INTRODUCTION — Central sleep apnea (CSA) is a disorder characterized by repetitive cessation or decrease of both airflow and ventilatory effort during sleep. It can be primary (ie, idiopathic CSA) or secondary. Examples of secondary CSA include CSA associated with Cheyne-Stokes breathing, a medical condition, a drug or substance, or high altitude periodic breathing [1].
CSA associated with Cheyne-Stokes breathing is particularly common among patients who have heart failure or have had a stroke. It is characterized by central apneas that occur during the decrescendo portion of the cyclic crescendo-decrescendo respiratory pattern. (See "Classification of sleep disorders", section on 'Sleep-related breathing disorders' and "Sleep-disordered breathing in heart failure".)
CSA can alternatively be categorized as hyperventilation- or hypoventilation-related. Hyperventilation-related CSA encompasses most of the types of CSA mentioned above. Hypoventilation-related CSA occurs in disorders in which there is alveolar hypoventilation that is so severe that central apneas occur when the patient falls asleep because the wakefulness stimulus to breathe disappears. Central apneas tend to be a minor component of such disorders. Examples of contexts in which hypoventilation-related CSA may occur include neuromuscular disease, such as muscular dystrophy or myasthenia gravis, and severe abnormalities in pulmonary mechanics (eg, kyphoscoliosis). (See "Central sleep apnea: Pathogenesis", section on 'Hyperventilation-related central apnea' and "Central sleep apnea: Pathogenesis", section on 'Hypoventilation-related central apnea'.)
CSA associated with a drug or substance tends to have features of both hyperventilation and hypoventilation. The evaluation and management of sleep-disordered breathing in this context is reviewed separately. (See "Sleep-disordered breathing in patients chronically using opioids".)
In this topic review, we present an approach to treating CSA and then we describe the evidence supporting each intervention. The definition, risk factors, clinical presentation, diagnosis, and pathogenesis of CSA are discussed separately. (See "Central sleep apnea: Risk factors, clinical presentation, and diagnosis" and "Central sleep apnea: Pathogenesis".)
GOALS OF THERAPY — Untreated central sleep apnea (CSA) disrupts sleep. This may lead to difficulty maintaining sleep (ie, awakenings) as well as daytime sequelae that include excessive daytime sleepiness, poor concentration, morning headaches, and increased risk for motor vehicle crashes or workplace accidents.
The goals of therapy in patients with CSA are to normalize sleep-related breathing patterns (ie, abolish central apneas, decrease or eliminate oxygen desaturations) and thereby improve both the quality of sleep and daytime symptoms and function.
The urgency with which to initiate therapy for CSA depends upon the severity of daytime symptoms as well as the severity of the physiological sequelae (eg, oxyhemoglobin desaturation during sleep) attributable to CSA.
●In patients with mild symptoms and sequelae, initial treatment can be directed at optimizing the underlying condition that is causing or exacerbating CSA (eg, heart failure). If the CSA persists despite such therapy, or if the condition is not expected to improve with additional therapy, CSA-specific therapies (eg, positive airway pressure therapy) are indicated in an attempt to improve symptoms.
●In patients with severe consequences of CSA (eg, prolonged severe oxyhemoglobin desaturation precipitating myocardial ischemia, nocturnal arrhythmias), CSA-specific therapy is typically started at the same time as attempts to treat or optimize the underlying medical condition.
ADDRESSING THE UNDERLYING CAUSE — Most cases of CSA are secondary to an underlying medical condition, central nervous system pathology, or medication side effect. In such cases, treatment of the underlying condition or removal of the offending medication or substance may result in improvement in the CSA. The extent to which CSA is expected to improve varies by the condition but is rarely complete.
In patients with heart failure, a variety of interventions (ie, medical therapy, cardiac resynchronization, ventricular assist devices or transplantation) have been associated with improvements in the severity of sleep apnea. For the most part, however, these interventions do not lead to complete resolution of the abnormal breathing pattern and should be considered complementary to CSA-specific therapy. (See "Sleep-disordered breathing in heart failure", section on 'Management'.)
In patients with CSA related to central nervous system suppressing drugs or substances, attempts should be made to eliminate or at least reduce the offending drugs. (See "Sleep-disordered breathing in patients chronically using opioids", section on 'Treatment'.)
PATIENTS WITH HYPERVENTILATION-RELATED CSA — Hyperventilation-related CSA (ie, high loop gain CSA) is the most common form of CSA. It includes primary CSA and CSA associated with Cheyne-Stokes breathing, a medical condition such as heart failure, or high altitude periodic breathing. (See "Central sleep apnea: Risk factors, clinical presentation, and diagnosis", section on 'Diagnostic criteria'.)
Continuous positive airway pressure — Continuous positive airway pressure (CPAP) is the preferred first-line therapy for symptomatic patients with hyperventilation-related CSA. This approach is largely based upon the extrapolation of evidence from studies that focused on CSA in patients with heart failure [2]; there is a paucity of direct data regarding the treatment of patients with other types of hyperventilation-related CSA.
Efficacy — The rationale for CPAP as first-line therapy is based on results of a limited number of randomized trials, mostly small, which have consistently demonstrated that CPAP decreases the frequency of central apneas in patients with hyperventilation-related CSA associated with heart failure [3-7].
In the largest trial (CANPAP), 258 patients who had CSA associated with heart failure (mean apnea-hypopnea index [AHI] of 40 events per hour) were randomly assigned to receive CPAP or no CPAP and followed for a mean of two years [4]. At three months, the CPAP group had greater improvement in the AHI from baseline (-21 versus -2 events per hour, p<0.001), improved mean nocturnal oxygen saturation (1.6 versus 0.2 percent, p<0.001), and longer six-minute walk distance compared with the control group. There were no differences in quality of life, transplant-free survival, or hospitalizations.
A major limitation of the CANPAP study was that the CPAP level was not titrated to effect. In a post hoc analysis in which patients were stratified according to whether or not the CPAP had sufficiently treated the CSA (defined as reducing the AHI to fewer than 15 events per hour three months after randomization), there was significantly better transplant-free survival among patients who were sufficiently treated with CPAP (hazard ratio 0.37, 95% CI 0.14-0.97) [8]. This was a not a pre-planned analysis, however, and therefore it remains possible that factors other than CPAP were responsible for the observed difference.
Other patient-important outcomes, including quality of life, have been inadequately studied. The effect of CPAP on heart failure-related outcomes (eg, ejection fraction) is described separately. (See "Sleep-disordered breathing in heart failure", section on 'Positive airway pressure therapy'.)
In contrast to CSA in patients with heart failure, there is a lack of direct data regarding the impact of CPAP on other types of hyperventilation-related CSA. CPAP is also suggested as first-line therapy in such patients, however, based on available data in heart failure patients [2].
Mechanism of effect — CPAP reduces the frequency of central apneas, probably by preventing pharyngeal airway narrowing and occlusion during a central apnea. The mechanism of action of CPAP in patients with CSA can be conceptualized as follows:
●The pharyngeal airway typically narrows and may occlude during a central apnea [9]. As a result, greater negative airway pressure must be generated for breathing to resume [10].
●Greater negative airway pressure induces airway deformation, which can trigger hyperpnea, ventilatory overshoot, hypocapnia, and additional central apneas [11].
●By preventing pharyngeal airway narrowing, CPAP mitigates the need for greater negative airway pressure and its sequelae. The observation that arterial carbon dioxide tension increases after initiation of CPAP therapy supports this theory [3].
Titration — The goal of CPAP titration is to determine the minimal pressure required to resolve all apneas, hypopneas, and other sleep-related respiratory events, in all stages of sleep and in all sleep positions. There is no established method for CPAP titration in patients with CSA and no comparative data between various titration methods.
Most patients should undergo titration for CPAP in an attended in-laboratory setting, which offers an opportunity to troubleshoot mask interfaces and quickly correct or adjust settings. Unattended in-home or inpatient CPAP initiation is an alternative if in-laboratory titration during polysomnography is not possible. CPAP titration in patients with CSA is reviewed in detail separately. (See "Mode selection for positive airway pressure titration in adult patients with central sleep apnea syndromes".)
Supplemental oxygen during sleep — Supplemental oxygen during sleep is indicated for patients with hyperventilation-related CSA who have hypoxemia during sleep. It can be used along with positive airway pressure therapy and is also indicated for patients who do not tolerate or fail positive airway pressure therapy.
Supplemental oxygen during sleep not only mitigates hypoxemia during sleep, but it may also reduce the AHI. This has been demonstrated by several small randomized trials and crossover studies in patients with CSA due to heart failure [12]. The relative reduction in AHI achieved by oxygen therapy in various small studies ranged from 30 to 80 percent compared with baseline [12]; in the largest study (n = 36), the mean AHI decreased from 49 to 29 events per hour [13]. The quality of the evidence is limited by small sample sizes and use of short-term and surrogate outcomes. A multicenter trial of low-flow nocturnal oxygen therapy in patients with heart failure and CSA (LOFT-HF) is ongoing [14,15].
It seems reasonable that supplemental oxygen during sleep may reduce the symptoms of disrupted sleep since it improves the AHI, but this has not been measured directly. There are few data regarding the effects of supplemental oxygen during sleep in patients with other types of hyperventilation-related CSA.
The mechanism by which supplemental oxygen during sleep improves CSA is unknown. It has been hypothesized that supplemental oxygen during sleep either increases the cerebral carbon dioxide tension, thereby preventing the hyperventilation and ventilatory overshoot that results in central apneas, or reduces carbon dioxide chemoreflex sensitivity [2].
Insurance coverage for nocturnal oxygen may be difficult in some cases and varies by region. In the United States, Medicare criteria govern coverage for long-term oxygen therapy. For patients with awake arterial oxygen tension (PaO2) ≥56 mmHg or arterial pulse oxygen saturation (SpO2) ≥89 percent, provision of supplemental oxygen during sleep requires that additional testing (during sleep) show either of the following [16]:
●A decrease in PaO2 to 55 mmHg or below (or SpO2 to 88 percent or below) for five minutes or longer during sleep.
●A decrease in PaO2 by 10 mmHg from baseline (or SpO2 by more than five percentage points from baseline) for at least five minutes during sleep and associated with symptoms or signs reasonably attributable to hypoxemia. Some examples of symptoms are impairment of cognitive processes, nocturnal restlessness, and insomnia; some examples of signs are cor pulmonale, "P" pulmonale on electrocardiogram (EKG), documented pulmonary hypertension, and erythrocytosis.
Additional aspects of long-term oxygen therapy are discussed separately. (See "Long-term supplemental oxygen therapy".)
CPAP failure or intolerance — The optimal therapy for patients with CSA who fail or do not tolerate continuous positive airway pressure (CPAP) is not known, and there are few comparative studies upon which to base recommendations. Further studies to help define patient selection and long-term safety and effectiveness are needed. Treatment decisions should be individualized based on the underlying etiology of the CSA and individual patient characteristics.
Patients with ejection fraction ≤45 percent — For patients with CSA due to heart failure with reduced ejection fraction (≤45 percent) who do not tolerate or respond to CPAP, the optimal approach is uncertain.
Based on results of the SERVE-HF randomized trial discussed below, we recommend not using adaptive servo-ventilation (ASV), which had previously been used as a second-line therapy for patients with CSA who had failed or did not tolerate CPAP. This recommendation is consistent with updated clinical practice guidelines of the American Academy of Sleep Medicine, which recommend against the use of ASV to treat heart failure-associated CSA in patients with an ejection fraction ≤45 percent and moderate or severe CSA [17]. (See 'Supplemental oxygen during sleep' above and 'Addressing the underlying cause' above.)
In these patients, supplemental nocturnal oxygen may be the best approach. It may also be beneficial to revisit whether heart failure can be further optimized with medical therapy. (See 'Supplemental oxygen during sleep' above and 'Addressing the underlying cause' above.)
Increased caution when considering the use of ASV in patients with heart failure is based on results of SERVE-HF, a multicenter, open-label trial that randomly assigned 1325 patients with moderate to severe predominant CSA (AHI ≥15, central apnea index >10) and symptomatic heart failure (defined as an EF ≤45 percent and New York Heart Association [NYHA] Class III, NYHA class IV, or NYHA class II with ≥1 hospitalization for heart failure in the previous 24 months) to ASV plus standard medical therapy or medical therapy alone [18]. The primary endpoint was 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. Secondary endpoints included time to death from any cause, time to death from cardiovascular causes, change in NYHA class, and change in timed-walk distance. Results included the following:
●The mean age of the patients was 69 years, and approximately 90 percent were male. The majority of patients had NYHA class III heart failure, and the mean ejection fraction was 32 percent. Approximately 50 percent of the patients in both groups had an implanted defibrillator device. The mean baseline AHI in the two treatment groups was 31 to 32 events per hour, and central AHI to total AHI ratio was 81 to 82 percent.
●ASV effectively reduced the number of apneas and hypopneas as measured by polysomnography: the mean AHI in the ASV group decreased from 31 events per hour at baseline to 6.6 events per hour at 12 months.
●The incidence of the primary endpoint was similar in the ASV group compared with controls (54 versus 51 percent; hazard ratio [HR] 1.13, 95% CI 0.97-1.31). In addition, ASV did not improve symptoms or ejection fraction.
●Compared with controls, patients assigned to ASV had significantly higher all-cause mortality (35 versus 29 percent; HR 1.28, 95% CI 1.06-1.55) and cardiovascular mortality (30 versus 24 percent; HR 1.34, 95% CI 1.09-1.65).
The mechanism whereby ASV led to an increase in mortality in this trial is not known. One possibility is that ASV led to a decrease in cardiac output and stroke volume in susceptible patients, such as those with low preload. Another possibility is that Cheyne-Stokes respiration serves a compensatory function in heart failure, and that ASV proved detrimental by diminishing this compensatory respiratory pattern [19,20]. In the SERVE-HF trial, a subgroup analysis found a positive association between the proportion of Cheyne-Stokes breathing and the risk of cardiovascular mortality [18].
A much smaller trial in 126 hospitalized heart failure patients with predominant CSA (81 percent with reduced ejection fraction) randomized patients to receive ASV plus optimal medical therapy or medical therapy alone [21]. While the trial showed no difference in the primary outcome of death, cardiovascular hospitalizations, and timed walk distance, the confidence intervals were wide and there was a suggestion of increased harm in the ASV group (HR 1.06, 95% CI 0.75-1.51). The trial was stopped early, in part due to the SERVE-HF trial results.
The use of bilevel positive airway pressure (BPAP) with a back-up respiratory rate in patients with CSA due to heart failure with reduced ejection fraction should also be approached with caution and on a case-by-case basis. One reason for concern is the analogous mechanism of effect between ASV and BPAP with a back-up rate, the relative paucity of data on the effectiveness of BPAP on patient-important outcomes, and the data that ASV may cause harm in these patients. Given the limited available therapies in this patient group, however, BPAP with a back-up rate may be the only available option for some patients. (See 'Bilevel positive airway pressure (BPAP) with back-up rate' below.)
Patients with ejection fraction >45 percent — Treatment options for patients with CSA and an ejection fraction >45 percent who fail or do not tolerate CPAP include ASV and BPAP with a back-up respiratory rate. In addition, all patients with hypoxemia during sleep should receive supplemental oxygen during sleep. (See 'Supplemental oxygen during sleep' above.)
Adaptive servo-ventilation — Adaptive servo-ventilation (ASV) provides a varying amount of inspiratory pressure superimposed on a low level of CPAP, with a back-up respiratory rate. The magnitude of the inspiratory pressure is reciprocal to the amount of respiratory effort, determined over a three- to four-minute moving window (figure 1). (See "Mode selection for positive airway pressure titration in adult patients with central sleep apnea syndromes", section on 'Devices'.)
ASV remains an option in patients with hyperventilation-related CSA and a preserved ejection fraction, although treatment decisions in such patients should be individualized, and there is a paucity of direct data in these patients [17].
Patients who are already using ASV for other indications (eg, heart failure with preserved ejection fraction, primary CSA, treatment-emergent CSA) should be informed about the safety signal from the SERVE-HF trial; in some cases the balance of risks and benefits may still favor ASV therapy, particularly in patients who are responding to therapy and have failed prior CPAP. (See "Treatment-emergent central sleep apnea", section on 'Treatment'.)
Previously published randomized trials and uncontrolled trials in patients with hyperventilation-related CSA due to heart failure had demonstrated that ASV decreases the frequency of central apneas [22-30]. In a systematic review, a meta-analysis of nine trials (127 patients) found that ASV decreases the AHI by a mean of 30 events per hour [2]. In small studies, ASV also appeared to improve symptoms of disrupted sleep, left ventricular ejection fraction, exercise capacity [22,31,32], and arrhythmic events in patients with implanted cardioverter-defibrillator devices [33]. However, the impact of ASV on other patient-important outcomes (ie, quality of life) is uncertain because such outcomes have not been studied.
Fewer studies have directly compared ASV to CPAP in heart failure patients with sleep-disordered breathing [29,34]. A meta-analysis of randomized studies found a small advantage to ASV over CPAP in two studies (weighted mean difference in AHI of -0.65 events/hour favoring ASV, 95% CI -1.06 to -0.25) [32]. No randomized studies have compared the two treatments directly in heart failure patients with predominantly Cheyne-Stokes breathing (as opposed to obstructive sleep apnea (OSA) or a combination of obstructive and central breathing patterns).
As with CPAP, in-laboratory titration of ASV is preferred in patients with CSA in order to made individualized adjustments to pressure limits and assure good mask fit. ASV titration is reviewed separately. (See "Mode selection for positive airway pressure titration in adult patients with central sleep apnea syndromes", section on 'Titration modules'.)
Bilevel positive airway pressure (BPAP) with back-up rate — BPAP therapy is an option when used in the spontaneous timed (ST) mode (ie, with a back-up rate) targeted to normalize the apnea-hypopnea index (AHI); it should only be considered for the treatment of CSA if there is no response to CPAP or oxygen therapy.
BPAP delivers positive airway pressure at different levels during inspiration and expiration. The level during inspiration is called the inspiratory positive airway pressure (IPAP), and the level during expiration is called the expiratory positive airway pressure (EPAP). BPAP has two major effects. First, it splints the upper airway open. Second, it increases alveolar ventilation by augmenting the tidal volume. The tidal volume is directly related to the difference between the IPAP and EPAP. As an example, the tidal volume is greater when the IPAP is set at 15 cm H2O and the EPAP at 5 cm H2O (difference of 10 cm H2O) than when the IPAP is set at 10 cm H2O and the EPAP at 5 cm H2O (difference of 5 cm H2O). It is of note that BPAP in the spontaneous mode (ie, without a back-up rate) may induce hypocapnia and hence exacerbate central apnea. Therefore, a back-up respiratory rate is required if BPAP is used for the treatment of central apnea.
Available data on BPAP in patients with CSA includes two uncontrolled trials and one nonrandomized controlled trial of BPAP with a back-up respiratory rate in patients with CSA due to heart failure [35-37]. A meta-analysis of these three trials reported a mean decrease in the AHI of 44 events per hour [2]. In addition, improved exercise capacity was reported with BPAP in the only trial that measured this outcome [36]. There is a paucity of data regarding the effects of BPAP in patients with hyperventilation-related CSA other than CSA due to heart failure. (See "Central sleep apnea: Pathogenesis", section on 'Hyperventilation-related central apnea'.)
In patients with hyperventilation-related CSA that is not due to heart failure, trials of BPAP are typically reserved for those who have failed or not tolerated trials of CPAP and ASV. In these patients, BPAP should only be used with a back-up respiratory rate because BPAP without a back-up rate may exacerbate hyperventilation, hypocapnia, and central apnea by augmenting the tidal volume [38,39]. Titration of BPAP in patients with CSA is reviewed in more detail separately. (See "Mode selection for positive airway pressure titration in adult patients with central sleep apnea syndromes", section on 'Bilevel positive airway pressure (BPAP)' and "Mode selection for positive airway pressure titration in adult patients with central sleep apnea syndromes", section on 'Attended in-laboratory BPAP titration'.)
At present, the use of BPAP with a back-up rate in patients with CSA due to heart failure with reduced ejection fraction should be approached with caution and on a case-by-case basis. (See 'Patients with ejection fraction ≤45 percent' above.)
Pharmacologic therapy — Patients who do not tolerate or benefit from positive airway pressure therapy or supplemental oxygen during sleep may benefit from treatment with a respiratory stimulant, such as acetazolamide; however, such medications can have harmful side effects and should be monitored closely. There are no randomized clinical trials addressing long-term efficacy of pharmacologic therapy.
Acetazolamide is a carbonic anhydrase inhibitor and a weak diuretic. It causes mild metabolic acidosis, which stimulates respiration and decreases the frequency of central apneas [40,41]. Acetazolamide has been studied in patients with hyperventilation-related CSA, specifically CSA associated with Cheyne-Stokes breathing and primary CSA. A meta-analysis that included seven small trials in 160 patients with CSA found that acetazolamide reduced the central apnea index by 12 events per hour compared with placebo or pre-acetazolamide baseline (most trials were cross-over or pre-post design) [42]. All studies were short-term, with treatment periods ranging from three days to one month using acetazolamide doses ranging from 250 mg once to four times daily.
Phrenic nerve stimulation — An implantable device that causes diaphragmatic contraction via unilateral transvenous phrenic nerve stimulation (remedē System) may be an option in selected patients with symptomatic CSA who fail or do not tolerate CPAP or other therapies. While the device has been approved by the US Food and Drug Administration [43], additional studies on cardiovascular outcomes and long-term safety are needed to better determine the its role in relation to other therapies for CSA [44].
The device consists of a neurostimulator (similar in size and appearance to a cardiac pacemaker) implanted in the upper chest and connected to two leads, one of which delivers transvenous stimulation to the phrenic nerve to achieve diaphragmatic contraction similar to a normal breath. The device is programmed to deliver stimulation during sleep and senses respiration via a lead in a thoracic vein.
Regulatory approval was based on results of a multicenter trial in which 151 patients with moderate to severe CSA (AHI >20 events per hour, ≥50 percent central apneas) underwent device implantation and were randomly assigned to active stimulation or no stimulation for six months [45]. Treatment assignment was known to both patients and clinicians. The majority of patients were males (89 percent) with a mean age of 65 years and mean AHI of 46 events per hour; 64 percent had heart failure and 54 percent had a concomitant implantable cardioverter-defibrillation or other cardiac device. At six months, the stimulation group was more likely to achieve ≥50 percent reduction in AHI from baseline (51 versus 11 percent). In a per-protocol analysis that included 131 patients, secondary outcomes were also improved in the stimulation group compared with controls, including blinded sleep parameters (eg, mean AHI, arousal index, oxygen desaturation index) and unblinded patient assessments of quality of life and daytime sleepiness. Two patients in each group died during the six-month treatment period and the rate of serious adverse events was similar (8 versus 9 percent). Therapy-related discomfort was reported by over one-third of patients but resolved with device reprogramming in all but one.
Observational follow-up of the trial cohort suggests that reductions in AHI and improvements in patient-reported outcomes are sustained over 12 to 36 months [46-50].
As with other treatment options for patients who fail CPAP, the device has not yet been compared with CPAP or other therapies for CSA, and further studies are needed to help define optimal patient selection and long-term safety.
PATIENTS WITH HYPOVENTILATION-RELATED CSA — Hypoventilation-related CSA is less common and includes CSA associated with central nervous system diseases, central nervous system suppressing drugs or substances (such as opioids), neuromuscular diseases, or severe abnormalities in pulmonary mechanics. (See "Sleep-disordered breathing in patients chronically using opioids".)
In patients with CSA whose central apneas are due to hypoventilation, BPAP is first-line therapy. In such patients, BPAP may be used with or without a back-up respiratory rate, or in an assured tidal volume mode. We advocate using a back-up rate to ensure adequate ventilation if the patient's ventilatory output is insufficient to reliably trigger mechanical inspiration. Mode selection is reviewed in more detail separately. (See "Mode selection for positive airway pressure titration in adult patients with central sleep apnea syndromes", section on 'Patients with CSA and opioid use'.)
Management of patients with severe alveolar hypoventilation, with or without CSA, is described in greater detail separately. (See "Noninvasive ventilation in adults with chronic respiratory failure from neuromuscular and chest wall diseases: Patient selection and alternative modes of ventilatory support" and "Noninvasive ventilation in adults with chronic respiratory failure from neuromuscular and chest wall diseases: Practical aspects of initiation".)
Patients may also benefit from treatment with a pharmacological respiratory stimulant, such as acetazolamide, but such medications can have harmful side effects and should be monitored closely. (See 'Supplemental oxygen during sleep' above and 'Pharmacologic therapy' above.)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Sleep-related breathing disorders in adults".)
SUMMARY AND RECOMMENDATIONS
●Central sleep apnea (CSA) can be primary (ie, idiopathic CSA) or secondary. Examples of secondary CSA include CSA associated with Cheyne-Stokes breathing, a medical condition, a drug or substance, or high altitude periodic breathing. CSA associated with Cheyne-Stokes breathing is particularly common among patients who have heart failure and a low ejection fraction. (See 'Introduction' above.)
●Initial treatment of CSA should be directed at any condition that may be causing or exacerbating the CSA. If the CSA persists despite such therapy, CSA-specific therapies are indicated for patients with symptoms or significant physiological sequelae attributable to CSA (eg, daytime sleepiness, prolonged or repetitive oxyhemoglobin desaturation during sleep). (See 'Goals of therapy' above and 'Addressing the underlying cause' above.)
●The following recommendations are pertinent to patients with HYPERventilation-related CSA for whom it has been decided that CSA-specific therapy is indicated:
•For patients with CSA associated with heart failure, we recommend a trial of continuous positive airway pressure (CPAP) (Grade 1B). (See 'Continuous positive airway pressure' above.)
•We recommend not using adaptive servo-ventilation (ASV) in patients with CSA due to heart failure with reduced ejection fraction (Grade 1B). We also suggest not using BPAP with a back-up rate in these patients, based on the analogous mechanism of effect between ASV and BPAP with a back-up rate (Grade 2C). (See 'Patients with ejection fraction ≤45 percent' above.)
•Treatment options for heart failure patients with CSA and reduced ejection fraction who fail or do not tolerate CPAP include nocturnal oxygen and medical management of heart failure. (See 'Supplemental oxygen during sleep' above.)
•For patients with HYPERventilation-related CSA other than CSA due to heart failure with reduced ejection fraction (eg, heart failure with preserved ejection fraction, primary CSA), we suggest a trial of CPAP (Grade 2C). For those who fail or do not tolerate CPAP, we suggest a trial of ASV (Grade 2C). However, some patients may reasonably choose to avoid ASV therapy until further data are available regarding known risks of ASV in other patient populations. For those who fail or do not tolerate both CPAP and ASV, we suggest a trial of BPAP with a back-up respiratory rate (Grade 2C). (See 'Continuous positive airway pressure' above and 'Patients with ejection fraction >45 percent' above.)
•For all patients with HYPERventilation-related CSA and hypoxemia during sleep, we suggest supplemental oxygen during sleep (Grade 2C). (See 'Supplemental oxygen during sleep' above.)
•For any patient with HYPERventilation-related CSA who does not tolerate positive airway pressure therapy and remains symptomatic, we suggest a trial of acetazolamide (Grade 2C). (See 'Pharmacologic therapy' above.)
●The role of implantable transvenous phrenic nerve stimulation in selected patients with symptomatic CSA remains to be determined based on long-term follow-up studies and comparative data. (See 'Phrenic nerve stimulation' above.)
●For patients with HYPOventilation-related CSA for whom it has been decided that CSA-specific therapy is indicated, we suggest BPAP as first-line therapy (Grade 2B). We advocate using a back-up respiratory rate. (See 'Patients with hypoventilation-related CSA' above.)
