Author + information
- Received July 3, 2012
- Revision received October 30, 2012
- Accepted November 15, 2012
- Published online February 1, 2013.
- Takatoshi Kasai, MD, PhD∗,†,‡∗ (, )
- Satoshi Kasagi, MD, PhD∗,
- Ken-ichi Maeno, MD∗,
- Tomotaka Dohi, MD, PhD∗,‡,§,
- Fusae Kawana, BSc∗,†,‖,
- Mitsue Kato, MT∗,†,‖,
- Ryo Naito, MD‡,
- Sugao Ishiwata, MD, PhD§,‖,
- Minoru Ohno, MD§,
- Tetsu Yamaguchi, MD§,
- Koji Narui, MD∗ and
- Shin-ichi Momomura, MD¶
- ↵∗Reprint requests and correspondence:
Dr. Takatoshi Kasai, Sleep Center, Toranomon Hospital, 2-2-2 Toranomon, Minato-ku, Tokyo 105-8470, Japan.
Objectives The aim of this study was to investigate whether effective suppression of central sleep apnea (CSA) by adaptive servo-ventilation (ASV) improves underlying cardiac dysfunction among patients with heart failure (HF) in whom CSA was not effectively suppressed by continuous positive airway pressure (CPAP).
Background The presence of CSA in HF is associated with a poor prognosis, whereas CPAP treatment improves HF. However, in a large-scale trial, CPAP failed to improve survival, probably due to insufficient CSA suppression. Recently, ASV was reported as the most effective alternative to CSA suppression. However, the effects of sufficient CSA suppression by ASV on cardiac function are unknown.
Methods Patients with New York Heart Association class ≥II HF, left ventricular ejection fraction <50%, and CSA that was unsuppressed (defined as an apnea-hypopnea index ≥15) despite ≥3 months of CPAP were randomly assigned to receive ASV in either CPAP mode or ASV mode.
Results Of 23 patients enrolled, 12 were assigned to the ASV-mode group and 11 were assigned to the CPAP-mode group. Three months after randomization, the ASV mode was significantly more effective in suppressing the apnea-hypopnea index (from 25.0 ± 6.9 events/h to 2.0 ± 1.4 events/h; p < 0.001) compared to the CPAP mode. Compliance was signi-ficantly greater with the ASV mode than with the CPAP mode. Improvement in left ventricular ejection fraction was greater with the ASV mode (32.0 ± 7.9% to 37.8 ± 9.1%; p < 0.001) than with the CPAP mode.
Conclusions Patients with HF and unsuppressed CSA despite receiving CPAP may receive additional benefit by having CPAP replaced with ASV. Additionally, effective suppression of CSA may improve cardiac function in HF patients.
Central sleep apnea (CSA) is frequently observed (1) and is associated with increased mortality in patients with heart failure (HF) (2). Treatment of CSA has been shown to improve underlying HF; for instance, continuous positive airway pressure (CPAP) can not only suppress abnormal breathing patterns but also improve cardiac function in HF patients with CSA (3–6). The authors used data from the CANPAP (Canadian Continuous Positive Airway Pressure for Patients with Central Sleep Apnea and Heart Failure) study to analyze the long-term efficacy of CPAP in patients with HF and CSA but reported that CPAP did not improve transplantation-free survival (7).
Insufficient CSA suppression may be a possible explanation for these unexpected results in the CANPAP trial. Post-hoc analysis of data from the CANPAP trial showed that 43% of patients with CPAP still had significant CSA (i.e., apnea-hypopnea index [AHI]: ≥15 events/h) 3 months after CPAP initiation, and that greater increases in left ventricular ejection fraction (LVEF) and transplant-free survival over controls were observed only in patients showing suppressed CSA (8). Thus, AHI may remain ≥15 events/h on CPAP in a subset of HF patients with CSA, and effective AHI reduction may be a predictor of improved cardiac function and outcome. However, it remains unknown whether unsuppressed CSA by CPAP should be an interventional target.
Adaptive servo-ventilation (ASV) is reportedly the most effective alternative for CSA suppression (9). Furthermore, in HF patients with CPAP-unsuppressed CSA, ASV use was associated with significant CSA improvement (10). Several groups have reported efficacy of ASV in the treatment of underlying cardiac dysfunction in HF patients with CSA (11,12). Therefore, the objective of this study was to determine whether further suppression of AHI by ASV produces incremental benefits in measures of cardiac function among HF patients with CPAP-unsuppressed CSA.
In this prospective, single-center, randomized, single-blind trial (for more details on the methods of this study, see Online Appendix), subjects were recruited among patients who had been followed up at the Cardiovascular and Sleep Center of Toranomon Hospital (Tokyo, Japan). Inclusion criteria were as follows: 1) previous diagnosis of moderate to severe CSA, defined as AHI ≥15 events/h, of which ≥50% were central events; 2) on CPAP therapy for ≥3 months but current (within 1 month) AHI on CPAP remained ≥15 events/h with polysomnography; 3) history of persistent systolic HF (even after CPAP initiation), defined as an LVEF <50% on echocardiography and New York Heart Association class ≥II; and 4) absence of HF exacerbations while receiving stable-dose optimal medical therapy within the previous 3 months.
Exclusion criteria were: 1) age <20 or ≥80 years; 2) HF primarily due to organic valvular heart disease with anatomic alteration of the valve; 3) need for cardiac resynchronization therapy; 4) chronic obstructive pulmonary disease with ventilatory impairment; 5) need for dialysis; and 6) history of stroke with neurological deficit.
All patients provided written informed consent to participate in this study. The study protocol complied with the Declaration of Helsinki and was approved by the ethics committee at Toranomon Hospital.
All enrolled subjects were provided an ASV device (HEART PAP, Respironics, Murrysville, Pennsylvania) to use in place of their conventional CPAP device and were randomly assigned to ASV or CPAP mode. Patients underwent titration to the assigned mode. In the CPAP-mode group, pressure levels were manually modulated and an appropriate fixed pressure was identified. In the ASV-mode group, previously described principles of titration for ASV were used (11). Briefly, expiratory positive airway pressure (EPAP) was manually titrated. Minimal inspiratory positive airway pressure (IPAP) was set at the determined EPAP level or to EPAP + 2 cm H2O. Maximal IPAP was set to 10 cm H2O above minimal IPAP. Backup rate was set as an automatic or fixed rate of >10 breaths/min. If continued periodic breathing was observed, maximal IPAP was raised by 2 cm H2O. All subjects were instructed to use the ASV device at home with the blinded assigned mode.
The effectiveness of CSA treatment was evaluated by overnight polysomnography at baseline and 3 months after randomization. CPAP level before study enrollment and nightly CPAP usage for 3 months before enrollment were also determined.
Assessments at baseline and follow-up included: polysomnography; body mass index; subjective sleepiness; arterial blood gas; cardiovascular variables, such as blood pressure and heart rate; plasma B-type natriuretic peptide; 24-h urinary norepinephrine excretion (UNE); 6-min walking distance (6MWD); echocardiographic parameters, including LVEF, left ventricular end-diastolic and end-systolic diameters, and mitral regurgitation (MR) area; and quality of life (QOL), as measured using the 36-Item Short Form Health Survey (SF-36).
All values are shown as the mean ± SD or for normally distributed data and as the median (interquartile range) for non-normally distributed data. Categorical variables are expressed as numbers and percentages. Patient characteristics at baseline were compared using the Student t test for normally distributed data, the Mann-Whitney U test for non-normally distributed data for continuous variables, and the chi-square test or Fisher exact test for categorical variables. Two-way repeated-measures analysis of variance, followed by the Tukey test, was used to compare within- and between-group differences in variables measured at baseline and 3 months. For non-normally distributed data, natural logarithm–transformed values were used for analyses. A p value <0.05 was considered statistically significant.
Patient characteristics and follow-up
Seventy-four HF patients with CSA on CPAP were screened. Among them, 41 failed to meet the criteria and 10 declined participation. Of the 23 remaining patients, 12 were assigned to the ASV-mode group and 11 were assigned to the CPAP-mode group. There were no significant differences in patients’ characteristics at baseline (Table 1).
All of the patients completed the study. No changes were made to patients’ medications during the study period. None of the patients were using narcotic analgesics at baseline or during the study period. Mean change in weight at 3 months was not significantly different between the ASV and CPAP groups (–1.1 ± 4.3 kg vs. –0.9 ± 2.6 kg; p = 0.883).
Polysomnography data, device usage, sleepiness, and arterial blood gas
At baseline, there were no significant differences in polysomnography data, sleepiness, and arterial blood gas variables between the 2 groups (Table 2). Mean AHI on baseline polysomnography with the currently used CPAP level was similar between groups (p = 0.529) (Fig. 1A), and nearly all respiratory events were central in nature (percentage of central apnea or hypopnea: 97.7 ± 3.5% in the ASV-mode group vs. 97.6 ± 2.2% in the CPAP-mode group; p = 0.951).
All of the patients used the same type of the nasal mask during the study period. At 3 months, reductions in AHI and arousal index were significantly greater in the ASV-mode group than in the CPAP-mode group. The increase in the mean lowest arterial oxygen saturation measurement was significantly greater, and sleep structure was significantly better, with ASV mode compared to CPAP mode (Table 2, Fig. 1A).
In the CPAP-mode group, mean provided pressure was 7.7 ± 1.3 cm H2O. Conversely, in the ASV-mode group, mean IPAP and EPAP were 12.4 ± 2.4 cm H2O and 6.6 ± 2.0 cm H2O, respectively. Seven patients used automatic backup and 4 patients used fixed backup (rates: 10 and 12 backups/min, respectively). In the ASV-mode group, nightly device usage increased significantly from baseline to the end of the study period, from 3.5 ± 1.1 h/night to 4.7 ± 0.6 h/night (p = 0.006), whereas the change in the CPAP-mode group was nonsignificant (from 3.2 ± 0.9 h/night to 3.3 ± 1.2 h/night; p = 0.990). Change in nightly device usage was significantly greater in the ASV-mode group than in the CPAP-mode group (p = 0.027).
There were no significant differences in sleepiness or arterial oxygen pressure within or between the 2 groups. However, the increase in arterial carbon dioxide pressure was significantly greater in the ASV-mode group than in the CPAP-mode group (Table 2).
Cardiovascular variables and QOL
As shown in Table 3, the ASV-mode group showed significant reductions in plasma B-type natriuretic peptide in both within- and between-group comparisons. Although UNE and 6MWD were not significantly changed within groups, the reduction in UNE and increase in 6MWD were significantly greater in the ASV-mode group than in the CPAP-mode group. The ASV-mode group showed significant reductions in left ventricular end-systolic diameter in both within- and between-group comparisons, whereas the changes in the left ventricular end-diastolic diameter were nonsignificant both within and between the 2 groups. MR area was decreased significantly within the ASV-mode group, and the reduction was significantly greater than that in the CPAP-mode group. LVEF was significantly increased within the ASV-mode group, and the LVEF increase was significantly greater in the ASV-mode group than in the CPAP-mode group (Fig. 1B). Improvement in QOL was significantly greater in the ASV-mode group than in the CPAP-mode group (Online Fig. 1).
The authors previously reported that patients HF and CSA in whom AHI was not sufficiently suppressed with CPAP experienced further and effective reduction of AHI via bi-level positive airway pressure and, subsequently, improved cardiac function (13). However, this was not a randomized study; none of the subjects underwent follow-up polysomnography, although some patients still had mild to moderate CSA with bi-level positive airway pressure; and ASV, not bi-level positive airway pressure, was used. In addition, patients in the previous study (13) were defined as nonresponders to CPAP after 1 night of CPAP application, whereas previous studies have reported that in HF patients, CPAP alleviated CSA and improved cardiac function when applied by gradually increasing pressure levels (4–8). Thus, patients in the authors’ previous study differed from the CPAP-unsuppressed CSA patients in post-hoc analysis of the CANPAP trial and from those in the present study in which CPAP was applied for at least 3 months. Therefore, the present study first shows that further reduction of AHI resulted in incremental benefits in cardiac function among HF patients with CPAP-unsuppressed CSA, suggesting that CPAP-unsuppressed CSA in HF patients may be an interventional target in conjunction with treatments that effectively suppress AHI.
Several possible mechanisms may explain how further CSA suppression by ASV exerts efficacy on cardiac function. One such mechanism includes the suppression of sympathetic nerve activity (SNA). In the present study, a significantly greater reduction in UNE was observed in the ASV-mode group than in the CPAP-mode group. This reduction may have been associated not only with a significant reduction of AHI but also with more favorable changes in sleep quality and further improvement of oxygenation in the ASV-mode group. HF patients with CSA were reported to have had greater SNA than those without CSA, and suppressing CSA may have reduced such overactivity (5). Augmented SNA causes increases in peripheral vascular resistance, blood pressure, and cardiac afterload. Thus, increased SNA can exacerbate cardiac dysfunction via increased cardiac afterload and have direct and negative effects on myocytes. Reduction of SNA by further suppressing CSA through the use of ASV may improve cardiac contractility, stroke volume by reducing peripheral vascular resistance, and circulatory delay. This hypothesis is supported by findings from a recent study reporting that chronic ASV treatment increased LVEF and stroke volume in association with decreased systemic vascular resistance and severity of MR in HF patients (14). Indeed, in the present study, MR area was significantly reduced in the ASV-mode group. Independent of CSA suppression, positive airway pressure has a direct hemodynamic effect on reduction in cardiac preload and afterload in patients with HF by increasing the intrathoracic pressure (15,16). Although provided pressure levels during exhalation were lower in the ASV-mode group than in the CPAP-mode group, the net applied pressure during the entire respiratory cycle and during the entire night may be higher with the ASV mode. However, the actual net applied pressure was not determined in the ASV-mode group because the device used in the present study did not indicate such pressure. Improvements in LVEF in the ASV-mode group were corroborated by a decrease in left ventricular end-systolic diameter, although left ventricular end-diastolic diameter was not significantly changed. Thus, effective suppression of CSA with ASV may retard or reverse ventricular remodeling through these mechanisms. These findings are consistent with those from recent studies reporting that long-term ASV treatment was associated with ventricular reverse remodeling (11,14). In the ASV-mode group, improvement in LVEF, unloading of the left ventricle, and subsequent improvement in pulmonary congestion may have affected reductions in plasma B-type natriuretic peptide and slightly increased the 6MWD.
HF patients with CSA have low carbon dioxide pressure (pco2) associated with hyperventilation due to stimulation of pulmonary vagal irritant receptors by pulmonary congestion and increased carbon dioxide chemosensitivity (17). ASV can keep pco2 levels consistent and improve the chronic hyperventilation state, thereby increasing daytime pco2 levels by improving pulmonary congestion (18). Indeed, in the present study, a significant difference in arterial carbon dioxide pressure change was observed between the 2 groups. In contrast, improvement of pco2 fluctuation by ASV may also contribute to spontaneous suppression of abnormal breathing patterns and improvements in cardiac function by improving respiratory control instability, circulatory delay, and lung volume.
Better treatment compliance in the ASV-mode group may have contributed to improvements in cardiac function. An earlier small-scale study reported significant improvement in transplantation-free survival only in a subgroup of patients who were compliant with CPAP (2). Furthermore, sleep quality was better in the ASV-mode group than in the CPAP-mode group, perhaps due to better treatment compliance in the ASV-mode group.
The number of patients enrolled was small; moreover, they were all Japanese men. Therefore, the effects of ASV on improvements in cardiac function and neurohormonal status in women with HF and CSA unsuppressed by CPAP and in non-Japanese patients remain unclear. Also, because patients undergoing cardiac resynchronization therapy were excluded, the results of this study may not be applicable to such a patient population. Furthermore, although subjects were blinded to their treatment modes, they may have been aware of the assigned treatment mode because of the perception of pressure support or backup ventilation while awakening after sleep onset. Assumptions about treatment assignment may have affected treatment compliance and/or QOL responses. Additionally, blinding was not possible for investigators, although sonographers were blinded to treatment assignments. Finally, all patients used the CPAP device with fair compliance for at least 3 months before inclusion into the study and may have been familiar with using the mask-ventilation system. Although this experience may have minimized bias in preference of the mask-ventilation system, the findings cannot be applied to HF patients with predominant CSA who have never used a mask-ventilation system.
The results suggest that ASV treatment of CSA improves underlying cardiac dysfunction in male Japanese HF patients with CPAP-unsuppressed CSA. Moreover, effective suppression of CSA may be beneficial in improving cardiac function in HF patients. However, the long-term effects of CSA treatment by ASV are unknown. Therefore, randomized trials assessing long-term outcomes including mortality are warranted.
For a supplementary methods section, please see the online version of this article.
This research was funded by grants from the Okinaka Memorial Institute for Medical Research, Tokyo, Japan. Dr. Kasai has received an unrestricted research fellowship from Fuji-Respironics Inc. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- distance walked in 6 min
- apnea-hypopnea index
- adaptive servo-ventilation
- continuous positive airway pressure
- central sleep apnea
- expiratory positive airway pressure
- heart failure
- inspiratory positive airway pressure
- left ventricular ejection fraction
- mitral regurgitation
- carbon dioxide pressure
- quality of life
- sympathetic nerve activity
- 24-h urinary norepinephrine excretion
- Received July 3, 2012.
- Revision received October 30, 2012.
- Accepted November 15, 2012.
- American College of Cardiology Foundation
- Sin D.D.,
- Logan A.G.,
- Fitzgerald F.S.,
- Liu P.P.,
- Bradley T.D.
- Tkacova R.,
- Liu P.P.,
- Naughton M.T.,
- Bradley T.D.
- Arzt M.,
- Floras J.S.,
- Logan A.G.,
- et al.
- Kasai T.,
- Usui Y.,
- Yoshioka T.,
- et al.
- Naughton M.T.,
- Rahman M.A.,
- Hara K.,
- Floras J.S.,
- Bradley T.D.