Received: 10 November 2010 / Revised: 27 December 2010 / Accepted: 13 January 2011 / Published online: 25 February 2011 # Springer-Verlag 2011

Transcription

1 Sleep Breath (2012) 16:71 78 DOI /s ORIGINAL ARTICLE Prevalence and treatment of central sleep apnoea emerging after initiation of continuous positive airway pressure in patients with obstructive sleep apnoea without evidence of heart failure Michael Westhoff & Michael Arzt & Patric Litterst Received: 10 November 2010 / Revised: 27 December 2010 / Accepted: 13 January 2011 / Published online: 25 February 2011 # Springer-Verlag 2011 Abstract Background This study aimed to assess the prevalence of complex sleep apnoea (CompSA), defined as central sleep apnoea (CSA) emerging after the initiation of continuous positive airway pressure (CPAP) therapy for obstructive sleep apnoea (OSA), in patients with normal brain natriuretic peptide (BNP) levels, along with assessing the prevalence of CSA persisting in such patients after the onset of CPAP therapy. We hypothesised that the prevalence of CompSA and persistent CSA after CPAP initiation would be low in patients with OSA and normal BNP levels. Material and methods Between April 2004 and July 2007, CPAP was initiated for all patients with OSA for two nights using a standardised protocol. The prevalence of CompSA syndrome (CompSAS) and persisting CSA [central apnoea index (CAI) >5/h and apnoea hypopnoea index (AHI) > 15/h with >50% central events during CPAP therapy] was prospectively assessed in patients with normal BNP levels. Patients with CompSAS or persisting CSA upon CPAP treatment received adaptive servoventilation (ASV). M. Westhoff : P. Litterst Department of Pulmonary, Critical Care and Sleep Medicine, Hemer Lung Clinic, Theo-Funccius-Str. 1, Hemer, Germany M. Arzt Center for Sleep Medicine, Department of Internal Medicine II, Division of Respirology, University of Regensburg, Regensburg, Germany M. Westhoff (*) Lungenklinik Hemer, Theo-Funccius-Str. 1, Hemer, Germany woelkewesthoff@aol.com Results Of 1,776 patients with OSA receiving CPAP, 28 patients (1.57%) had CSA at the time of CPAP therapy and normal BNP levels. Additionally, 10 patients had CompSAS (0.56%) and 18 patients (1.01%) had persisting CSA. In patients with CompSA or persisting CSA, the AHI was significantly lower with CPAP therapy than at the time of diagnosis (34±15/h vs. 47±20/h, p=0.005). The CAI increased from 10±10/h to 18/h±13/h (p=0.009) upon initiation of CPAP therapy. ASV reduced the AHI to 6±12/h (p<0.001) during the first night of use. Conclusion The prevalence of CompSA or persisting CSA in patients with OSA and normal BNP levels who are receiving CPAP therapy is low (1.57%). ASV is an effective treatment for these patients. Keywords Adaptive servoventilation. Central sleep apnoea. Complex sleep apnoea. Brain natriuretic peptide Introduction It is a well-known phenomenon in sleep medicine that patients with obstructive sleep apnoea (OSA) may show central sleep apnoea (CSA) under continuous positive airway pressure (CPAP) therapy. Thomas et al. [1] described this condition as complex sleep-disordered breathing, and Morgenthaler et al. [2] used the term complex sleep apnoea. This new malady received considerable attention but was not accepted in the international classification of sleep disorders until recently. Thomas et al. [1] reported results for patients who were young, had a normal body mass index (BMI), did not have further comorbidities, and were primarily kept in the supine position for CPAP titration. The patients in the study published by Morgenthaler et al. [2] had cardiac comorbid-

2 72 Sleep Breath (2012) 16:71 78 ities and elevated BMIs. Furthermore, Morgenthaler et al. [2] did not exclude patients with diastolic dysfunction and preserved systolic LV function, and brain natriuretic peptide (BNP) was not measured. Therefore, heart failure must be considered a confounding factor in that study. The authors also applied split-night procedures for the diagnosis of obstructive sleep apnoea and therapy to treat the disorder. Patients with OSA may show central apnoeas at baseline, some of which may persist under CPAP therapy. However, the detection of such apnoeas can be missed using splitnight procedures because of a short diagnostic recording time. In general, there are few data about the prevalence of central sleep apnoea which emerges upon CPAP therapy for OSA (termed complex sleep apnoea syndrome or CompSA) or CSA which persists under those conditions. Therefore, in 2004, we began an observational study to determine the prevalence of these conditions. CSA is seen more often in heart failure patients, and fast CPAP titration may lead to higher pressures and induce CSA. Therefore, the present study of the prevalence of CompSAS and persistent CSA upon initiation of CPAP is based on a stepwise and standardised protocol for CPAP titration over two nights and excluded patients with evidence of heart failure. We hypothesised that, using this diagnostic approach, the prevalence of CompSAS and persistent CSA upon CPAP would be low. Hyperventilation plays a major role in the pathophysiology and generation of central apnoeas, not only in heart failure patients with Cheyne Stokes respiration and central sleep apnoea (CSR-CSA), but also in patients CompSAS and persistent CSA [3, 4]. Based on these observations and studies showing that adaptive servoventilation (ASV) normalises respiratory patterns and apnoea hypopnoea index (AHI) in cases of CSR-CSA [5 9], ASV was chosen for therapy in patients with CompSA and persistent CSA. device (Somnosmart II, Weinmann Inc., Germany) was used. This apparatus is based on oscillatory pressure signal, flow and snoring signals. Manual titration with polysomnographic evaluation was carried out the following night (i.e. the second night on CPAP). The starting pressure was chosen as the mean pressure level of the APAP device. The pressure was increased stepwise by 1 cmh 2 O with the occurrence or persistence of obstructive and/or central events. If central apnoeas persisted up to a pressure of 12 cmh 2 O, the pressure was reduced to the level which was effective to eliminate obstructive apnoeas. Patients with central apnoea indices (CAIs) >5/h and AHIs 15/h under effective CPAP therapy for OSA were tested for BNP levels (Biosite ). All patients with elevated BNP (>100 pg/ml) were excluded. The remaining patients, defined as having persisting CSA (CAI >5/h at baseline) or CompSAS (CAI <5/h at baseline and >5/h under CPAP), were treated with ASV (Auto-CS-2, Resmed Corp., Australia or Auto-SV, Respironics, Murraysville, PA) (Fig. 1). The expiratory positive airway pressure (EPAP) was set at the pressure that had proved to be effective in eliminating obstructive apnoeas in REM sleep in the supine position during the CPAP study. Inspiratory pressure support for the Auto-CS-2 was 3 10 cmh 2 O with a backup rate of 15 breaths/min. The IPAP range in the Auto-SV apparatus was set from 0 to 10 cmh 2 O over EPAP, and for the backup Suspicion of SAS OSA or CSA OSA (AHI > 15/h) PSG Methods Study design From April 1, 2004 to July 1, 2007, all patients of the Sleep Laboratory of the Hemer Lung Clinic who underwent polysomnography (PSG) were tested for inclusion and exclusion criteria. Inclusion criteria for this observational study consisted of having CPAP therapy scheduled for the treatment of OSA with an AHI 15/h. Patients with CSA and opioid therapy or CSR-CSA and elevated BNP were excluded. For all included patients, CPAP was initiated using a standardised protocol. On the first night of CPAP titration, an automatic positive airway pressure (APAP) Fig. 1 Study progression CAI > 5/h and AHI > 15/h and BNP < 100 pg/ml Adaptive Servoventilation 2 x CPAP + PSG

3 Sleep Breath (2012) 16: rate, the auto mode was used. The patients underwent a full-night polysomnographic study while on ASV. A follow-up with full-night polysomnography was performed after 6 weeks of therapy with ASV. The study was approved by the institutional review board of the Hemer Lung Clinic. Patients gave their informed consent. Polysomnography measurement Attended polysomnography was performed with the Alice 3.5 system (Respironics). The following parameters were recorded: EEG C4A1 and C3A2; submental and pretibial electromyography; electro-oculography; ECG; respiratory effort (thoracic and abdominal induction plethysmography); respiratory flow, both with a flow-pressure monitor (Heinen und Löwenstein, Bad Ems, Germany) and with an oral/ nasal thermocouple (Respironics); snoring signals (Alice laryngeal microphone and Backelektret microphone type ME 2/V 413, Peiker Inc.) and oxygen saturation by finger pulse oximetry (Novametrix Inc.). Airflow during polysomnography with CPAP and ASV was monitored only with the flow-pressure monitor. The polysomnographic recordings were scored manually (by M.W. and P.L). Sleep stages and arousals were analysed according to the criteria of Rechtschaffen and Kales [10] and the American Sleep Disorders Association [11]. A central apnoea was defined as an absence of airflow for 10 s associated with an absence of chest and abdominal movement; obstructive apnoea was defined as an absence of respiratory flow for 10 s despite ongoing respiratory effort. Hypopnoea was defined as a reduction in airflow of at least 50% compared with baseline for 10 s or a <50% amplitude reduction associated with either an oxygen desaturation of >3% or an arousal. Mixed apnoea was defined as an apnoea 10 s showing central as well as obstructive components. Snoring detected by online audio video registration was chosen as an additional criterion for underlying obstruction in mixed apnoeas and hypopnoea. The AHI was defined as the total number of apnoeas and hypopnoea divided by the total sleep time and is expressed as the number of events per hour. Polysomnography was performed in a sleep lab that fulfilled the quality criteria of, and was accredited by, the German Sleep Society. Statistical analysis All data were analysed using a statistical software package (SPSS, version 16.0; SPSS, Inc., Chicago, IL, USA). Data are shown as mean ± SD. To assess differences in the number of respiratory events, oxygen saturation and polysomnographic measures of sleep quality between the sleep study at baseline, during CPAP titration and during treatment with ASV, repeated-measures analysis of variance (ANOVA) was carried out with post hoc contrast using t tests. A two-sided p value of less than 0.05 was considered to indicate statistical significance. Results Patients From April 1, 2004 to July 1, 2007, a total of 1,776 patients with obstructive sleep apnoea (AHI >15/h) were treated with CPAP. CPAP therapy was applied using a standardised protocol. Prevalence of CompSA and persistent CSA Fifty patients had central apnoeas after CPAP therapy for obstructive sleep apnoea. Twenty two showed typical CSA-CSR with elevated BNP and were therefore excluded. Twenty-eight patients (1.57%), 1 female and 27 male, exhibited CSA under CPAP therapy. Eighteen patients (1.01%) had persistent CSA, and ten patients exhibited central apnoeas under CPAP therapy. Thus, the prevalence of complex sleep apnoea was 0.56% (Fig. 2). There were no differences in age, BMI, Epworth Sleepiness Scale (ESS), or pco 2 between the two groups (Table 1). Study population Of the patients in the study, 21 (72%) had hypertension, 9 (31%) had coronary artery disease, and 5 (17%) had atrial fibrillation. In terms of pharmaceutical use, 16 patients (55%) were on ß-blockers, 14 (48%) were on ACE inhibitors/at1 receptor antagonists, 11 (38%) were on diuretics, and 7 (24%) were on calcium-channel blockers. Total AHI was elevated to 47±20/h, the CAI was 10± 10/h, and the arousal index was 35±16/h. With the exception of CAI (2/h vs. 17/h), the total AHI, sleep stages, and mean and minimum SaO 2 at baseline polysomnography were not significantly different between the patients with persistent CSA and CompSA (Table 2). Treatment of CompSA and persistent CSA Obstructive apnoeas were effectively suppressed by CPAP of 5.8±1.2 cmh 2 O in all patients with CompSA or persistent CSA. This treatment resulted in a significant reduction in the total AHI by CPAP therapy but not in normalisation of this parameter (47±20/h to 34±15/h, p=0.005). Higher pressure did not lead to further reduction

4 74 Sleep Breath (2012) 16:71 78 persistent CSA n=18(1,01%) CPAP-therapy in OSA n=1776(100%) persistent CSA and Complex sleep apnea with BNP <100 pg/ml n=28(1,57%) Complex sleep apnea n=10(0,56%) Fig. 2 Prevalence of persistent CSA and CompSA (April 1, 2004 to July 1, 2007) of the AHI. The CAI increased from 10±10/h to 18/h±12/h (p=0.009) in the whole study population; in patients with persistent central apnoeas, the CAI increased from 15±9/h to 18±14/h (p=0.26), and in CompSA patients, it increased from 2±2/h to 17±8/h. The CAI under CPAP treatment was not significantly different between the two groups (p=0.75). ASV therapy was initiated in 28 patients with CompSA or persistent CSA. Twenty-two patients used the Auto- CS2 system (Resmed; Australia), and six used the BiPAP Auto-SV system (Respironics). ASV led to a significant reduction in the AHI, from 47±20/h to 7±11/h (p<0.001), and a significant decrease in the arousal index, from 35± 16/h to 17/h±8/h (p<0.001) after the first night of ASV therapy. Only a single patient did not show a normalised AHI under ASV treatment. His AHI was 28/h at baseline, increased to 56/h under CPAP, and reached 58/h under ASV. He refused further therapy. After 6 weeks, 27 patients who adhered to the ASV therapy underwent PSG. The ESS scores of these patients were significantly reduced, from 9.0±4.7 to 6.4±3.7 (p=0.014), compared to baseline. The AHI, CAI, and arousal index remained significantly improved. However, sleep stages did not change significantly compared to baseline, with only REM showing a trend in improvement (Table 3). There were no differences in ESS score or sleep data between patients with persistent CSA and CompSA (Table 4). Table 2 Apnoea indices, arousal index, and sleep stages in all patients and patients with central sleep apnoea persisting (group A) or emerging (group B) under CPAP therapy Variable Discussion All patients (n=28) Persistent CSA (n=18) CompSA (n=10) p value AHI (n/h) 47±20 51±18 40± CAI (n/h) 10±10 15±9 2± Arousals (n/h) 35±18 34±12 36± NREM 1/2 71±14 68±14 75± NREM 3/4 13±13 15±14 11± REM 14±8 15±9 12± SaO 2 mean 94.2± ± ± SaO 2 minimum 82.6± ± ± Baseline sleep data The present data show that the incidence of central apnoeas that persist or emerge under CPAP therapy for obstructive sleep apnoea (i.e. CompSA) is low in patients with normal BNP, when using a stepwise diagnostic and therapeutic approach with at least two full-night polysomnographic studies under CPAP treatment. BNP is a sensitive marker for systolic and diastolic heart failure [12 14]. Levels of BNP are elevated in cases of CSA-CSR and are associated with the severity of cardiac and sleep-related disease [14, 15]. In contrast to the inclusion criteria of prior studies about central or complex sleep apnoea [2, 16 18], this is the first study measuring BNP as an additional marker for the exclusion of possible confounding CSA-CSR cases. This exclusion is especially important for patients with a history of coronary artery disease, hypertension, and/or atrial fibrillation, which are predisposing factors for heart failure. The high prevalence of coronary artery disease and hypertension in our study population is comparable to the data reported by Lehman et al. [17] for patients with Table 3 ESS score, apnoea indices, arousal index, and sleep stages before and after 6 weeks of ASV in 27 patients adhering to ASV treatment Variable Baseline (n=27) ASV, 6 weeks (n=27) p value Table 1 Characteristics of the study population Variable All Patients (n=28) Persistent CSA (n=18) CompSA (n=10) p value Age, years 64.6± ± ± BMI, kg/m ± ± ± ESS score 8.8± ± ± pco 2, mmhg 36.7± ± ± ESS 9.0± ± AHI 48±20 4±4 < CAI 11±10 1±3 < Arousal index 35±16 18±9 < NREM 1/2 71±14 66± NREM 3/4 13±13 15± REM 14±8 18± Sleep data before and after 6 weeks of ASV

5 Sleep Breath (2012) 16: Table 4 ESS score, apnoea indices, arousal index, and sleep stages in 27 patients adhering to therapy after 6 weeks of ASV, analysed by subgroup (persistent CSA and CompSA) Variable Persistent CSA (n=18) CompSA (n=9) p value ESS 6.6± ± AHI (n/h) 4±3 4± CAI (n/h) 2±4 1± Arousals (n/h) 17±10 20± NREM 1/2 68±18 62± NREM 3/4 16±16 15± REM 16±8 22± ESS and sleep data for patient subgroups after 6 weeks of ASV CompSA. The prevalence of atrial fibrillation is lower than that reported by Leung et al. [19], who saw an incidence of 27%. This difference can be attributed to the exclusion of patients with elevated BNP in our study. Nevertheless, a significant number of patients in our study had cardiovascular disease. Medical treatment of arterial hypertension and coronary artery disease can explain normal values of BNP in this group of patients. Mansfield et al. [20] reported a study on a group of patients who had CSA-CSR before heart transplantation and CSA afterwards. The patients with CompSA or persistent CSA might be those who are at risk of developing typical CSR-CSA with progression to heart failure and elevated BNP. In the present study, all patients had obstructive events at baseline, which is consistent with observations by Thomas et al. [1] and Gilmartin et al. [21]. These authors described an obstructive sleep-disordered breathing with a dominant cycling alternating pattern and partial treatment failure in patients with lower body mass, terming this complex sleep-disordered breathing. However, the high prevalence of supine dependency of complex sleep-disordered breathing reported by Thomas et al. [1] was intended, as patients were examined predominantly in the supine position. These results suggest an ongoing obstructive component. The association between central and obstructive events, especially in complex sleep-disordered breathing, can be explained by a considerable overlap in the pathogenesis and pathophysiology of central and obstructive apnoeas [22 24]. The behaviour of the upper airway in the case of obstruction may not be uniform [25]. Airway obstruction, especially in the supine position, can mimic central apnoea, as shown by Issa and Sullivan [22]. Recently, Luo et al. [26] confirmed these observations and reported that nearly 30% of obstructive events appear phenomenologically as central apnoeas. On the other hand, central hypocapnic hypopnoea may exhibit obstructive features [24]. Therefore, all patients in our sleep lab underwent CPAP titration over two nights, with different pressure levels between 4 and 12 mbar to eliminate obstructive events. The APAP device chosen for the first night of CPAP titration used an additional forced oscillation technique for the evaluation of the patency of the upper airways. This procedure revealed that obstructive apnoeas were absent by applying comparatively low mean pressures in the group of patients with emergent or persistent CSA. Higher CPAP/ EPAP levels than those chosen for the EPAP during ASV therapy were not effective at eliminating or preventing the occurrence of phenomenologically central events, which were preserved until NREM sleep. Additional video and microphone recordings gave no further clues about airway obstruction, especially in the supine position. Nevertheless, the EPAP levels in our study were lower compared to the CPAP levels reported in the study by Javaheri et al. [27], the results of which are otherwise comparable to ours. This discrepancy may be attributed to the significantly lower BMIs of the patients in our study. In addition, Javaheri et al. [27] intended to eliminate obstructive and central events completely, whereas the EPAP level for ASV in our study was chosen for the effective elimination of obstructive apnoeas. Higher CPAP levels had not been beneficial in our study to reduce or eliminate central apnoeas. In agreement with our work, Javaheri et al. [38] recently showed that the CAI is not influenced by the CPAP level over a range of 5 17 cmh 2 O. One third of our patients presented CompSA, whereas most patients already had central apnoeas at baseline polysomnography. Similar observations were reported by Lehman et al. [17] and Javaheri et al. [27]. They observed that 46% and 32%, respectively, of their patients who developed central apnoeas under CPAP therapy for obstructive sleep apnoea syndrome also had central apnoeas at baseline. Therefore, the difference in the prevalence of central apnoeas at baseline compared to Morgenthaler et al. [2] and Thomas et al. [1] must be attributed to different diagnostic procedures, i.e. full-night baseline polysomnography vs. split-night procedures or polysomnography strictly in the supine position. Javaheri et al. [27] also pointed out that many central apnoeas at baseline might have been missed in split-night procedures. The duration of baseline polysomnography and the time spent in the supine body position can influence the number of central apnoeas detected. In contrast to previously published studies, we used a full-night diagnostic polysomnography, at least two polysomnographies with CPAP and two further polysomnographies with ASV at baseline and after 6 weeks of follow-up. However, Morgenthaler et al. [2] only applied a single split-night procedure for diagnosis and therapy, resulting in the use of higher CPAP levels than in our study. Such usage was associated with an incidence of CompSAS of 15%. Kuzniar et al. [18] noticed a reduction of CompSAS under long-term CPAP therapy in nearly 50% of patients,

6 76 Sleep Breath (2012) 16:71 78 which is a good indication that the high incidence of CompSAS reported by Morgenthaler et al. [18] is a procedure-related (i.e. a split-night polysomnographyrelated) phenomenon. The data of Javaheri et al. [27] are similar to our results and stand in contrast to previous studies assessing the prevalence of CompSAS [2, 18, 32]. Javaheri et al. [27] also used full-night PSG, but only one night with CPAP titration, and performed a follow-up study. After 6 weeks of CPAP therapy, the AHI under CPAP (30±32/h) was nearly the same as in the present study (34±15/h ), although 42 patients in the study by Javaheri et al. [27] were lost in follow-up, so that the estimation of the prevalence was only based on the remaining 42 patients. The prevalence of central apneas (1.5%) reported by Javaheri et al. [27] after 6 weeks of CPAP therapy is lower than reported ever before and the same as in the present study (1.57%). We did not schedule a follow-up study after 6 weeks, but optimised the pressure over two nights of CPAP therapy. Nevertheless, this procedure provides a comparable result to the approach with CPAP titration and a follow-up after 6 weeks and it does not overestimate the prevalence of CPAP-resistant central apneas. Even the distribution of the subgroups, CPAP-emergent (1.01% vs. 0.8%) and CPAP-persistent (0.56% vs. 0.6%) central apneas are nearly identical in the present study and the study of Javaheri et al. [27]. In contrast to our data, Javaheri et al. [27] calculated the prevalence for the whole study population, assuming that the number of CPAP-resistant CSA would have been the same in the 42 patients lost in follow-up as in the 42 patients seen after 6 weeks of CPAP therapy. However, this might result in an under as well as overestimation of CompSA, because the patients who were lost in follow-up might have had a higher or lower prevalence of CompSA. We found no difference in pco 2 levels between patients with central apnoeas persisting or emerging under CPAP therapy for OSA. The mean pco 2 of 36.7 mmhg in our patients corresponds to the pco 2 levels that Dernaika et al. [28] reported for their group of patients with CPAP-induced CSA. The BMIs in our study are comparable to the low BMIs reported by Thomas et al. [1] as characteristic of their patients with complex sleep-disordered breathing and to the BMIs of the subgroup of CPAP non-responders in the follow-up study of Kuzniar et al. [18]. Furthermore, there was no difference in BMI between both subgroups in our study. These data, along with the effectiveness of ASV in treating persistent or emerging central apnoeas occurring during CPAP therapy, support the assumption that there is no difference in the pathophysiology of the generation of central apnoeas in patients with persistent or emerging central apnoeas under CPAP therapy for obstructive sleep apnoea. It is possible that repeated PSG in patients with CompSA may show relevant central apnoeas at baseline as well. Differing data concerning the prevalence of CompSA raise questions concerning the comparability of study populations. The BMIs in our study are different from the BMIs of patients with complex sleep apnoea reported by Morgenthaler et al. [2, 29], Pusalavidyasagar et al. [30], Allam et al. [31], Kuzniar et al. [18], and Javaheri et al. [27]. In those studies, the BMI ranged between 31 and 34.4 kg/m 2, as is typical for a population with classic OSA but not for patients with CSA. The BMIs of our patients with central apnoeas during CPAP and their pco 2 values are comparable to the average BMI (31.3±5.0 kg/m 2 ) and pco 2 level (35.6±5.0 mmhg) of patients with mild systolic and diastolic heart failure and CSR-CSA [9]. However, the BMIs we measured are significantly different from the mean BMI (34.2±7.1 kg/m 2 ) of patients with OSA, as seen in our sleep lab (unpublished data) and reported by other authors [32, 33]. These anthropometric differences may suggest that patients with persistent CSA and CompSA in our study are different from most of the obstructive sleep apnoea patients who suffer from what has been described as complex sleep apnoea by Morgenthaler et al. [2, 29]. As the studies of the Morgenthaler group [29 31] did not exclude heart failure (in fact, a relevant number of patients with clinical signs of heart failure and patients taking opioid medication were included), the possibility of heart failure must be regarded as a further confounding factor affecting the incidence of central apnoeas. As we excluded heart failure patients by choosing elevated BNP as an exclusion criterion, this different approach, as well as the CPAP titration protocol, may explain the low number of central apnoeas occurring under CPAP therapy in our study. Otherwise, the higher prevalence data for CSA in other studies must be attributed to concomitant heart failure, which may trigger susceptibility to central apnoeas. Furthermore, the occurrence of CSR- CSA is not dependent on reduced systolic LV function. CSR-CSA can also be seen in patients with diastolic heart failure or only mildly reduced systolic LV function [9, 34]. Therefore, an ejection fraction (EF) cut-off of 40% is not sensitive enough to select patients with heart failure and CSR-CSA. These results further argue for choosing BNP levels as a criterion for detection or exclusion of heart failure in patients with CSA and CompSA. Regarding therapeutic approach, our data show that persistent CSA and CompSA can be successfully treated with ASV, with a low rate of non-responders. However, normalisation of the AHI only indicates a trend in improving sleepiness and sleep architecture, as was also seen in the study by Javaheri et al. [27]. Morgenthaler et al. [29] used ASV in a small group of patients with CSR-CSA, mixed CSA, and CompSAS in an acute setting and found normalisation of the AHI. Banno et

7 Sleep Breath (2012) 16: al. [16] saw normalisation of the AHI in a small feasibility study with three patients. The AHI also remained normal in a follow-up of 6 12 months. However, as described above, differences in study design and in CPAP titration influence the incidence of CompSA and persistent CSA, the number of patients treated with ASV and the effectiveness of ASV. ASV reduces ventilation overall and acts by the combination of CPAP with an inverse inspiratory pressure support. The normalisation of respiration and elimination of central apnoeas by ASV in contrast to CPAP therapy supports observations that hyperventilation plays an important role in the pathophysiology of both CSR-CSA and CSA [3, 4, 20, 35]. Hall et al. [36] described only gradual differences between these two conditions. One limitation of this study concerns the definition of central apnoeas by only polysomnographic criteria. Oesophageal pressure recording might have given more detailed information about the real type of apnoea, especially pseudo-central apnoeas. There was no follow-up study of our patients with central apnoeas under CPAP. This may have led to a further reduction in the AHI and the prevalence of central apnoeas, as has been shown in patients with CSA-CSR [37] and CompSA [27]. Otherwise, the prevalence of CSA after two nights of CPAP therapy is similar to the prevalence reported by Javaheri et al. [27] after a follow-up period of 6 weeks, indicating that a treatment lasting 6 weeks would not have led to a further reduction in the AHI. Elevation of BNP was regarded as a criterion for heart failure. BNP can be influenced by other diseases (renal failure, liver cirrhosis), medication, or obesity. The patients which had been excluded in this study because of an elevation of BNP all had cardiovascular comorbidity. However, it has to be discussed if some patients, despite having heart failure and an elevation of BNP, may exhibit a CSA pattern that is more CompSA-like than representing CSA-CSR. As the study did not intend to compare different ventilation therapies for CSA, bilevel therapy was not chosen. However, Johnson and Johnson [38] recently showed that bilevel therapy cannot be regarded as the treatment of choice in central apnoeas. The data from Morgenthaler et al. [29] and Allam et al. [31] also emphasise the superiority of ASV over non-invasive positive pressure ventilation. In summary, the prevalence of central apnoeas that do not respond to CPAP therapy or arise under CPAP therapy for OSA is low in patients with normal BNP, if patients using opioids are excluded and a CPAP titration protocol lasting at least two nights is used. Using APAP therapy based on oscillatory pressure signal, flow and snoring signals for titration for one night has no negative influence on the generation of central apnoeas. There are strong arguments that CSA persisting under CPAP and CompSA are not pathophysiologically different. ASV is an effective treatment for this entity of sleepdisordered breathing. The results are comparable to those found for CSR-CSA. The influence of the supine body position and of cardiovascular diseases, especially atrial fibrillation, despite a lack of heart failure, on the generation of central apnoeas merits further attention and evaluation. Conflict of interest Dr. P. Litterst reports that no significant conflicts of interest exist with any companies/organisations whose products or services are discussed in this article. There was no financial support for the study itself from any company. Michael Westhoff and Michael Arzt have received lecture fees from Respironics and Resmed, Germany; Michael Arzt has received unrestricted grant support from Resmed, Germany and grant support from the German Foundation for Cardiac Research (Deutsche Stiftung für Herzforschung). References 1. Thomas RJ, Terzano MG, Parrino L, Weiss JW (2004) Obstructive sleep-disordered breathing with a dominant cyclic alternating pattern a recognizable polysomnographic variant with practical clinical implications. Sleep 27: Morgenthaler TI, Kagramanov V, Hanak V, Decker PA (2006) Complex sleep apnea syndrome: is it a unique clinical syndrome? Sleep 29: Xie A, Rutherford R, Rankin F, Wong B, Bradley TD (1995) Hypocapnia and increased ventilator responsiveness in patients with idiopathic central sleep apnea. Am J Respir Crit Care Med 152: Solin P (2000) Peripheral and central ventilatory responses in central sleep apnea with and without congestive heart failure. Am J Respir Crit Care Med 162: Philippe C, Stoica-Herman M, Drouot X, Raffestin B, Escourrou P, Hittinger L, Michel PL, Rouault S, d Ortho MP (2006) Compliance with and efficacy of adaptive servo-ventilation (ASV) versus continuous positive airway pressure (CPAP) in the treatment of Cheyne Stokes respiration in heart failure over a six month period. Heart 92: Teschler H, Döhring J, Wang YM, Berthon-Jones M (2001) Adaptive pressure support servo-ventilation. Am J Respir Crit Care Med 164: Pepperell JC, Maskell NA, Jones DR, Langford-Wiley BA, Crosthwaite N, Stradling JR, Davies RJ (2003) A randomized controlled trial of adaptive ventilation for Cheyne Stokes breathing in heart failure. Am J Respir Crit Care Med 168: Töpfer V, El-Sebai M, Wessendorf TE, Moaridis I, Teschler H (2004) Adaptive servoventilation bei chronischer Herzinsuffizienz: Wirkung auf Cheyne-Stokes-Atmung und Lebensqualität. Pneumologie 58: Westhoff M, Arzt M, Litterst P (2010) Influence of adaptive servoventilation on B-type natriuretic peptide in patients with Cheyne Stokes respiration and mild to moderate systolic and diastolic heart failure. Pneumologie 64: Rechtschaffen A, Kales A (1968) A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Bethesda, Maryland, U.S. Department of Health, Education, and Welfare Public Health service, National Institutes of Health

8 78 Sleep Breath (2012) 16: AASM (1999) 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 22: McDonagh TA, Robb SD, Murdoch DR, Morton JJ, Ford I, Morrison CE, Tunstall-Pedoe H, McMurray JJ, Dargie HJ (1998) Biochemical detection of left-ventricular systolic dysfunction. Lancet 351: Lubien E, DeMaria A, Krishnaswamy P, Clopton P, Koon J, Kazanegra R, Gardetto N, Wanner E, Maisel AS (2002) Utility of B-natriuretic peptide in detecting diastolic dysfunction: comparison with Doppler velocity recording. Circulation 105: Steg PG, Joubin L, McCord J, Abraham WT, Hollander JE, Omland T, Mentré F, McCullough PA, Maisel AS (2005) B-type natriuretic peptide and echocardiographic determination of ejection fraction in the diagnosis of congestive heart failure in patients with acute dyspnea. Chest 128: Christ M, Sharkova Y, Fenske H, Rostig S, Herzum I, Becker HF, Mueller C, Grimm W (2007) Brain natriuretic peptide for prediction of Cheyne Stokes respiration in heart failure patients. Int J Cardiol 116: Banno K, Okamura K, Kryger MH (2006) Adaptive servoventilation in patients with idiopathic Cheyne Stokes breathing. J Clin Sleep Med 2: Lehman S, Antic NA, Thompson C, Catcheside PG, Mercer J, McEvoy RD (2007) Central sleep apnea on commencement of continuous positive airway pressure in patients with a primary diagnosis of obstructive sleep apnea-hypopnea. J Clin Sleep Med 3: Kuzniar TJ, Pusalavidyasagar S, Gay PC, Morgenthaler TI (2008) Natural course of complex sleep apnea a retrospective study. Sleep Breath 12: Leung RS, Huber MA, Rogge T, Maimon N, Chiu KL, Bradley TD (2005) Association between atrial fibrillation and central sleep apnea. Sleep 28: Mansfield DR, Solin P, Roebuck T, Bergin P, Kaye DM, Naughton MT (2003) The effect of successful heart transplant treatment of heart failure on central sleep apnea. Chest 124: Gilmartin GS, Daly RW, Thomax JR (2005) Recognition and management of complex sleep-disordered breathing. Curr Opin Pulm Med 11: Issa FG, Sullivan CE (1986) Reversal of central sleep apnea using nasal CPAP. Chest 90: Eckert DJ, Jordan AS, Merchia P, Malhotra A (2007) Central sleep apnea: pathophysiology and treatment. Chest 131: Sankri-Tarbichi AG, Rowley JA, Badr MS (2009) Expiratory pharyngeal narrowing during central hypocapnic hypopnea. Am J Respir Crit Care Med 179: Badr MS, Toiber F, Skatrud JB, Dempsey J (1995) Pharyngeal narrowing/occlusion during central sleep apnea. J Appl Physiol 78: Luo YM, Tang J, Jolley C et al (2009) Distinguishing obstructive from central sleep apnea events. Diaphragm electromyogram and esophageal pressure compared. Chest 135: Javaheri S, Smith J, Chung E (2009) The prevalence and natural history of complex sleep apnea. J Clin Sleep Med 5: Dernaika T, Tawk M, Nazir S, Younis W, Kinasewitz GT (2007) The significance and outcome of continuous positive airway pressure-related central sleep apnea during split-night sleep studies. Chest 132: Morgenthaler TI, Gay PC, Gordon N, Brown LK (2007) Adaptive servoventilation versus noninvasive positive pressure ventilation for central, mixed, and complex sleep apnea syndromes. Sleep 30: Pusalavidyasagar SS, Olson EJ, Gay PC, Morgenthaler TI (2006) Treatment of complex sleep apnea syndrome: a retrospective comparative review. Sleep Med 7: Allam JS, Olson EJ, Gay PC, Morgenthaler TI (2007) Efficacy of adaptive servoventilation in treatment of complex and central sleep apnea syndromes. Chest 132: Randerath WJ, Galetke W, Rühle KH (2003) Auto-adjusting CPAP based on impedance versus bilevel pressure in difficult-totreat sleep apnea syndrome: a prospective randomized crossover study. Med Sci Monit 9:CR353 CR Zirlik S, Schahin SP, Premm W, Hahn EG, Fuchs FS (2009) Lung volumes and mean apnea duration in obstructive sleep apnea. Respir Physiol Neurobiol 168: Bitter T, Westerheide N, Faber L, Hering D, Prinz C, Langer C, Horstkotte D, Oldenburg O (2010) Adaptive servoventilation in diastolic heart failure. Eur Respir J 36: Solin P, Jackson DM, Roebuck T, Naughton MT (2002) Cardiac diastolic function and hypercapnic ventilatory responses in central sleep apnoea. Eur Respir J 20: Hall MJ, Xie A, Rutherford R, Ando S, Floras JS, Bradley TD (1996) Cycle length of periodic breathing in patients with and without heart failure. Am J Respir Crit Care Med 154: Arzt M, Schulz M, Schroll S, Budweiser S, Bradley TD, Riegger GA, Pfeifer M (2009) Time course of continuous positive airway pressure effects on central sleep apnoea in patients with chronic heart failure. J Sleep Res 18: Johnson KG, Johnson DC (2005) Bilevel positive airway pressure worsens central apnoeas during sleep. Chest 128: