Issue Navigator

Volume 12 No. 01
Earn CME
Accepted Papers

Journal Club

Adaptive Servo-Ventilation and Central Apnea Associated with Systolic Heart Failure

Shirin Shafazand, MD, MS1; M. Safwan Badr, MD, MBA2
1Associate Professor of Medicine, Pulmonary, Critical Care and Sleep Medicine, University of Miami, Miller School of Medicine; 2Professor and Chief, Pulmonary Critical Care and Sleep Medicine, Wayne State University, School of Medicine

Summary and Commentary on Cowie et al. Adaptive servo-ventilation for central sleep apnea in systolic heart failure. N Engl J Med 2015;373:1095–1105.



In individuals with symptomatic congestive heart failure (left ventricular ejection fraction ≤ 45%) and predominately central sleep apnea, does treatment of the sleep apnea with adaptive servo-ventilation (ASV), in addition to guideline based medical treatment change time to first event (composite of first event of death from any cause, lifesaving cardiovascular intervention [cardiac transplantation, implantation of a ventricular assist device, resuscitation after sudden cardiac arrest, or appropriate lifesaving shock], or unplanned hospitalization for worsening heart failure)?



Multi-center, randomized, controlled trial; identifier: NCT00733343.


Randomization was conducted by a centralized computer system. Allocation was stratified by country. The randomization sequence was concealed.


The investigators and participants were not blinded to study arm assignment. The primary outcome measurements were blinded.

Follow-up period

This was an event driven analysis. Participants were recruited over a period of 60 months with a total of 2-year follow-up until the target 651 observed events was reached.


Patients were recruited from each investigator's network of sleep and cardiology practices.


INCLUSION CRITERIA: (1) Age ≥ 22 years, (2) Left ventricular ejection fraction of ≤ 45%, (3) New York Heart Association (NYHA) class III or IV heart failure or NYHA class II heart failure with at least one heart failure related hospitalization within the 24 months prior to randomization, (4) Stable medical therapy, and (5) Predominantly central sleep apnea (AHI ≥ 15 events per hour, with > 50% central events [apnea or hypopnea] and a central AHI ≥ 10 events per hour).

EXCLUSION CRITERIA: (1) Significant chronic obstructive pulmonary disease (COPD) with a forced expiratory volume in 1 second (FEV1) < 50% of predicted, (2) Oxygen saturation ≤ 90% at rest during the day, (3) Current use of positive airway pressure (PAP) therapy, (4) Life expectancy < 1 year for diseases unrelated to chronic HF, (5) Cardiac surgery, percutaneous coronary intervention, myocardial infarction or unstable angina within the previous 6 months, (6) Transient ischemic attack or stroke within the previous 3 months, (7) Hemodynamically significant uncorrected valvular heart disease, (8) untreated restless legs syndrome, (9) Pregnancy, and (10) Contraindication to the use of AutoSet device.


Patients meeting eligibility criteria were randomized to either control (optimal medical treatment alone) or active treatment (optimal medical treatment plus use of ASV) in a 1:1 ratio. ASV default settings were used: expiratory positive airway (EPAP) pressure of 5 cm of water; minimum pressure support of 3 cm of water; and maximum pressure support of 10 cm of water. EPAP was increased manually in a monitored setting to control obstructive sleep apnea, and the maximum pressure support was increased to control central sleep apnea. A full face mask was recommended for the initiation of ASV. Patients were advised to use the ASV device for at least 5 hours per night, 7 days per week. Adherence to therapy was objectively measured. The target was to reduce the AHI to < 10 events per hour within 14 days of starting ASV. Clinic visits took place at baseline, after 2 weeks, at 3 and 12 months, and every 12 months until the end of the study.


The primary endpoint was defined as a composite endpoint of time to first event of death from any cause, lifesaving cardiovascular intervention (cardiac transplantation, implantation of a ventricular assist device, resuscitation after sudden cardiac arrest, or appropriate lifesaving shock), or unplanned hospitalization for worsening heart failure. The first secondary outcome was the hierarchical primary endpoint if the null hypothesis was rejected, except cardiovascular death was included instead of all-cause mortality. The second secondary outcome was the primary outcome except unplanned all cause hospitalization was included instead of heart failure specific hospitalization. Other secondary outcomes included were time to death from any cause, the time to death from cardiovascular causes, change in NYHA class change in the 6-minute walk distance, quality of life, and sleepiness measures.

This was a multicenter, international phase-IV study with parallel group design. A total of 651 events had to be collected for the study to have a power of 80% to show a 20% reduction in the rate of the first primary end-point event with ASV at an overall two-sided type I error rate of 5%. A planned total of 1,193 patients needed to be recruited over a period of 60 months to meet the target estimated total events. This was an intention-to-treat analysis.

Patient follow-up

1,325 patients were enrolled from February 2008 through May 2013 at 91 centers and included in the intention to treat analysis; 659 patients were assigned to the control group and 666 to the ASV group. The median follow-up duration was 31 months. In the ASV group, 82 subjects withdrew consent and 1 was lost to follow-up. In the control group, 73 participants withdrew consent and 8 were lost to follow-up.

Main Results

Over 90% of participants were male, mean age 69 years, and 69% were NYHA class III. There were no significant differences in baseline characteristics (including AHI, central AHI, and ODI) between the control group and the ASV group, except for the rate of anti-arrhythmic drug use, which was higher in the ASV group (P = 0.005). Sensitivity analyses showed that this baseline difference did not have an impact on study outcomes.

A total of 60% of the patients in the ASV group used device for an average of ≥ 3 hours per night. At 12 months, the mean AHI was 6.6 events per hour, and oxygen desaturation index (ODI) 3% was 8.6 events per hour. The median inspiratory positive airway (IPAP) pressure was 9.8 cm H2O and median EPAP pressure was 5.7 cm H2O at 12 months.

The incidence of the primary end point did not differ significantly between the ASV group and the control group (54.1% and 50.8%, respectively; hazard ratio, 1.13; 95% confidence interval [CI], 0.97 to 1.31; P = 0.10). All-cause mortality and cardiovascular mortality were significantly higher in the ASV group than in the control group (hazard ratio for death from any cause 1.28; 95% CI, 1.06 to 1.55; P = 0.01; and hazard ratio for cardiovascular death 1.34; 95% CI, 1.09 to 1.65; P = 0.006). In subgroup analyses, the signal for the primary end point was stronger in patients with a higher proportion of Cheyne-Stokes respiration and the signal for cardiovascular death was stronger in patients with very low left ventricular ejection fraction. Quality of life and NYHA class did not differ between the two groups at study end.


In predominately male subjects with central sleep apnea and an ejection fraction of ≤ 45%, randomized to ASV therapy in addition to guideline driven heart failure therapy or medical therapy alone, there was no statistically significant difference in a composite primary outcome of time to first event of death from any cause, lifesaving cardiovascular intervention (cardiac transplantation, implantation of a ventricular assist device, resuscitation after sudden cardiac arrest, or appropriate lifesaving shock), or unplanned hospitalization for worsening heart failure. However, all-cause and cardiovascular mortality were significantly higher in the ASV group than in the control group.

Sources of funding: The study was supported by ResMed and by grants from the National Institute for Health Research (NIHR) Cardiovascular Biomedical Research Unit (to Dr. Cowie), the NIHR Respiratory Biomedical Research Unit (Dr. Simonds), and the National Institutes of Health (R01HL065176 Dr. Somers).

For correspondence: Dr. Cowie at the Department of Cardiology, Imperial College London, Dovehouse St., London SW3 6LY, United Kingdom,


This significant study addresses a difficult clinical problem, namely central apnea in patients with congestive heart failure (CHF). Most previous studies and published guidelines have relied on intermediate outcomes, such as AHI, ejection fraction, or quality of life metrics.1,2 In contrast, this study tested a robust composite outcome combining death, lifesaving cardiovascular intervention, or unplanned hospitalization for worsening heart failure. Casting a wide net strengthened the results by capturing myriad relevant outcome measures. However, the surprising finding was that all-cause mortality and cardiovascular mortality were higher in the patients receiving ASV than in the control group, despite convincing efficacy in reducing AHI and central apnea index. This study provides another cautionary note against total reliance on intermediate outcome to drive treatment decisions. Several explanations have been put forward to explain this critical finding, but none provide a complete explanation for this unexpected finding.

First, the study population was predominantly males (90%), non-obese (BMI < 30) who had central sleep apnea (CSA; central AHI / total AHI > 80%). It is unclear if any of these characteristics influenced the response to ASV in this study. It is also unclear if the response was due to the specific algorithm used in the device, or to any mode of ASV.

One possible explanation is decreased cardiac output when positive pressure ventilation is used. This could occur in patients with decreased pre-load who are more vulnerable to positive pressure effects on venous return. However, previous studies, such as CANPAP, did not demonstrate increased mortality in the CPAP group.3 In fact, those who responded to CPAP, with decreased AHI, had better outcome as evidenced by increased ventricular ejection fraction and heart transplant-free survival if CSA was suppressed by CPAP therapy.4 The other intriguing possibility is that Cheyne-Stokes respiration may be an adaptive not a pathologic mechanism that should be ameliorated.5 It is also possible that participants who were randomized to ASV were overall sicker than the control group. The difference in the arrhythmic drug use between the two groups suggests that some baseline differences may have been present.

Another potential issue is the patient-machine interaction, and the relationship between mechanical inflation and neural inspiratory activity. Two patterns of interaction may be noted: inhibition6,7 and entrainment.8 Mechanical ventilation may result in inhibition of ventilatory motor output, via hypocapnia9; this inhibition would be masked in the presence of a backup respiratory rate. Mechanical ventilation could also cause inhibition of the ventilatory motor output, even under normocapnic conditions, through large tidal volume, respiratory frequency or their combination.7 In fact, a respiratory rate of 15/minute may promote inhibition of ventilatory motor output during prolonged mechanical ventilation. The second pattern of interaction is respiratory entrainment, which occurs when neural inspiratory activity and mechanical inflation are coupled. In other words, entrainment represents a matching of neural and mechanical outputs. There is evidence that entrainment occurs during mechanical ventilation in wakefulness and sleep.8 Entrainment may be a highly adaptive response by matching drive to the breath volume. Interestingly, the effect of ASV on the ventilatory motor output to respiratory muscles is unknown. However, it is unclear whether entrainment or inhibition occurs during ASV. Inhibition of respiratory muscle activity may have long-term deleterious effects on respiratory muscle strength and potentially on daytime ventilation.

Cheyne-Stokes respiration is associated with oscillating ventilation and ventilatory motor output, with proper matching of ventilatory drive and tidal volume during spontaneous breathing. In contrast, ASV provides pressure support that targets 90% of the patient's recent average ventilation. Thus, the delivered volume may be less or more that the intrinsic ventilatory drive for a given breath. What are the physiologic implications to delivering a small breath despite a high ventila-tory drive? Is this neuromechanical dissociation injurious? It is important to determine the effect of this modality on respiratory muscle function, especially in patients with CHF who may suffer from respiratory muscle dysfunction.10 These are unanswered questions that may require detailed physiologic studies to ascertain the effects of ASV on respiratory muscle strength and endurance.

The results of the current study are applicable only to ASV in patients with HF with reduced EF and should not be extrapolated to other conditions associated with central apnea, nor should they be extrapolated to other therapeutic modalities such as CPAP. Therefore, current recommendations regarding treatment of central apnea are unchanged.2 Initial treatment for central apnea associated with HF should emphasize optimal treatment of heart failure. In fact, it is important to emphasize that CPAP and ASV are fundamentally different modalities. Unlike ASV, CPAP is a pneumatic split and does not provide pressure-support ventilation. CPAP may mitigate post-apneic ventilatory overshoot by increasing lung volumes and decreasing loop gain.11 The majority of patients with central apnea have concomitant obstructive apnea,12 and many others demonstrate upper airway narrowing or occlusion during central apneic events.13 Therefore, a trial of CPAP therapy remains the treatment of choice. However, definitive studies addressing the long-term effectiveness of CPAP in patients with central apnea are still needed. Supplemental oxygen therapy14 is another physiologically appealing intervention; however, desaturation criteria must be met to ensure third-party coverage.

In conclusion, the failure of one modality in a very limited setting does not merit therapeutic nihilism in other settings. Instead, we should continue to use CPAP, BPAP, or ASV where indicated, and be prepared to update our practice as new exciting discoveries illuminate the road ahead.


The authors have indicated no financial conflicts of interest.


Shafazand S, Badr MS. Adaptive servo-ventilation and central apnea associated with systolic heart failure. J Clin Sleep Med 2016;12(1):147–150.



Badr MS, author. Central sleep apnea. Prim Care. 2005;32:361–74. vi. [PubMed]


Aurora RN, Chowdhuri S, Ramar K, et al., authors. The treatment of central sleep apnea syndromes in adults: practice parameters with an evidence-based literature review and meta-analyses. Sleep. 2012;35:17–40. [PubMed Central][PubMed]


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


Arzt M, Floras JS, Logan AG, et al., authors. Suppression of central sleep apnea by continuous positive airway pressure and transplant-free survival in heart failure: a post hoc analysis of the Canadian Continuous Positive Airway Pressure for Patients with Central Sleep Apnea and Heart Failure Trial (CANPAP). Circulation. 2007;115:3173–80. [PubMed]


Naughton MT, author. Cheyne-Stokes respiration: friend or foe? Thorax. 2012;67:357–60. [PubMed]


Simon PM, Dempsey JA, Landry DM, Skatrud JB, authors. Effect of sleep on respiratory muscle activity during mechanical ventilation. Am Rev Respir Dis. 1993;147:32–7. [PubMed]


Manchanda S, Leevers AM, Wilson CR, Simon PM, Skatrud JB, Dempsey JA, authors. Frequency and volume thresholds for inhibition of inspiratory motor output during mechanical ventilation. Respir Physiol. 1996;105:1–16. [PubMed]


Simon PM, Habel AM, Daubenspeck JA, Leiter JC, authors. Vagal feedback in the entrainment of respiration to mechanical ventilation in sleeping humans. J Appl Physiol. 2000;89:760–9. [PubMed]


Rowley JA, Zhou XS, Diamond MP, Badr MS, authors. The determinants of the apnea threshold during NREM sleep in normal subjects. Sleep. 2006;29:95–103. [PubMed]


Meyer FJ, Borst MM, Zugck C, et al., authors. Respiratory muscle dysfunction in congestive heart failure: clinical correlation and prognostic significance. Circulation. 2001;103:2153–8. [PubMed]


Sands SA, Edwards BA, Kee K, et al., authors. Loop gain as a means to predict a positive airway pressure suppression of Cheyne-Stokes respiration in patients with heart failure. Am J Respir Crit Care Med. 2011;184:1067–75. [PubMed]


Chowdhuri S, Ghabsha A, Sinha P, Kadri M, Narula S, Badr MS, authors. Treatment of central sleep apnea in U.S. veterans. J Clin Sleep Med. 2012;8:555–63. [PubMed Central][PubMed]


Badr MS, Toiber F, Skatrud JB, Dempsey J, authors. Pharyngeal narrowing/occlusion during central sleep apnea. J Appl Physiol. 1995;78:1806–15. [PubMed]


Franklin KA, Eriksson P, Sahlin C, Lundgren R, authors. Reversal of central sleep apnea with oxygen. Chest. 1997;111:163–9. [PubMed]