European Journal of Heart Failure:

Adaptive servo-ventilation for central sleep apnoea in systolic heart failure: results of the major substudy of SERVE-HF

Martin R. Cowie1*, Holger Woehrle2,3, Karl Wegscheider4, Eik Vettorazzi4, Susanne Lezius4, Wolfgang Koenig5,6,7, Frank Weidemann8,9, Gillian Smith10, Christiane Angermann11, Marie-Pia d’Ortho12, Erland Erdmann13, Patrick Levy14, Anita K. Simonds, MD10, Virend K. Somers15, Faiez Zannad16, Helmut Teschler17

1Imperial College London, London, United Kingdom; 2ResMed Science Center, ResMed Germany Inc., Martinsried, Germany; 3Sleep and Ventilation Center Blaubeuren, Respiratory Center Ulm, Ulm, Germany; 4Department of Medical Biometry and Epidemiology, University Medical Center Eppendorf, Hamburg, Germany; 5Deutsches Herzzentrum München, Technische Universität München, Munich, Germany; 6DZHK (German Centre for Cardiovascular Research), partner site Munich Heart Alliance, Munich, Germany; 7Department of Internal Medicine II-Cardiology University of Ulm Medical Center, Ulm, Germany; 8Department of Medicine I, University and University Hospital Würzburg, Würzburg, Germany; 9Katharinen-Hospital Unna, Unna, Germany; 10Royal Brompton Hospital, London, United Kingdom; 11Comprehensive Heart Failure Center, University Hospital and University of Würzburg, Würzburg, Germany; 12University Paris Diderot, Sorbonne Paris Cité, Hôpital Bichat, Explorations Fonctionnelles, DHU FIRE, AP-HP, Paris, France; 13Heart Center, University of Cologne, Cologne, Germany; 14CHU de Grenoble, Grenoble, France; 15Mayo Clinic and Mayo Foundation, Rochester, Minnesota, USA; 16Inserm, Université de Lorraine, CHU Nancy, France; 17Department of Pneumology, Ruhrlandklinik, West German Lung Center, University Hospital Essen, University Duisburg-Essen, Essen, Germany

*Corresponding author: Martin R. Cowie, Professor of Cardiology, Imperial College London, London, United Kingdom; Email: m.cowie@imperial.ac.uk

ABSTRACT

Aims The SERVE-HF trial investigated the impact of treating central sleep apnoea (CSA) with adaptive servo ventilation (ASV) in patients with systolic heart failure. A preplanned substudy was conducted to provide insight into mechanistic changes underlying the observed effects of ASV, including assessment of changes in left ventricular function, ventricular remodelling, and cardiac, renal and inflammatory biomarkers.

Methods and results In a subset of the 1325 randomised patients, echocardiography and biomarker analysis were performed at baseline, and 3 and 12 months. In secondary analyses, data for patients with baseline and 12-month values were evaluated. 312 patients participated in the substudy. The primary endpoint, change in echocardiographically-determined left ventricular ejection fraction from baseline to 12 months, did not differ significantly between the ASV and control groups. There were also no significant between-group differences for changes in left ventricular dimensions, wall thickness, diastolic function or right ventricular dimensions, and ejection fraction (echocardiography). Plasma N-terminal proB-type natriuretic peptide concentration decreased in both groups, and values were similar at 12 months. There were no significant between-group differences in changes in cardiac, renal and systemic inflammation biomarkers.

Conclusions The SERVE-HF substudy did not identify any effect of ASV on cardiac structure and function, cardiac biomarkers, renal function and systemic inflammation over 12 months in patients with systolic HF and CSA. These findings suggest that the increased cardiovascular mortality reported in SERVE-HF may not be related to adverse remodelling or worsening heart failure.

Keywords: heart failure; central sleep apnoea; adaptive servo-ventilation; cardiac function; biomarkers

Introduction

Sleep-disordered breathing can have a negative effect on cardiac function in patients with heart failure (HF).1 The prevalence of central sleep apnoea (CSA) in these patients is 25–40%,2 with rates increasing in parallel with HF severity.3 In addition, the presence of CSA in patients with HF is a marker of worse prognosis.4-6

Adaptive servo-ventilation (ASV) is a type of positive airway pressure that is particularly effective in ameliorating CSA.7 Some smaller studies, not all randomised, confirmed by later meta-analysis, have reported an improvement in left ventricular ejection fraction (LVEF) in patients with heart failure being treated with ASV.8-11 ASV has also been shown to improve levels of the biomarker B-type natriuretic peptide (BNP)7,12,13 and functional status9 in heart failure patients with central apnoea.

The SERVE-HF (Treatment of Sleep-Disordered Breathing With Predominant Central Sleep Apnea by Adaptive Servo-Ventilation in Patients With Heart Failure) trial investigated the effects of adding ASV to guideline-based medical management on survival and cardiovascular outcomes in patients with HF with reduced ejection fraction (HFrEF) and predominant CSA.14 The study did not show a difference between the ASV and control groups for the primary endpoint – a composite of time to first event of death from any cause, lifesaving cardiovascular intervention (transplantation, implantation of a long-term ventricular assist device, resuscitation after sudden cardiac arrest, or appropriate lifesaving shock) or unplanned hospitalisation for worsening HF – but there was a statistically and clinically significant increased risk of all-cause, and cardiovascular, mortality in the ASV versus control group.14

On the basis of existing data,8-11 including a meta-analysis,15 it was felt that the most likely mechanism of benefit of ASV in patients with HF would be improvements in left ventricular geometry. Therefore, SERVE-HF was designed to include a major substudy (NCT01164592),16 with the aim of providing insight into mechanistic changes underlying outcomes data for a subset of study participants; the primary endpoint was change in LVEF from baseline to 12 months. This paper reports cardiac function and biomarker results from the SERVE-HF substudy.

Methods

Study population

Of 91 centres participating in SERVE-HF, 29 contributed patients to the major substudy. Participants in the substudy were a subgroup of those enrolled in the SERVE-HF trial. Inclusion and exclusion criteria have been reported in detail previously.14,16 Briefly, patients were aged ≥22 years and had symptomatic chronic heart failure (New York Heart Association [NYHA] class III or IV, or class II with ≥1 HF-related hospitalisation in the previous 24 months) and reduced ejection fraction (LVEF ≤45%). All were receiving stable, guideline-based medical treatment for HF. With respect to sleep-disordered breathing, subjects had predominant CSA defined as an apnoea-hypopnoea index (AHI) of ≥15/h, with >50% central events and a central AHI of ≥10/h), derived from full polysomnography (PSG; based on total sleep time), documented <4 weeks before randomisation, with flow measurements performed with a nasal cannula. Additional exclusion criteria were amyloidosis, hypertrophic cardiomyopathy or arteriovenous fistulas, and diuretic dosage more than doubled within the 4 weeks prior to randomisation.

The substudy protocol was approved by the appropriate local or regional ethics committees (110420d/110420f [Adelaide], 2011-06-303 [Brisbane], HREC-D 153-11 [Melbourne], HPH323 [Perth], HREC/11/WMEAD/124 [Sydney], 27PZT/2012 [Czech Republic], H-D-2008-034 [Denmark], 293/13/03/01/2011 [Finland], 08-RESM-1 [France], 010/1553 [Germany], AA11 [Netherlands], 2009/2083/REK vest [Norway], dnr M38-08 [Sweden], Rif CE 2581 [Switzerland], 08/H1307/41 [United Kingdom]). The trial was conducted according to Good Clinical Practice and the Principles of the Declaration of Helsinki 2002. All participants gave written informed consent.

Study intervention

SERVE-HF participants were randomised to receive optimal medical therapy17 for HF alone, or in combination with ASV (Auto Set CS, ResMed). Full details of ASV titration and settings have been reported previously.14

Assessments and follow-up

Substudy evaluations were performed at baseline, and at 3 and 12 months after randomisation. These included echocardiography to determine left ventricular volumes, mass and ejection fraction.16 The substudy was completed when all 312 patients had been followed for 12 months.

Echocardiography was performed ≥3 hours after the end of ASV therapy according to standard operating procedures and using equipment that met predefined technical specifications; all study centres were required to undergo a qualification process prior to the study. All analyses were performed centrally at a core laboratory by experts blinded to treatment allocation. LVEF was determined using the modified biplane Simpson method, averaged over three beats. Tricuspid valve gradient was measured using continuous wave Doppler; IVRT, DT and E and A waves were measured using pulse wave Doppler, and septal and lateral mitral annulus motion was measured using tissue Doppler. Systolic pulmonary artery pressure was derived from peak tricuspid valve regurgitation jet velocity in combination with an estimated right atrial pressure from inferior vena cava diameter and changes with respiration. Blood was drawn at substudy visits, centrifuged locally and stored locally at –20°C or below, and then shipped frozen to the core lab where samples were stored at –80°C until biomarker analyses were performed.

Study outcomes

The primary endpoint of the substudy was the change in echocardiographically-determined LVEF from baseline to 12 months. Secondary endpoints included changes in left and right ventricular function, left ventricular systolic and diastolic indexed volumes, left and right ventricular mass, left ventricular sphericity index, LV end-systolic global wall stress, and biomarker levels (N-terminal pro BNP [NT-proBNP], soluble growth STimulation expressed gene 2 (sST2), interleukin-1 receptor-like 1, galectin-3, high-sensitivity [hs] troponin T and troponin I, neutrophil gelatinase-associated lipocalin [NGAL], cystatin C, creatinine, hs-C-reactive protein [CRP], tumour necrosis factor-α [TNF-α], ferritin, and leptin). These biomarkers were included to provide information about the heart failure syndrome, renal function and systemic inflammation.

Sample size

The primary endpoint was the difference in ventricular remodelling between the ASV and control groups from baseline to 12 months, as determined by echocardiographic measurement of LVEF. The substudy sample size calculation was based on the assumption that ASV treatment would increase the LVEF by 4% from baseline over 12 months and that there would be no change from baseline in LVEF in the control group (a 4% improvement in LVEF was assumed to be clinically meaningful). Residual standard deviation (SD) of measurement of was determined to be 11.5% based on published data 18, with α=0.05 and 1-β=0.80. Based on these inputs it was calculated that a total sample size of 240 evaluable patients would be required. With a conservative estimated drop-out rate of about 20%, the target sample size was 300 patients.

Statistical analyses

Analysis of the primary endpoint was performed using ANCOVA (analysis of covariance) with study group as the only factor and baseline LVEF as the only covariate. The study hypothesis was tested using the two-sided parameter coefficient t-test of the intervention group with a two-sided α=0.05. The primary analysis was performed in the intention-to-treat population, consisting of all substudy patients. Missing LVEF values (baseline and 12 months) were imputed for the primary analysis. A multiple imputation procedure with sequential imputation using chained equations (MICE), 49 baseline variables, and 20 repetitions was performed in the pooled data set. Alternative imputation methods were performed to study the sensitivity of the results to assumptions: multiple imputations only for 12-month values, Expectation-Maximisation [EM] algorithm to find maximum-likelihood estimators, and hot deck imputation. These imputation procedures were developed under the missing-at-random (MAR) assumption. However, because of the possibility that at least censoring due to sudden death may be non-random (i.e. informative on LVEF values), additional not MAR (NMAR) imputation procedures were applied (last observation carried forward [LOCF] potentially favouring non-survivors, upper half estimates and worst case estimates where the lowest observations were imputed for missing values potentially favouring survivors and compliant patients, and a mixture of worst case for non-survivors and LOCF for other missing values). An analysis of complete cases only, without any imputation, was also performed. To check the plausibility of NMAR imputation, 12-month LVEF changes from baseline were groupwise compared between survivors and non-survivors who died after more than 12 months in the trial.

For analyses of changes in the secondary endpoints at 12 months, the ANCOVA model of the primary analysis was used analogously. Because we did not impute missing values for secondary outcomes, only data for patients with baseline and 12-month values (variablewise complete cases) were evaluated. Biomarker changes were log-transformed, analysed and back transformed as percentages to the original scale. Mixed models with random intercept for patient and allowing for first order autoregression were used to visualise differences between groups over the course of the study (baseline, 3 months and 12 months).

Results

Patients

SERVE-HF centres in Germany (recruiting 246 patients), France (16 patients), Finland (7 patients), United Kingdom (3 patients), Australia (29 patients), Czech Republic (7 patients), Switzerland (3 patients) and Netherlands (1 patient) contributed patients to the substudy (total n=312). Table 1 details baseline demographics and characteristics for substudy patients. There were no significant differences in baseline characteristics between the control and ASV groups (Table 1). Baseline data for the substudy and overall SERVE-HF study populations are shown in Table S1. In centres contributing to the substudy, patients had a slightly higher BMI or NYHA class, and were more likely to have diabetes and to be receiving aldosterone antagonists. A CONSORT diagram showing patient flow in the substudy is provided in Figure 1. The number of patients who had data available for analysis varied for each parameter (range 82–249). Endpoint event rates in substudy participants were similar to those in non-participants (Table S2).

Echocardiography

A small increase in LVEF over the study period was seen in both treatment groups (Table 2). The primary endpoint parameter, change in LVEF from baseline to 12 months, did not differ significantly (p=0.222) between the ASV and control groups (Table 2, Figure S1). This result was robust when different imputation methods were used which assume MAR; even with methods that assume NMAR no significant differences were seen. However, with the most extreme imputations of low LVEF in non-survivors, differences as large as 4% in favour of control were within the 95% confidence intervals and therefore cannot be excluded (Table 2). Similarly, ASV and control group patients who died after more than 12 months showed a 12-month LVEF increase of 0.7% and 4.4%, respectively, resulting in a between-groups difference of –3.7% (95% CI –8.4-1.1) in favour of control, while LVEF increases were similar in surviving patients (by 2.2% and 2.9% for control and ASV; interaction p=0.264). There were also no significant differences between the two groups with respect to changes in left ventricular dimensions, wall thickness or measures of diastolic function or right ventricular dimensions and tricuspid annular plane systolic excursion (TAPSE) over the 12-month follow-up period (Table 3). The adjusted analysis also did not show any significant between-group differences.

Biomarkers

Reduction in plasma amino terminal pro B-type natriuretic peptide (NT-proBNP) concentration was seen in both the ASV and control groups, with no significant between-group difference at 12 months (Table 5, Figure S2). There were no significant differences between treatment groups in changes in troponin-T, troponin-I, sST-2, galectin-3, cystatin C, creatinine, NGAL, hs-CRP and TNF-α (Table 5). There were also no significant between-group differences in the adjusted analysis.

Discussion

In this analysis of SERVE-HF major substudy data, there were no statistically significant differences between the ASV and control groups with respect to changes in echocardiographically-determined cardiac structure or function, or a wide variety of cardiac, renal and systemic biomarkers. Although there were improvements in some of the parameters assessed during ASV therapy (e.g. LVEF and NT-proBNP), similar improvements also occurred in the control group, i.e. none of the between-group differences reached statistical significance. These findings suggest that any changes in the heart failure syndrome during the first year of SERVE-HF were of similar magnitude in the ASV and control groups. This would seem to be in keeping with the lack of effect of ASV on both general and disease-specific quality of life in the main SERVE-HF study, along with a lack of difference in HF-related hospitalisations between the ASV and control groups.14 The substudy findings are also in alignment with the results of multistate modelling analysis, which did not show any increase in heart failure hospitalisations and suggested that the mortality risk in SERVE-HF patients allocated to ASV is most apparent for cardiovascular death without preceding hospitalisation and therefore likely sudden (cardiac) death.19

The SAVIOR-C trial is another randomised clinical trial (RCT), albeit smaller and shorter in duration, of ASV in patients with HFrEF receiving guideline-based medical therapy, irrespective of the presence or severity of sleep-disordered breathing.20 As was the case in the major substudy of SERVE-HF, SAVIOR-C showed that LVEF improved from baseline in both the ASV and control groups, with no significant between-group differences. There were also no significant differences between the ASV and control groups in the change from baseline in all echocardiographic parameters and in plasma NT-proBNP concentration, consistent with our results. Also consistent with the main SERVE-HF study was a lack of significant difference between ASV and control with respect to disease-specific quality of life, although SAVIOR-C reported a significantly greater improvement in NYHA class in the ASV versus control group.

Although ASV did not improve left ventricular structure and function in HFrEF patients enrolled in the SERVE-HF study, these parameters have been modified by other forms of treatment for patients with HF. Compared with conventional therapy, use of cardiac resynchronisation therapy (CRT) has been associated with significant improvements in LVEF, and left ventricular end-diastolic and end-systolic volumes in patients with mild to moderate21-23 or moderate to severe24,25 heart failure. In terms of guideline-based optimal heart failure therapy, both ACE inhibitors and beta-blockers have been shown to improve left ventricular function.26,27 Beneficial effects on biomarkers have also been documented after successful CRT28 and in patients receiving medical heart failure therapies.29

SERVE-HF showed an increase in all-cause and cardiovascular mortality explained by an increase in sudden (presumed cardiac) death. There are no reliable biomarkers of sudden cardiac death risk. Our data suggest that the increased risk of cardiovascular mortality is not likely to be explained by adverse remodelling or worsening of the HF syndrome as assessed from the perspectives of NT-proBNP, renal function and systemic inflammation.

The lack of effect of ASV on LVEF in RCTs including HF patients with CSA is in contrast to positive effects of positive airway pressure therapy on left ventricular function that have been reported in patients with HF and obstructive sleep apnoea (OSA). Treatment with CPAP was associated with significant improvements in LVEF and quality of life versus control over one month in a randomised controlled study of 24 patients,30 and significantly improved LVEF and significantly decreased heart rate, systolic blood pressure and the left ventricular end-systolic dimension over 3 months in another randomised study of 55 patients.31

These data are from a substudy of a larger trial, but the study protocol was prespecified and analyses were conducted prospectively. The substudy population was slightly sicker at baseline than the remainder of the SERVE -HF study population, but sufficiently similar to allow robust evaluation and comparison. However, the substudy was only powered to detect changes in LVEF, and therefore may not have had adequate statistical power to reliably compare other endpoints (e.g. cardiac biomarkers) between the ASV and control groups. The fact that only a subset of SERVE-HF patients participated in the major substudy meant that the sample size was smaller, particularly for some subgroup analyses. Furthermore, there are no data on changes in LVEF in patients who died before the 12-month follow-up. These missing data could potentially influence the substudy findings. Several sensitivity analyses were performed to account for this, which differ with respect to their underlying assumptions about the nature of missing data. None of these analyses identified statistically significant differences between the ASV and control groups but confidence intervals were wide. Therefore, definitive statements about the effect, or lack of effect, of ASV on LVEF and the potential contribution of this to the increase in cardiovascular deaths observed in SERVE-HF patients randomised to ASV cannot be made.

In conclusion, the SERVE-HF major substudy did not identify any effect of ASV, either positive or negative, on cardiac structure and function, cardiac biomarkers, renal function and systemic inflammation over a 12-month period in patients with HFrEF. This suggests that adverse remodelling or worsening of the heart failure syndrome may not be the mechanisms underlying the increased cardiovascular mortality reported in the main SERVE-HF study. The relative stability of the HF syndrome observed is consistent with the lack of change in HF-related hospitalisations reported in the parent study. While the possibility of longer term effects during the overall follow-up of the SERVE-HF study cannot be excluded, the HF syndrome is more likely to show progressive decline that would be accompanied by changes in biomarkers such as BNP, which were not seen in this substudy. However, factors contributing to the increased risk of sudden (presumably cardiac) death remain to be identified. Additional large prospective randomised studies are required in this area.

Funding: This work was supported by ResMed. Representatives and scientists from the ResMed participated in the study including design, data collation, data analysis, and critical review of the paper.

Conflicts of interest: M.R.C. reports receiving consulting fees from Servier, Novartis, Pfizer, St. Jude Medical, Boston Scientific, Respicardia, and Medtronic and grant support through his institution from Bayer; H.W. was employed by ResMed during the conduct of the study, and reports receiving lecture fees from Vital Air, Boehringer Ingelheim and ResMed, and research support from ResMed; K.W. receiving grant support from ResMed and personal fees from Biotronik; E.V., receiving grant support from ResMed; S.L., receiving grant support from ResMed; W.K. Koenig receiving personal fees from AstraZeneca, Novartis, MSD, Amgen, Actavis, Novartis, Pfizer, The Medicines Company, GSK and Sanderling Ventures, grants and non-financial support from Abbott, Roche Diagnostics, Beckmann and Singulex; F.W., receiving lecture fees from ResMed; G.S., receiving lecture fees from ResMed; C.A., receiving grants, personal fees and non-financial support from ResMed, grants from Novartis, personal fees from Servier, grants and non-financial support from Thermo Fischer, grants and personal fees from Boehringer, grants and personal fees from Lundbeck, and grants and personal fees from Vifor; M-P.d’O., receiving fees for serving on advisory boards from ResMed and IP Santé, lecture fees from ResMed, Philips, IP Santé, and VitalAire, grant support from Fisher and Paykel Healthcare, ResMed, Philips, ADEP Assistance, and IP Santé, and small material donations from VitalAire; E.E., P.L. and A.K., no potential conflicts of interest outside the submitted work; V.S., receiving consulting fees from PriceWaterhouseCoopers, Sorin, GlaxoSmithKline, Respicardia, uHealth, Ronda Grey, Dane Garvin, Philips, Biosense Webster, Philips Respironics and ResMed, working with Mayo Medical Ventures on intellectual property related to sleep and cardiovascular disease, and having a pending patent (12/680073) related to biomarkers of sleep apnoea; F.Z., receiving personal fees from ResMed, Janssen, Bayer, Pfizer, Novartis, Boston Scientific, Takeda, Amgen and CVRx; and H.T., receiving consulting fees, grant support, and hardware and software for the development of devices from ResMed. The majority of these conflicts of interest are outside the submitted work.

Acknowledgements: The authors thank CRI (the clinical research institute) for their expertise in overseeing the SERVE-HF trial. Medical writing support was provided by Nicola Ryan, independent medical writer, and statistical calculations were supported by Anika Buchholz, Christine Eulenburg, and Anna Suling, funded by ResMed. MRC and AKS’s salaries are supported by the National Institute for Health Research Cardiovascular and Respiratory Biomedical Research Units, respectively, at the Royal Brompton Hospital, London, UK. VKS was supported by NIH R01HL065176. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Supplementary Information

Additional Supporting Information may be found in the online version of this article.

References

1.Pearse SG, Cowie MR. Sleep-disordered breathing in heart failure. Eur J Heart Fail 2016; 18(4): 353-61.

2.Levy LM, Pepin J-L, Tamisier R, Neuder Y, Baguet J-P, Javaheri S. Prevalence and impact of central sleep apnea in heart failure. Sleep Med Clin 2007; 2: 615-21.

3.Oldenburg O, Lamp B, Faber L, Teschler H, Horstkotte D, Topfer V. 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(3): 251-7.

4.Bitter T, Westerheide N, Prinz C, et al. Cheyne-Stokes respiration and obstructive sleep apnoea are independent risk factors for malignant ventricular arrhythmias requiring appropriate cardioverter-defibrillator therapies in patients with congestive heart failure. Eur Heart J 2011; 32(1): 61-74.

5.Javaheri S, Shukla R, Zeigler H, Wexler L. Central sleep apnea, right ventricular dysfunction, and low diastolic blood pressure are predictors of mortality in systolic heart failure. J Am Coll Cardiol 2007; 49(20): 2028-34.

6.Wang H, Parker JD, Newton GE, et al. Influence of obstructive sleep apnea on mortality in patients with heart failure. J Am Coll Cardiol 2007; 49(15): 1625-31.

7.Teschler H, Dohring J, Wang YM, Berthon-Jones M. Adaptive pressure support servo-ventilation: a novel treatment for Cheyne-Stokes respiration in heart failure. Am J Respir Crit Care Med 2001; 164(4): 614-9.

8.Hetland A, Haugaa KH, Olseng M, et al. Three-month treatment with adaptive servoventilation improves cardiac function and physical activity in patients with chronic heart failure and cheyne-stokes respiration: a prospective randomized controlled trial. Cardiology 2013; 126(2): 81-90.

9.Kourouklis SP, Vagiakis E, Paraskevaidis IA, et al. Effective sleep apnoea treatment improves cardiac function in patients with chronic heart failure. Int J Cardiol 2013; 168(1): 157-62.

10.Miyata M, Yoshihisa A, Suzuki S, et al. Adaptive servo ventilation improves Cheyne-Stokes respiration, cardiac function, and prognosis in chronic heart failure patients with cardiac resynchronization therapy. J Cardiol 2012; 60(3): 222-7.

11.Oldenburg O, Schmidt A, Lamp B, et al. Adaptive servoventilation improves cardiac function in patients with chronic heart failure and Cheyne-Stokes respiration. Eur J Heart Fail 2008; 10(6): 581-6.

12.Arzt M, Schroll S, Series F, et al. Auto-servoventilation in heart failure with sleep apnoea: a randomised controlled trial. Eur Respir J 2013; 42(5): 1244-54.

13.Randerath WJ, Nothofer G, Priegnitz C, et al. Long-term auto-servoventilation or constant positive pressure in heart failure and coexisting central with obstructive sleep apnea. Chest 2012; 142(2): 440-7.

14.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(12): 1095-105.

15.Sharma BK, Bakker JP, McSharry DG, Desai AS, Javaheri S, Malhotra A. Adaptive servoventilation for treatment of sleep-disordered breathing in heart failure: a systematic review and meta-analysis. Chest 2012; 142(5): 1211-21.

16.Cowie MR, Woehrle H, Wegscheider K, et al. Rationale and design of the SERVE-HF study: treatment of sleep-disordered breathing with predominant central sleep apnoea with adaptive servo-ventilation in patients with chronic heart failure. Eur J Heart Fail 2013; 15(8): 937-43.

17.Ponikowski P, Voors AA, Anker SD, et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: The Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur J Heart Fail 2016; 18(8): 891-975.

18.Hastings PC, Vazir A, Meadows GE, et al. Adaptive servo-ventilation in heart failure patients with sleep apnea: a real world study. Int J Cardiol 2010; 139(1): 17-24.

19.Eulenburg C, Wegscheider K, Woehrle H, et al. Mechanisms underlying increased mortality risk in patients with heart failure and reduced ejection fraction randomly assigned to adaptive servoventilation in the SERVE-HF study: results of a secondary multistate modelling analysis. Lancet Respir Med 2016: 30 Aug [Epub ahead of print].

20.Seino Y, Momomura S, Kihara Y, et al. Effects of adaptive servo-ventilation therapy on cardiac function and remodeling in patients with chronic heart failure (SAVIOR-C): study protocol for a randomized controlled trial. Trials 2015; 16: 14.

21.Abraham WT, Young JB, Leon AR, et al. Effects of cardiac resynchronization on disease progression in patients with left ventricular systolic dysfunction, an indication for an implantable cardioverter-defibrillator, and mildly symptomatic chronic heart failure. Circulation 2004; 110(18): 2864-8.

22.Daubert C, Gold MR, Abraham WT, et al. Prevention of disease progression by cardiac resynchronization therapy in patients with asymptomatic or mildly symptomatic left ventricular dysfunction: insights from the European cohort of the REVERSE (Resynchronization Reverses Remodeling in Systolic Left Ventricular Dysfunction) trial. J Am Coll Cardiol 2009; 54(20): 1837-46.

23.Linde C, Abraham WT, Gold MR, et al. Randomized trial of cardiac resynchronization in mildly symptomatic heart failure patients and in asymptomatic patients with left ventricular dysfunction and previous heart failure symptoms. J Am Coll Cardiol 2008; 52(23): 1834-43.

24.Abraham WT, Fisher WG, Smith AL, et al. Cardiac resynchronization in chronic heart failure. N Engl J Med 2002; 346(24): 1845-53.

25.Cleland JG, Daubert JC, Erdmann E, et al. The effect of cardiac resynchronization on morbidity and mortality in heart failure. N Engl J Med 2005; 352(15): 1539-49.

26.Hall SA, Cigarroa CG, Marcoux L, Risser RC, Grayburn PA, Eichhorn EJ. Time course of improvement in left ventricular function, mass and geometry in patients with congestive heart failure treated with beta-adrenergic blockade. J Am Coll Cardiol 1995; 25(5): 1154-61.

27.Pfeffer MA, Braunwald E, Moye LA, et al. Effect of captopril on mortality and morbidity in patients with left ventricular dysfunction after myocardial infarction. Results of the survival and ventricular enlargement trial. The SAVE Investigators. N Engl J Med 1992; 327(10): 669-77.

28.Dong YX, Burnett JC, Jr., Chen HH, et al. Effect of cardiac resynchronization therapy on broad neurohormone biomarkers in heart failure. J Interv Card Electrophysiol 2011; 30(3): 241-9.

29.Ordu S, Yildiz BS, Alihanoglu YI, et al. Effects of ivabradine therapy on heart failure biomarkers. Cardiol J 2015; 22(5): 501-9.

30.Mansfield DR, Gollogly NC, Kaye DM, Richardson M, Bergin P, Naughton MT. Controlled trial of continuous positive airway pressure in obstructive sleep apnea and heart failure. Am J Respir Crit Care Med 2004; 169(3): 361-6.

31.Kaneko Y, Floras JS, Usui K, et al. Cardiovascular effects of continuous positive airway pressure in patients with heart failure and obstructive sleep apnea. N Engl J Med 2003; 348(13): 1233-41.

Legends

Figure 1. CONSORT diagram showing patient flow in the SERVE-HF substudy. ASV, adaptive servo-ventilation; Echo, echocardiography; PSG, polysomnography.

Appendix: SERVE-HF core laboratories and study centres

Core laboratories

Analysis and quality control of echocardiography: Med Klinik und Poliklinik I der Universität Würzburg, Klinikstraße 6-8 97070, Würzburg, Germany.

Analysis and quality control of biomarkers: Department of Internal Medicine II-Cardiology, University of Ulm Medical Center, Ulm, Germany.

Study centres

The following is a list of sites for the SERVE-HF study. Each centre included at least two sites (i.e. referring cardiologist[s] and sleep laboratory).

Germany

Universitätsklinikum Schleswig-Holstein Campus Lübeck, Lübeck.

Charité Campus Mitte CCM, Berlin.

Praxis Dr Hein, Reinbek.

Klinikum der Universität Köln, Köln.

Facharztzentrum Dresden-Neustadt Gbr, Dresden.

Universitätsklinikum Aachen, Aachen.

Krankenhaus Reinbek St. Adolf-Stift, Reinbek.

Gemeinschaftspraxis, Drs Schmidt/Gronke, Würzburg.

Deutsches Zentrum für Herzinsuffizienz, CHFC, Universitätsklinikum Würzburg, Würzburg.

Florence Nightingale Krankenhaus, Düsseldorf.

Universität Leipzig – Herzzentrum, Leipzig.

Klinik Augustinum München, München.

Universitätsklinikum Münster/WWU, Münster.

DRK Krankenhaus, Alzey.

Malteser Krankenhaus St. Hildegardis, München.

Lungenärzte am Rotkreuzplatz, München.

Universitätsklinikum Schleswig-Holstein Campus, Lübeck.

Universitätsklinikum Hamburg-Eppendorf, Hamburg.

Missionsärztliche Klinik Würzburg, Würzburg.

Herzzentrum Universität Dresden, Dresden.

HDZ NRW, Bad Oeynhausen.

Helios Klinikum Borna, Borna.

Kardiologische Praxis Dr. Lodde, Dortmund.

AFPR e.V., Ruhrland-klinik Essen, Essen.

Kardiologische Praxis Dr. Furche, Herne.

Universitätsklinikum Freiburg, Freiburg.

Kardiologische Praxis Dr. Wetzel, Dortmund.

Augusta-Kranken-Anstalt gGmbH Thorax-zentrum Ruhrgebiet, Herne.

Herzzentrum Bad Krozingen, Bak Krozingen.

Kardiologische Praxis Dr. Schlichting, Herne.

Lungenklinik Herner, Herner.

Kardiologische Praxis Dr. Isburch, Castrop-Rauxel.

Asklepios Klinik Barmbek, Hamburg.

Praxis für Kardiologie Dr. med. Menz, Menden.

Kardiologische Praxis Marschner, Bonn.

Kardiologische Praxis Dr. Burkhard-Meier, Viersen.

Praxis Dr. Anselm Bäumer, Köln.

Gemeinschaftspraxis Drs Gysan/Heinzler/May, Köln.

Praxis Dr. Frölich, Ratingen.

Cardiopraxis Ingelheim, Ingelheim.

Praxis Dr. Tekiyeh, Essen.

Gemeinschaftspraxis Drs Leischik/Littwitz, Hagen.

Universitätsklinikum Heidelberg, Heidelberg.

Thoraxklinik Heidelberg gGmbH, Heidelberg.

Praxis für Lunge, Herz un Schlaf, Bielefeld.

Kardiologische Praxis Dr. Schön, Mühldorf.

Praxis Dr. Gerritsen, Waldkraiburg.

Gemeinschaftspraxis Drs Dobler/Turin, Karlstadt.

Kath. Klinikum Essen/Philippusstift, Essen.

Universitätsklinikum Regensburg, Regensburg.

Jüdisches Krankenhaus Berlin, Berlin.

Kardiology Oberkassel, Düsseldorf.

Evangelisches Krankenhaus Mülheim, Mülheim/Ruhr.

Marienhospital Stuttgart, Stuttgart.

Kardiologische Praxis Ludwigsburg, Ludwigsburg.

Klinikum Dortmund gGmbH, Dortmund.

POLIKUM Friedenau, Berlin.

Herz- & Gefäßpraxis, Metzingen.

American Sleep Clinic, Frankfurt.

Kardiologische Gemeinschaftspraxis Dr. K Vorbeck, Wiesbaden.

St. Elisabeth-Hospital Herten gGmbH, Herten.

Gemeinschaftspraxis PD Dr. Lankisch, Düsseldorf.

Praxis gemeinschaft Frille, Berlin.

Kardiologische Praxis Nienburg, Nienburg.

Praxis Dr. Diedrichs, Frechen.

CardioVasculares Centrum Frankfurt, Frankfurt.

Praxis Dr. Hecker, Dortmund.

Clinical Trial Site Services, Dortmund.

Kardiologie Praxis Dr. Bonnekamp, Essen.

B&B GmbH, Herne.

Unfalkrankenhause Berlin, Berlin.

Schreiber Klinik München, München.

Klinik Kitzinger Land, Kitzingen.

Katharinen Hospital Unna, Unna.

Kardiologische Praxis Gütersloh, Gütersloh.

Praxis Dr. Hohensee, Dresden.

Evangelisches Krankenhaus, Düsseldorf.

Gemeinschaftspraxis Kardiologie Dr. Becker, Soest.

Bethanienkrankenhaus Moers, Moers.

Australia

Hollywood Specialist Centre (CVS), Perth.

Royal Adelaide Hospital, Adelaide.

Rivercity Private Hospital, Brisbane.

Melbourne Sleep Disorders Centre, Melbourne.

Westmead Hospital and Specialist Services, Sydney.

Finland

Tampere University Hospital, Pirkanmaa sairaanhoitopiiri, Tampere.

Helsinki University Central Hospital, Helsinki.

Unesta Research Centre, Tampere.

France

CHU Grenoble, Hôpital Michallon, Grenoble.

CHU de Poitiers, Poitiers.

United Kingdom

Brompton Hospital, London.

Netherlands

University Medical Centre Groningen, Groningen.

Czech Republic

St Anne’s University Hospital, Brno.