmonary J receptors will amplify respiratory instability.12,13 Phasic indices (LA reservoir, conduit, or booster function) provide more detailed and

Received for publication December 23, 2015. Accepted February 9, 2016. *For a complete list of ADVENT-HF Committees and Principal Investigators, see Supplemental Appendix S1. Corresponding author: Dr John S. Floras, Suite 1614, Mount Sinai Hospital, 600 University Ave, Toronto, Ontario M5G 1X5, Canada. Tel.: þ1-416-586-8704; fax: þ1-416-586-8702. E-mail: john.fl[email protected] See page 8 for disclosure information. Clinical Trial Registration: www.clinicaltrials.gov NCT01128816.

http://dx.doi.org/10.1016/j.cjca.2016.02.047 0828-282X/Ó 2016 Canadian Cardiovascular Society. Published by Elsevier Inc. All rights reserved.

2

Canadian Journal of Cardiology Volume - 2016

Results: Each of EI (P ¼ 0.004), PEI (P < 0.001), and active emptying index (P ¼ 0.045) was less in participants with CSA compared with those with OSA, whereas average left ventricular ejection fraction and LA and left ventricular volumes were similar. Multivariable analysis identified an independent relationship between central AHI and LA EI (P ¼ 0.040) and PEI (P ¼ 0.005). In contrast, the obstructive AHI was unrelated to any LA phasic index, and slopes relating central AHI to EI and PEI differed significantly from corresponding relationships with obstructive AHI (P ¼ 0.018; P ¼ 0.006). Conclusions: In these ADVENT-HF patients with heart failure with reduced ejection fraction, all 3 components of LA phasic function (reservoir, conduit, and contractile) were significantly reduced in those with CSA compared with participants with OSA. The severity of CSA, but not OSA associated inversely and independently with LA reservoir and conduit function. Impaired LA phasic function might be consequent to or could exacerbate CSA.

(indice d’expansion), de conduite (indice de vidange passive) et de  à l’insu les contraction (indice de vidange active), nous avons calcule volumes maximum et minimum de l’oreillette gauche, ainsi que le volume avant la contraction auriculaire. sultats : Les indices d’expansion (p ¼ 0,004), de vidange passive Re taient moindres chez les (p < 0,001) et de vidange active (p ¼ 0,045) e patients atteints d’ACS par rapport à ceux atteints d’ASO, tandis que la jection ventriculaire gauche moyenne et les volumes de fraction d’e taient semblables. L’analyse l’oreillette et du ventricule gauches e e a permis d’e tablir une relation inde pendante entre l’IAH multivarie centrale et les indices d’expansion (p ¼ 0,040) et de vidange passive (p ¼ 0,005) de l’oreillette gauche. En revanche, l’IAH obstructive ne sentait aucune relation avec les indices phasiques de l’oreillette pre tablissant le lien entre l’IAH centrale et les gauche, et les courbes e raient de façon signiindices d’expansion et de vidange passive diffe es pour l’IAH obstructive (p ¼ 0,018; p ¼ ficative de celles observe 0,006). tude ADVENT-HF mene e auprès de Conclusions : Dans le cadre de l’e duction de la fraction patients atteints d’insuffisance cardiaque avec re jection ventriculaire gauche, les valeurs observe es pour les trois d’e composantes de la fonction phasique auriculaire gauche (fonctions de servoir, de conduite et de contraction) e taient significativement re rieures chez les patients atteints d’ACS par rapport à ceux atteints infe  de l’ACS e tait inversement d’ASO. Contrairement à l’ASO, la gravite e, de manière inde pendante, à une baisse des indices d’exassocie servoir et de conduite). pansion et de vidange passive (fonctions de re quence de Ce dysfonctionnement phasique pourrait être une conse l’ACS, ou encore constituer un facteur d’exacerbation de cette dernière.

dynamic information concerning LA function than global measures of LA volume. The aim of the present study was to test, in patients with heart failure with reduced ejection fraction (HFrEF) in sinus rhythm with untreated SA, the hypothesis that one or more of these components of LA phasic function is diminished in those with CSA relative to those with OSA.

pregnancy; and previous adaptive servo ventilation or other positive airway pressure therapy for SA. Baseline clinical data include documentation of prescribed medications and plasma concentrations of N-terminal pro B-type natriuretic peptide (NT-pro BNP) or B-type natriuretic peptide, if only the latter was available at the trial hospital. The ADVENT-HF trial protocol was approved by the University Health Network Research Ethics Board (09-0834-B).

Methods Study design and subjects The multicentre Adaptive Servo Ventilation for Therapy of Sleep Apnea in Heart Failure (ADVENT-HF) trial is testing, in patients with HFrEF, the hypothesis that treatment of CSA or of nonsleepy patients with OSA using adaptive servo ventilation will reduce the composite rate of deaths and cardiovascular hospitalizations. Inclusion criteria are: nonhospitalized adult patients diagnosed with American Heart Association stage B-D HF due to ischemic, hypertensive, or idiopathic dilated cardiomyopathy with a 2-dimensional echocardiography-quantified LV ejection fraction (LVEF)
45% receiving doses of optimal tolerated medical therapy for HFrEF unchanged for at least 2 weeks with SA with  15 apneas and hypopneas per hour of sleep (apnea hypopnea index [AHI]). Principal exclusion criteria are: primary valvular, hypertrophic obstructive, restrictive, or postpartum causes of cardiomyopathy; any 1 of acute myocardial infarction, cardiac surgery, percutaneous coronary intervention, defibrillator or cardiac resynchronization device implantation within 3 months; an LV assist device; previous cardiac transplantation;

Echocardiography At each site, images were stored electronically in Digital Imaging and Communications in Medicine format then submitted to the Toronto General Hospital Core Echocardiography Laboratory. An experienced echocardiographer (N.H.) blinded to subjects’ polysomnographic findings reviewed the first 155 visit 1 (screening) echocardiographic studies to identify those with image quality sufficient to quantify LA phasic function. Of these, atrial fibrillation (precluding determination of LA phasic function) was present in 22 participants, and these patients were excluded. One subject in sinus rhythm but with extraordinarily large LV dimensions also was excluded from analysis. In the remainder (n ¼ 132), LV end-diastolic volume, endsystolic volume, and LVEF were calculated using the biplane Simpson method from apical 4-chamber and 2-chamber views according to the recommendations of the American Society of Echocardiography.14 Left ventricular mass (LVM) was calculated using the Devereux formula15 then indexed to body surface area. Mitral regurgitation (MR) was assessed semiquantitatively according to current guidelines.16

Haruki et al. Sleep Apnea and Left Atrial Phasic Function

3

LA volumes were calculated using the biplane area-length method then indexed to body surface area.14 To determine the 3 components of LA phasic function, maximum LA volume (LAVmax) was measured at the point of mitral valve opening in end-systole, minimum LA volume (LAVmin) at the point of mitral valve closure in end-diastole, and LA volume before atrial contraction (LAVpreA) at the onset of the P wave in the simultaneously recorded electrocardiogram. LA reservoir function (during ventricular systole) was assessed by calculating the expansion index as 100  (LAVmax  LAVmin)/LAVmin. LA conduit function (during early diastole) was assessed by calculating the passive emptying index as 100  (LAVmax  LAVpreA)/LAVmax. Booster pump (or LA contractile) function (during late diastole) was assessed by calculating the active emptying index as 100  (LAVpreA  LAVmin)/LAVpreA. In previous publications the first author and his echocardiography colleagues have reported intraobserver and interobserver intraclass correlations for these specific measurements as 0.95 (0.95-0.96) and 0.92 (0.890.95), respectively.17-20 With respect to LAVmax, in a previous publication by the first author the reported intraobserver variability was 5% and the interobserver variability was 8%.17

multivariable linear regression was used to determine independent predictors of LA expansion and passive emptying indices. These models included clinically significant covariates (age, sex, body mass index, ischemic etiology of HF, moderate or greater MR, logNT-pro BNP, LVEF, LVM index, obstructive AHI [OAHI], and central AHI [CAHI]). The relative importance of each variable in the model was summarized by the Lindeman, Merenda, and Gold partial R2. For each component of LA phasic function we also fitted a linear regression with CAHI and OAHI as the 2 predictors and the atrial function indices as the outcome. A linear contrast was used to test whether the slopes for each SD change in CAHI and OAHI were equal. A 2-sided P value < 0.05 was considered statistically significant. Analyses were performed using JMP version 11.0 (SAS Institute Inc, Cary, NC) and R software version 3.0.2 (https://cran.r-project.org/bin/ windows/base/old/3.0.2).

Polysomnography

Subjects’ mean age was 66  10 years, 86% were male, and their mean LVEF was 30  8% (Table 1). Of the 132 participants, 82 had predominantly OSA. Compared with OSA patients, those with CSA were older, a greater proportion was male, blood pressure was lower, there was a greater prevalence of ischemic cardiomyopathy, and the AHI and the oxygen desaturation index were significantly higher, but there were no significant between-group differences with respect to height, weight, body mass index, prevalence of hypertension, and diabetes, plasma NT-pro BNP, HF symptom severity, medication use, reported HF duration (CSA: 84  94 months; OSA: 86  77 months; P ¼ 0.34) or the proportion with implanted cardiac devices (Tables 1 and 2).

Thoracoabdominal motion was measured using respiratory inductance plethysmography,21 nasal air flow was assessed using nasal pressure cannulae, and arterial oxyhemoglobin saturation was determined using pulse oximetry. Polysomnographic data were transmitted electronically to the Toronto Rehabilitation Institute Core Sleep Research Laboratory, where sleep stages and arousals from sleep were scored uniformly according to standard criteria22,23 then reviewed and interpreted by 1 of the 2 core sleep laboratory’s sleep physicians. Apneas and hypopneas were defined as a reduction in air flow from intranasal pressure of at least 90%, or between 50% and 90%, respectively, for at least 10 seconds. If the signal was attenuated because of mouth-breathing, the sum of thoracoabdominal motion was used to detect apneas and hypopneas. Apneas were classified as obstructive if thoracoabdominal motion was present and central if absent. Hypopneas were classified as obstructive if thoracoabdominal motion was out of phase or if air flow limitation was observed on the nasal pressure signal, whereas central hypopneas were those in which thoracoabdominal motion was in phase without evidence of air flow limitation on the nasal pressure signal. The AHI was defined as the number of apneas and hypopneas per hour of sleep. Subjects with an AHI  15 with  50% events of obstructive origin were classified as having OSA, whereas subjects with an AHI  15 with > 50% events of central origin were classified as having CSA.24 The oxygen desaturation index was quantified as the number of dips in arterial oxyhemoglobin saturation of  3% per hour of sleep. Statistical analysis Continuous data are expressed as mean  SD or median (interquartile range). Categorical data are presented as numbers with percentages. Differences in continuous variables between the 2 groups were evaluated using the Wilcoxon rank sum test and categorical variables were compared using Fisher exact test or the c2 test as appropriate. Fully adjusted

Results Patient characteristics

Echocardiographic data Left atria of this study cohort were markedly enlarged (LAVmax: 97  39 mL; maximal LA volume index: 49  19 mL/m2; LAVmin: 63  38 mL; minimum LA volume index: 32  18 mL/m2). As illustrated in Figure 1, all 3 components of LA phasic function were reduced significantly in subjects classified with CSA compared with those with OSA. OSA and CSA subjects had similar LV dimensions, LA and LV volumes, LVEF, LVM index, and prevalence of moderate or severe MR (Table 3). Similar significant reductions in all 3 components of LA phasic function were present if comparisons were restricted to participants in whom > 75% of events were either central (n ¼ 32) or obstructive (n ¼ 62). There were significant correlations between the AHI and the LA expansion index (r ¼ 0.24; P ¼ 0.006) and passive emptying index (r ¼ 0.20; P ¼ 0.02) and between the CAHI and all 3 indices of LA phasic function but none between the OAHI and LA phasic index (Fig. 2). There were significant between-group differences in the slopes of these relationships for the expansion index (P ¼ 0.018) and the passive emptying index (P ¼ 0.006) but not the active emptying index (P ¼ 0.23).

4

Canadian Journal of Cardiology Volume - 2016

Table 1. Patient characteristics Characteristic Age, years Male sex, % Height, cm Weight, kg Body mass index, kg/m2 Body surface area, m2 Comorbidities Hypertension Diabetes mellitus Heart rate, beats per minute Systolic blood pressure, mm Hg Diastolic blood pressure, mm Hg NT-pro BNP, pg/mL* ACC/AHA HF stage CþD Cardiac devices ICD CRT Pacemaker Ischemic etiology of HF Medications ACE-I/ARB b-Blocker Diuretics Aldosterone antagonist Nitrates Digoxin Statin

All subjects (n ¼ 132) 66  10 114 (86%) 171.2  9.7 86.9  23.4 28.9  4.9 1.98  0.24 78 (59%) 43 (33%) 62  9 116  18 70  9 912 (374-2212) 93 (70%)

CSA group (n ¼ 50)

OSA group (n ¼ 82)

69  11 48 (96%) 171.4  9.4 82.9  15.9 28.2  4.9 1.95  0.20

64  9 66 (80%) 171.1  10.0 89.4  26.7 29.4  4.9 2.00  0.26

27 (54%) 16 (32%) 61  8 113  18 66  9 1280 (460-2746) 39 (78%)

51 (62%) 27 (33%) 63  9 119  18 72  9 818 (322-2107) 54 (66%)

P 0.007 0.02 0.99 0.30 0.10 0.54 0.46 1.00 0.31 0.13 < 0.001 0.10 0.17

64 18 65 85

(48%) (14%) (49%) (64%)

23 8 25 39

(46%) (16%) (50%) (78%)

41 10 40 46

(50%) (12%) (49%) (56%)

0.72 0.60 1.00 0.02

107 121 97 51 49 15 99

(81%) (92%) (73%) (39%) (37%) (11%) (75%)

41 45 37 20 18 6 37

(82%) (90%) (74%) (40%) (36%) (12%) (74%)

66 76 60 31 31 9 62

(80%) (93%) (73%) (38%) (38%) (11%) (76%)

1.00 0.75 1.00 0.85 0.86 1.00 0.84

Values are mean  SD, n (percent), or median (25-75 percentile range). ACC, American College of Cardiology; ACE-I, angiotensin-converting enzyme inhibitor; AHA, American Heart Association; ARB, angiotensin-2 receptor antagonist; CRT, cardiac resynchronization therapy; CSA, central sleep apnea; HF, heart failure; ICD, implantable cardioverter-defibrillator; NT-pro BNP, N-terminal pro B-type natriuretic peptide; OSA, obstructive sleep apnea. * N ¼ 100.

Multivariable linear regression analysis revealed that CAHI, NT-pro BNP, and at least moderate MR were independent predictors of the LA expansion index and that CAHI and NT-pro BNP were independent predictors of the LA passive emptying index (Table 4). Diastolic functional parameters (not mandated by the ADVENT-HF echocardiographic protocol) were of sufficient quality to permit between-group comparisons of 1 or more of E and A wave velocity, the E/e’ ratio, and deceleration time in up to 40 subjects; there were no significant differences with respect to any of these variables between patients with OSA and CSA (eg, for E/e’: 13.8  3.8 [n ¼ 12] vs 13.5  8.5 [n ¼ 16]).

Discussion The major novel findings of this first study to relate LA phasic function in HFrEF to SA type were: (1) each phasic component of LA function (ie, LA expansion, passive and active emptying indices [representing LA reservoir, conduit, and booster function, respectively]) was attenuated significantly in patients with CSA compared with those with OSA; and (2) CAHI, but not OAHI correlated significantly with LA expansion and passive emptying. There being no significant between-group differences with respect to mean values for LVEF, LA, and LV volumes, LVM index, NT-pro BNP concentrations or MR severity, these observations are unlikely a simple consequence of more severe HF in those with CSA.

Table 2. Polysomnographic data Characteristic AHI/h Central AHI/h Obstructive AHI/h Total sleep time, minutes Stage 1 sleep, % Stage 2 sleep, % Slow wave sleep, % REM sleep, % Mean SaO2, % Minimum SaO2, % Oxygen desaturation index/h Total arousal index/h

All subjects (n ¼ 132) 40.1 16.2 23.8 290 14.2 63.1 10.8 12.0 93.4 81.1 36.1 37.4

           

18.8 18.4 16.7 75 13.1 12.5 8.7 7.0 2.3 8.9 19.0 17.3

CSA group (n ¼ 50) 48.7 35.9 12.8 283 18.9 59.0 9.7 12.3 93.6 81.8 41.5 40.2

           

16.3 14.7 9.1 71 17.3 14.1 8.0 7.3 2.3 7.5 16.5 17.2

OSA group (n ¼ 82)

P

           

< 0.001 < 0.001 < 0.001 0.30 0.003 0.014 0.39 0.91 0.24 0.58 0.003 0.09

34.8 4.2 30.6 295 11.2 65.6 11.4 11.7 93.2 80.7 32.8 35.7

18.3 5.8 16.7 78 8.6 10.8 9.1 6.8 2.4 9.7 19.7 17.2

Values are mean  SD. AHI, apnea-hypopnea index; CSA, central sleep apnea; OSA, obstructive sleep apnea; REM, rapid eye movement; SaO2, arterial oxyhemoglobin saturation.

Haruki et al. Sleep Apnea and Left Atrial Phasic Function

5

P = 0.0004

(%) 53.6

24.3

CSA

P < 0.0001

(%)

P = 0.045

(%)

80.1 41.5

17.3 6.7

23.1 7.8

19.6 8.7

24.3 13.0

OSA

CSA

OSA

CSA

OSA

Expansion index

Passive emptying index

Active emptying index

Figure 1. Comparison, in heart failure patients with reduced left ventricular ejection fraction, of the 3 components of left atrial phasic function between those with predominant central sleep apnea (CSA) and those with obstructive sleep apnea (OSA). Open circles represent data points for each trial participant. Mean value  SD.

Systolic HF and SA associate independently with LA enlargement.10 In HF, LA dilatation is not only a pathophysiological response to pressure and/or volume overload, but also anticipates cardiac hospitalization or death.8 OSA and CSA are characterized by recurring mechanical, chemical and autonomic forces4 that affect adversely atrial and ventricular wall stress, structure, contraction, and relaxation.25,26 In HFrEF patients, LA volume index correlates directly with CO2 chemosensitivity and the frequency of CSA.27 In a longitudinal study of HFrEF patients with CSA, AHI and LA area were the only independent predictors of cardiac death.28

Vazir et al. described greater LA dimensions in HFrEF patients with CSA than OSA whereas LVEF and LV size were similar.29 Conversely, Oldenburg et al. reported no difference in LA diameter between OSA and CSA in HFrEF.30 Their conclusions were derived from a single-dimension measurement, but the LA enlarges asymmetrically in HFrEF.31 We quantified instead LA volume and found no significant between-group difference. The LA acts as a reservoir during systole, as a conduit during early diastole, and as a booster pump to augment LV filling during late diastole.32 Consequently, analysis of LA

Table 3. Echocardiographic indices Measure Interventricular septal thickness, mm Posterior wall thickness, mm LV diastolic dimension, mm LV systolic dimension, mm LVEDV, mL LVEDV index, mL/m2 LVESV, mL LVESV index, mL/m2 LVEF, % LVM index (g/m2) LAVmax, mL LAVImax, mL/m2 LAVpreA, mL LAVIpreA, mL/m2 LAVmin, mL LAVImin, mL/m2 Prevalence of moderate or severe MR, %

CSA group (n ¼ 50)

OSA group (n ¼ 82)

P

8.9  1.7 8.8  1.6 63.6  7.9 57.0  8.4 227.7  76.3 117.2  38.5 166.9  74.1 86.1  38.4 28.5  8.7 121.8  30.3 96.8  42.1 49.4  19.5 81.3  41.0 41.4  19.1 67.6  41.8 34.4  19.7 2 (4%)

8.9  1.8 8.8  1.6 64.6  7.9 57.3  10.4 220.1  61.3 110.3  28.5 154.5  53.6 77.4  25.3 30.8  7.8 121.3  29.5 97.0  37.5 48.7  18.6 76.0  34.3 38.0  16.8 60.2  35.0 30.2  17.4 7 (9%)

0.90 0.81 0.26 0.48 1.00 0.76 0.78 0.52 0.24 0.97 0.81 0.98 0.58 0.34 0.20 0.11 0.48

Values are mean  SD. CSA, central sleep apnea; LAVmin, minimum left atrial volume; LAVImin, minimum left atrial volume index; LAVImax, maximum left atrial volume index; LAVIpreA, left atrial volume index before atrial contraction; LAVmax, maximum left atrial volume; LAVpreA, left atrial volume before atrial contraction; LV, left ventricular; LVEDV, LV end-diastolic volume; LVEF, LV ejection fraction; LVESV, LV end-systolic volume; LVM, LV mass; MR, mitral regurgitation; OSA, obstructive sleep apnea.

6

Canadian Journal of Cardiology Volume - 2016

Expansion index (%)

Passive emptying index (%)

r = –0.30 P < 0.001

CAHI (events per hour)

CAHI (events per hour)

r = 0.05 P = 0.55

CAHI (events per hour)

r = 0.002 P = 0.98

r = 0.11 P = 0.23 Active emptying index (%)

Passive emptying index (%)

Expansion index (%)

OAHI (events per hour)

r = –0.17 P = 0.046 Active emptying index (%)

r = –0.29 P < 0.001

OAHI (events per hour)

OAHI (events per hour)

Figure 2. Correlations between each of left atrial expansion index, passive emptying index, and active emptying index, and either the central sleep apnea severity (central apnea hypopnea index [CAHI]; upper panels) or the obstructive sleep apnea severity (obstructive apnea hypopnea index [OAHI]; lower panels) of each participant. Significant correlations are specific to central sleep apnea severity. Relationships between the apnea indices and the left atrial phasic indices were significantly different with respect to apnea type for the expansion index (P ¼ 0.018) and for the passive emptying index (P ¼ 0.006).

phasic function provides more detailed and dynamic information than LA volume.7 Here, all 3 components of LA phasic function were less in those with CSA. In HFrEF, due to compromised contractility and reduced LV compliance, the LA is exposed to higher LV pressure during diastole; LA pressure is obliged to increase to maintain adequate LV filling. Persistent LA pressure elevation leads to LA wall stiffening, decreased atrial compliance, chamber enlargement, and diminished LA phasic function.33 LA reservoir function is governed by atrial compliance and LV contraction. LA conduit function is also influenced by atrial compliance and is reciprocally related to its reservoir function.32 Decreased LA reservoir function is associated with elevated LV end-diastolic pressure (LVEDP) and pulmonary capillary wedge pressure, and LA conduit function relates inversely to LVEDP.34 LA reservoir and LA conduit function related independently to the severity of CSA but not OSA. Such dysfunction, in turn, might promote the progression of CSA and HF. In the present context, in CSA patients relative to OSA patients (with their comparatively preserved LA function), for any given LVEDP the combination of decreased LA compliance and impaired LA phasic function would be expected to generate a greater LA pressure at similar LA volumes and thus

cause more pulmonary congestion. Stimulation of pulmonary J receptors by such congestion augments ventilatory loop gain and elicits reflexively hyperpnea. The resulting hypocapnea elicits central apnea when PCO2 falls below the apnea threshold.3 This pathophysiology can be simulated in HFrEF patients with CSA during wakefulness: when fluid is shifted rapidly into the thorax, ventilation augments and PCO2 falls toward the apnea threshold.13 Impaired LA buffering, resulting in greater exposure of the pulmonary venous circulation to LV pressures throughout diastole, would render an HFrEF patient more susceptible to CSA. Some limitations merit acknowledgement. Patients whose HF was attributed primarily to MR were not enrolled in the ADVENT-HF trial; nor were HFrEF patients without SA. Echocardiographic and polysomnographic data were acquired when participants were recruited; in HFrEF the predominant SA type might vary over time.35,36 In the present cohort, a greater proportion of CSA patients had ischemic cardiomyopathy. More impaired LA function, attributed to atrial myopathy, has been observed in patients with idiopathic dilated relative to ischemic cardiomyopathy despite similar loading conditions and LA volumes.37 Ageing also impairs LA function.19 However, by the seventh decade, the small (4-year) age difference between groups would not explain the

Haruki et al. Sleep Apnea and Left Atrial Phasic Function

7

Table 4. Multivariable linear regression analyses Variable LA expansion index Age Sex BMI Ischemic etiology of HF MR  moderate Log NT-pro BNP* LVEF LVM index Obstructive AHI Central AHI Multivariable model: R2 ¼ P < 0.001 LA passive emptying index Age Sex BMI Ischemic etiology of HF MR  moderate Log NT-pro BNP* LVEF LVM index Obstructive AHI Central AHI Multivariable model: R2 ¼ P < 0.001

Estimate

SE

P

Partial R2

0.261 0.332 0.4334 1.9 14.502 9.128 0.1157 1.4 0.108 0.655 0.8695 0.6 0.588 6.916 0.9324 0.3 32.354 13.316 0.0171 4.9 28.734 5.918 < 0.0001 21.3 0.522 0.418 0.2146 6.3 0.043 0.113 0.7019 1.2 0.038 0.213 0.8585 0.5 0.445 0.213 0.0396 5.8 0.441; adjusted R2 ¼ 0.379; F10,89 ¼ 7.03; 0.010 0.077 0.8928 0.6 3.645 2.129 0.0903 2.7 0.134 0.153 0.3814 0.3 1.164 1.613 0.4722 0.2 4.041 3.105 0.1965 1.5 4.334 1.380 0.0023 10.4 0.072 0.097 0.4597 4.0 0.014 0.026 0.5837 1.2 0.015 0.050 0.7570 1.2 0.144 0.050 0.0048 10.3 0.323; adjusted R2 ¼ 0.247; F10,89 ¼ 4.34;

AHI, apnea-hypopnea index; BMI, body mass index; HF, heart failure; LVEF, left ventricular ejection fraction; LVM, left ventricular mass; MR, mitral regurgitation; NT-pro BNP, N-terminal pro B-type natriuretic peptide. * n ¼ 100.

present differences. We compared only patients in sinus rhythm when imaged, but some might have had previous atrial fibrillation, which can increase the risk of developing CSA38 and can impair LA function.39 For pragmatic reasons the ADVENT-HF echocardiography protocol did not mandate acquisition of indices of LV diastolic filling, such as E/A or E/e’, but when measured, there were no between-group differences. Nor were there any between-group differences with respect to plasma NT-pro BNP concentration, an index of LV end-diastolic wall stress, or mean LA volume. Such differences would be anticipated if diastolic filling was more impaired in one specific group. Estimates of LV diastolic function are influenced instantaneously by changes in hemodynamics and loading conditions; the reliability of E/e’ in advanced HFrEF is now doubted.40 Importantly, upper airway obstruction evokes a substantially greater increase in negative intrathoracic pressure than central apnea; this in turn has a much greater proportionate effect on atrial than on ventricular transmural pressure, wall stress, and diastolic relaxation,4 yet it was the CSA participants who exhibited significant impairment of all elements of LA phasic function. Absence of diastolic filling estimates in all subjects does not render any less definitive either the present novel finding of differences in all 3 components of LA phasic function between CSA and OSA patients with HF of similar severity or their specific relationships with the central but not the OAHI. SA is present in approximately 50% of HFrEF patients.1 The 2013 American College of Cardiology Foundation/ American Heart Association guideline for HF assigns “continuous positive airway pressure can be beneficial to increase LVEF and improve functional status in patients with

HF and SA” a ‘class IIa’ recommendation, supported by ‘grade B’ evidence.41 However, the reasoning substantiating this recommendation conflated OSA and CSA, despite their differing pathophysiology, and the most compelling evidence in support of this proposition, namely the Canadian Continuous Positive Airway Pressure for Patients With Central Sleep Apnea and Heart Failure Trial (CANPAP)24 was described incorrectly as a long-term randomized trial of treating OSA (page e263). Although we do not as yet know the true effect of suppressing CSA on cardiovascular end points, the initial Treatment of Sleep-Disordered Breathing with Predominant Central Sleep Apnea by Adaptive Servo-Ventilation in Patients with Heart Failure Trial (SERVE-HF) publication, which reported only the results of an allocation strategy42 is likely to stimulate, in future, different approaches to the therapy of CSA and OSA.43 The present findings raise the intriguing possibility that in HFrEF patients at risk for SA, LA phasic characteristics, particularly if markedly impaired, might offer a readily available rapid screening clue to the presence of CSA: no CSA patient had an expansion index > 103%, a passive emptying index > 32% or an active emptying index > 38% (Fig. 1). A larger ADVENTHF data set would enable the construction of receiver operating curves to test this hypothesis. The present study thus provides the first data relating LA phasic function to SA type in HFrEF. Correlates of HF severity were similar in the 2 cohorts, yet LA reservoir, conduit, and contractile functions were less in CSA patients relative to those with OSA. Unlike OSA severity, CSA severity associated inversely and independently with LA expansion index and passive emptying index. Should such impaired LA function lead to diminished LA compliance the resulting pulmonary congestion could worsen CSA and accelerate HF progression. Impaired LA function might also place such individuals at increased risk of developing atrial fibrillation, hospitalization for HF, and death.7,44 The present results of distinctly different LA phasic properties and their relation to apnea severity in OSA and CSA advance our understanding of the pathophysiology of SA and HFrEF. Acknowledgements Dr Floras takes responsibility for the content of the report, including the data and analysis. Dr Bradley is the Chair of the ADVENT-HF Clinical Trial Steering Committee, and Dr Logan and Dr Floras are its Vice Chairs. Dr Woo is the Director of the Toronto General Hospital Echocardiography Laboratory and the ADVENT-HF Core Echocardiography Laboratory lead. Drs Tsang and Thavendiranathan are her associates. All 3 were engaged in ensuring the quality of the core laboratory echocardiograms, in mentoring Dr Haruki, in the interpretation of data drafting and in the drafting of the report. Dr Haruki, who has several previous publications in this field, conducted under supervision blinded analysis of all echocardiograms, assembled all clinical data provided by the trial’s core Sleep Laboratory. Dr Tomlinson, our ADVENTHF statistician, conducted the principal statistical analyses. All authors made substantial contributions to one or more of conception and design, to acquisition of data, to analysis and interpretation of data, and to the drafting and critical revision of the report for important intellectual content and all have provided final approval of the version to be published.

8

For a complete list of ADVENT-HF Committees and Principal Investigators, see Supplemental Appendix S1. Funding Sources The ADVENT-HF trial is supported by a Clinical Trials grant from the Canadian Institutes of Health Research (IS295225) and an unrestricted gift from Philips Respironics Inc to the Toronto Rehabilitation Institute of the University Health Network. Dr Haruki was supported by a Bluma Appel International Fellowship of the Mount Sinai Hospital Department of Medicine, Toronto. Dr Bradley holds the Clifford Nordel Chair in Sleep Apnea and Rehabilitation Research at the University Health Network. Dr Floras holds the Tier 1 Canada Research Chair in Integrative Cardiovascular Biology. None of these entities had any role in the development of the research and report. Disclosures Dr Bradley is Chair and Drs Floras and Logan are ViceChairs of the ADVENT-HF trial. The remaining authors have no conflicts of interest to disclose. References 1. Kasai T, Floras JS, Bradley TD. Sleep apnea and cardiovascular disease: a bidirectional relationship. Circulation 2012;126:1495-510. 2. Lyons OD, Bradley TD. Heart failure and sleep apnea. Can J Cardiol 2015;31:898-908. 3. Bradley TD, Floras JS. Sleep apnea and heart failure: part II: central sleep apnea. Circulation 2003;107:1822-6. 4. Bradley TD, Floras JS. Sleep apnea and heart failure: part I: obstructive sleep apnea. Circulation 2003;107:1671-8. 5. Melenovsky V, Hwang SJ, Redfield MM, et al. Left atrial remodeling and function in advanced heart failure with preserved or reduced ejection fraction. Circ Heart Fail 2015;8:295-303. 6. Fang F, Lee AP, Yu CM. Left atrial function in heart failure with impaired and preserved ejection fraction. Curr Opin Cardiol 2014;29: 430-6. 7. Hsiao SH, Chiou KR. Left atrial expansion index predicts all-cause mortality and heart failure admissions in dyspnoea. Eur J Heart Fail 2013;15:1245-52.

Canadian Journal of Cardiology Volume - 2016 12. Lorenzi-Filho G, Azevedo ER, Parker JD, Bradley TD. Relationship of carbon dioxide tension in arterial blood to pulmonary wedge pressure in heart failure. Eur Respir J 2002;19:37-40. 13. Kasai T, Motwani SS, Yumino D, et al. Contrasting effects of lower body positive pressure on upper airways resistance and partial pressure of carbon dioxide in men with heart failure and obstructive or central sleep apnea. J Am Coll Cardiol 2013;61:1157-66. 14. Lang RM, Bierig M, Devereux RB, et al. Recommendations for chamber quantification: a report from the American Society of Echocardiography’s Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr 2005;18:1440-63. 15. Devereux RB. Detection of left ventricular hypertrophy by M-mode echocardiography. Anatomic validation, standardization, and comparison to other methods. Hypertension 1987;9:II19-26. 16. Zoghbi WA, Enriquez-Sarano M, Foster E, et al. Recommendations for evaluation of the severity of native valvular regurgitation with twodimensional and Doppler echocardiography. J Am Soc Echocardiogr 2003;16:777-802. 17. Haruki N, Takeuchi M, Kaku K, et al. Comparison of acute and chronic impact of adaptive servo-ventilation on left chamber geometry and function in patients with chronic heart failure. Eur J Heart Fail 2011;13: 1140-6. 18. Iwataki M, Takeuchi M, Otani K, et al. Measurement of left atrial volume from transthoracic three-dimensional echocardiographic datasets using the biplane Simpson’s technique. J Am Soc Echocardiogr 2012;25: 1319-26. 19. Okamatsu K, Takeuchi M, Nakai H, et al. Effects of aging on left atrial function assessed by two-dimensional speckle tracking echocardiography. J Am Soc Echocardiogr 2009;22:70-5. 20. Wu VC, Takeuchi M, Kuwaki H, et al. Prognostic value of LA volumes assessed by transthoracic 3D echocardiography: comparison with 2D echocardiography. JACC Cardiovasc Imaging 2013;6:1025-35. 21. Chadha TS, Watson H, Birch S, et al. Validation of respiratory inductive plethysmography using different calibration procedures. Am Rev Respir Dis 1982;125:644-9. 22. Sleep-related breathing disorders in adults: recommendations for syndrome definition and measurement techniques in clinical research. The Report of an American Academy of Sleep Medicine Task Force. Sleep 1999;22:667-89.

8. Rossi A, Temporelli PL, Quintana M, et al. Independent relationship of left atrial size and mortality in patients with heart failure: an individual patient meta-analysis of longitudinal data (MeRGE Heart Failure). Eur J Heart Fail 2009;11:929-36.

23. Rechtschaffen A, Kales A. A Manual of Standardized Terminology, Techniques and Scoring System for Sleep Stages of Human Subjects. NIH Publication No 204. Bethesda, MD: US National Institute of Neurological Diseases and Blindness, Neurological Information Network, 1968.

9. Pellicori P, Zhang J, Lukaschuk E, et al. Left atrial function measured by cardiac magnetic resonance imaging in patients with heart failure: clinical associations and prognostic value. Eur Heart J 2015;36:733-42.

24. Bradley TD, Logan AG, Kimoff RJ, et al. Continuous positive airway pressure for central sleep apnea and heart failure. N Engl J Med 2005;353:2025-33.

10. Mauser W, Sandrock S, Demming T, Kotzott L, Bonnemeier H. Sleep disordered breathing is an independent risk factor for left atrial enlargement in patients with congestive heart failure. Int J Cardiol 2013;167: 2323-4.

25. Oliveira W, Campos O, Bezerra Lira-Filho E, et al. Left atrial volume and function in patients with obstructive sleep apnea assessed by real-time three-dimensional echocardiography. J Am Soc Echocardiogr 2008;21: 1355-61.

11. Solin P, Bergin P, Richardson M, et al. Influence of pulmonary capillary wedge pressure on central apnea in heart failure. Circulation 1999;99: 1574-9.

26. Kasai T, Yumino D, Redolfi S, et al. Overnight effects of obstructive sleep apnea and its treatment on stroke volume in patients with heart failure. Can J Cardiol 2015;31:832-8.

Haruki et al. Sleep Apnea and Left Atrial Phasic Function 27. Calvin AD, Somers VK, Johnson BD, Scott CG, Olson LJ. Left atrial size, chemosensitivity, and central sleep apnea in heart failure. Chest 2014;146:96-103. 28. Lanfranchi PA, Braghiroli A, Bosimini E, et al. Prognostic value of nocturnal Cheyne-Stokes respiration in chronic heart failure. Circulation 1999;99:1435-40. 29. Vazir A, Hastings PC, Dayer M, et al. A high prevalence of sleep disordered breathing in men with mild symptomatic chronic heart failure due to left ventricular systolic dysfunction. Eur J Heart Fail 2007;9: 243-50. 30. Oldenburg O, Lamp B, Faber L, et al. Sleep-disordered breathing in patients with symptomatic heart failure: a contemporary study of prevalence in and characteristics of 700 patients. Eur J Heart Fail 2007;9: 251-7. 31. Tsang TS, Abhayaratna WP, Barnes ME, et al. Prediction of cardiovascular outcomes with left atrial size: is volume superior to area or diameter? J Am Coll Cardiol 2006;47:1018-23. 32. Hoit BD. Left atrial size and function: role in prognosis. J Am Coll Cardiol 2014;63:493-505. 33. Lancellotti P, Henri C. The left atrium: an old ‘barometer’ which can reveal great secrets. Eur J Heart Fail 2014;16:1047-8. 34. Wakami K, Ohte N, Asada K, et al. Correlation between left ventricular end-diastolic pressure and peak left atrial wall strain during left ventricular systole. J Am Soc Echocardiogr 2009;22:847-51.

9

38. Sin DD, Fitzgerald F, Parker JD, et al. Risk factors for central and obstructive sleep apnea in 450 men and women with congestive heart failure. Am J Respir Crit Care Med 1999;160:1101-6. 39. Toh N, Kanzaki H, Nakatani S, et al. Left atrial volume combined with atrial pump function identifies hypertensive patients with a history of paroxysmal atrial fibrillation. Hypertension 2010;55:1150-6. 40. Mullens W, Borowski AG, Curtin RJ, Thomas JD, Tang WH. Tissue Doppler imaging in the estimation of intracardiac filling pressure in decompensated patients with advanced systolic heart failure. Circulation 2009;119:62-70. 41. Yancy CW, Jessup M, Bozkurt B, et al. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on practice guidelines. Circulation 2013;128:e240-327. 42. Cowie MR, Woehrle H, Wegscheider K, et al. Adaptive servo-ventilation for central sleep apnea in systolic heart failure. N Engl J Med 2015;373: 1095-105. 43. Magalang UJ, Pack AI. Heart failure and sleep-disordered breathingethe plot thickens. N Engl J Med 2015;373:1166-7. 44. Hsiao SH, Chiou KR. Left atrial expansion index predicts atrial fibrillation in dyspnea. Circ J 2013;77:2712-21.

35. Tkacova R, Wang H, Bradley TD. Night-to-night alterations in sleep apnea type in patients with heart failure. J Sleep Res 2006;15:321-8. 36. Tamura A, Kawano Y, Kadota J. Carvedilol reduces the severity of central sleep apnea in chronic heart failure. Circ J 2009;73:295-8. 37. Triposkiadis F, Moyssakis I, Hadjinikolaou L, et al. Left atrial systolic function is depressed in idiopathic and preserved in ischemic dilated cardiomyopathy. Eur J Clin Invest 1999;29:905-12.

Supplementary Material To access the supplementary material accompanying this article, visit the online version of the Canadian Journal of Cardiology at www.onlinecjc.ca and at http://dx.doi.org/10. 1016/j.cjca.2016.02.047.