Introduction

Sleep disordered breathing (SDB), including central sleep apnea (CSA) and obstructive sleep apnea (OSA), is common in patients with heart failure (HF), although the diagnosis is frequently missed [1]. The prevalence of SDB is estimated to be as high as 47–76% among those with HF and reduced ejection fraction (EF) [2] and 55% in those with HF and preserved EF [3].SDB in congestive heart failure (CHF) can be broadly classified into two types: CSA with Cheyne-Stokes breathing (CSA-CSB), and OSA, and the two may exist together. CSB is characterized by crescendo-decrescendo changes in tidal volume that result in central apneas (lack of airflow without respiratory effort (Fig. 1). OSA is characterized by repeated pharyngeal airway collapse during sleep, resulting in repetitive episodes of oxygen desaturation despite ongoing respiratory effort, and arousals (Fig. 2).

Figure 1
figure 1

Polysomnogram showing crescendo–decrescendo pattern of breathing (as shown by the rib cage and abdomen movements), known as Cheyne Stokes breathing. Note the central apneas in which cessation of airflow (shown in nasal pressure channel) occurs without respiratory effort (in the rib cage and abdomen belt channels). In addition, desaturations are observed with each apnea, but somewhat delayed in time due to slowed circulation. Arousals from sleep commonly occur during periods of hyperpnea, yielding paroxysmal nocturnal dyspnea.

Full size image
Figure 2
figure 2

Polysomnogram showing obstructive apnea. Note the cessation of airflow for more than 10 s (detected by nasal thermistor, CFLOW channel) associated with continued respiratory effort as shown by thoracic and abdominal movements (detected in the thoracic and abdominal belt channels). There is also associated snoring. The top 6 channels (F4-M1 to O1-M2) show the electroencephalogram with stage 2 sleep.

Full size image

Although originally ascribed to reduced stimulation of central chemoreceptors in the setting of diminished cardiac output, CSA in CHF is now thought to be due to instability of the ventilatory system due to increased chemo-responsiveness (which means that the respiratory system is more responsive to even small changes in arterial partial pressure of carbon-dioxide [PaCO2]) [4, 5••, 6, 7]. Hypocapnia due to hyperventilation in response to stimulation of pulmonary vagal irritant receptors (J-receptors) by pulmonary edema [8–11] suppresses ventilation (undershoot), leading to central apneas and associated hypercapnia, which in turn stimulate excessive ventilation (overshoot). This ventilatory undershoot-overshoot cycle leads to CSB. Thus the prerequisites for CSB are pulmonary edema and an unstable ventilatory system. On the other hand, OSA in CHF appears to be due to narrowing of the upper airway due to unfavorable craniofacial structure, perhaps due to pharyngeal wall edema [12] and/or co-existent obesity. Pharyngeal dilator muscle dysfunction and ventilatory control instability also likely play a role, although they have been less carefully studied in CHF patients with OSA.

The presence of SDB in heart failure patients may be associated with adverse prognosis, perhaps due to worsening of ventricular function and heart failure symptoms. OSA may reduce systolic cardiac function by increasing afterload due to negative intrathoracic pressure generated during respiratory efforts against an occluded upper airway and is frequently associated with hypertension and atherosclerotic vascular disease [13, 14]. Both CSA and OSA may also be associated with a higher incidence of atrial and ventricular arrhythmias in HF patients [15, 16]. In some studies, SDB induces cardiac electrical instability (as assessed by indices such as T wave alternans), thus increasing the risk of sudden cardiac death [17]. Interventional studies have shown some improvement in ventricular ectopy with continuous positive airway pressure (CPAP) therapy [18]. Nasal CPAP therapy has yielded improvements in EF in small studies [19–21], although definitive trials are ongoing.

The first goal in the treatment of SDB in CHF is to optimize CHF treatment. Conservative measures for OSA such as weight reduction, avoidance of supine position during sleep, and avoiding alcohol and sedative medications before sleep, are also useful. Nocturnal CPAP therapy may be useful in treating SDB in CHF [1]. Other forms of PAP therapy, such as bilevel PAP (with a backup rate) [20, 21] or the newer adaptive servo-ventilation (ASV) (Table 1) [22–29], [30••], [31, 32], have also been used to treat SDB in CHF, although trials are ongoing regarding effects on outcome. Novel treatment modalities such as carbon dioxide therapy [33], acetazolamide [5••], and theophylline [34] have also been used, although the data are more limited.

Table 1 Published studies showing clinical outcomes with adaptive servo-ventilation in congestive heart failure
Full size table

Treatment

Optimal HF treatment

  • The first step in the management of SDB in CHF should be optimization of CHF treatment in accordance with published guidelines [35].

  • Administration of diuretics to restore normal fluid balance has been shown to improve OSA in CHF patients with hypervolemia [12].

  • Angiotensin-converting enzyme (ACE) inhibitors (eg, captopril) [36] and beta-blockers (eg, carvedilol) [37] reduce CSA in CHF.

  • Other treatment modalities for advanced heart failure (HF), such as atrial overdrive pacing (AOP), left ventricular assist device (LVAD), and cardiac resynchronization therapy (CRT), can additionally improve SDB. These assistive devices are used primarily as a treatment for HF and are not generally recommended to treat SDB in HF. However, their use may have a beneficial effect on SDB in HF.

Positive airway pressure therapy

Rationale :

PAP applies positive pressure to the upper airway and effectively acts like a pneumatic splint to maintain upper airway patency in OSA. It also reduces pulmonary edema by reducing cardiac preload and afterload, thus reducing CSA.

Evidence :

CPAP therapy has been shown to treat OSA, left ventricular EF (LVEF) [38, 39, 41, 44, 45] and reduce sympathetic activity in small studies of CHF [45]. CPAP also improves SDB [38–45] and reduces urinary catecholamine levels [40] in CHF with CSA. The Canadian Positive Airway Pressure Trial for patients with central sleep apnea and heart failure (CANPAP) trial randomized CSA patients with HF due to reduced EF to either CPAP or no CPAP. After 2 years of follow-up, CPAP treatment yielded no improvement in the primary outcome of transplant-free survival although CSA, EF, and exercise capacity were somewhat improved [46]. However, post hoc analysis of the CANPAP study revealed that adequate suppression of CSA by CPAP was associated with improved heart transplant–free survival [47••]. Thus, the data remain unclear whether elimination of apnea yields improved outcome or whether improvement in apnea suggests a good prognosis. Bilevel PAP (BPAP) is as effective as CPAP in improving circulation time, New York Heart Association functional class, and sleep quality in CHF with CSA [20]. Flow or volume-targeted dynamic BPAP may be more effective than CPAP or fixed BPAP in treating CSA in CHF, although outcome data are still evolving [21].

ASV is a new PAP technology that adjusts the delivered pressure support according to the ventilation of the patient. Various companies have slightly different algorithms for these devices, although no comparative effectiveness data are available regarding these devices. ASV reduces SDB [24–27, 30] and systemic inflammation [29] and improves LVEF [22, 28, 31], quality of life [24], exercise capacity [22, 27, 28], and respiratory instability [22] in CHF patients with CSA. ASV was also better than CPAP in improving parameters like SDB, LVEF, and treatment compliance in systolic CHF patients with SDB [31, 32]. Randomized trials are now underway to assess the impact of ASV on outcome of CHF patients with SDB [48, 49].

Standard dosage :

The level of CPAP (provides same PAP in both inspiration and expiration) or BPAP (can provide different PAP in inspiration and expiration) used to treat OSA in CHF is usually determined by titrating the PAP level that suppresses SDB. BPAP should generally be provided with a back up rate to avoid any potential worsening of breathing instability. Treatment of CSA with PAP often requires follow-up after weeks to months on therapy to determine the extent of resolution of CSA.

Contraindications :

There are no absolute contraindications for CPAP therapy to treat SDB in CHF. However, relative contraindications/caution is advised in active emesis, increased upper airway secretions (for example, due to poor cough or impaired swallowing), potential for upper airway obstruction (upper airway or head and neck tumors), active cardiac arrhythmias, and bullous lung disease. Hypovolemic patients may do poorly with CPAP therapy due to a further drop in preload.

Complications :

Dry nose, sore throat, nasal congestion, runny nose, sneezing, irritation of the eyes and the skin on the face, abdominal bloating, and nose bleeds may result due to PAP therapy, although these effects are generally self-limited.

Cost/cost-effectiveness :

The fixed cost of a CPAP/BPAP device with the tubing and the mask ranges between $500 and $2,500 depending on the device features and the manufacturer [50]. Some modest variable cost for replacement masks and hoses is also present.

Medications

Nocturnal oxygen therapy

Rationale :

Various mechanisms have been proposed to explain therapy of CSA by O2 administration, which include increase in baseline PCO2 by O2 administration, thus reducing the propensity for apnea caused by CO2 below the apnea threshold [51], and reduction of hypercapnic ventilatory drive [52].

Evidence :

Nocturnal oxygen therapy has been shown to reduce CSA and sympathetic activity [53, 54] and improve quality of life [54], LVEF [55], and exercise capacity in CHF patients with CSA [53].

Dosage :

The dosage of nocturnal oxygen used in the literature to treat CSA in CHF was 2 to 3 L/min [55].

Contraindications :

Active smokers. Caution advised in hypercapnic patients (monitor arterial PCO2 level).

Complications :

Elevation of PCO2 level in hypercapnic patients (headache, confusion), oxygen toxicity including cardiac suppression in CHF with high oxygen tensions.

Cost/cost-effectiveness :

The average monthly cost of nocturnal oxygen therapy is $200.

Carbon dioxide supplementation

Rationale :

CO2 supplementation can cause a small increase in baseline PCO2, which reduces CSA by the mechanism described earlier [51].

Evidence :

CO2 supplementation is done by adding dead space using a facemask attached to a cylinder of adjustable volume. The procedure reduced CSA and improved sleep quality [33], but is not commonly used due to the side-effects such as panic and insomnia.

Dosage :

The amount of CO2 rebreathed depends on the amount of dead space added (400–600 mL).

Contraindications :

Hypercapnic patients.

Complications :

Insomnia, anxiety.

Theophylline

Rationale :

Theophylline stimulates respiration by antagonizing adenosine. Though the exact mechanisms are unknown, an increase in ventilation by theophylline could decrease the likelihood of developing CSA during sleep by widening the difference between the prevailing PCO2 and the apneic threshold PCO2.

Evidence :

In a small randomized, placebo-controlled study, theophylline reduced CSA by 50% [34]. Theophylline is not commonly used for treatment of CSA in CHF due to its narrow therapeutic window and potential for arrhythmias.

Dosage :

3.3 mg/kg body weight orally twice daily for 5 days [34].

Contraindications :

Hypersensitivity to the drug, active peptic ulcer, porphyria.

Adverse effects/complications :

Anaphylaxis, seizures, insomnia, cardiac arrhythmias, nausea, vomiting, abdominal pain, diarrhea.

Cost :

$ 0.30 per 100-mg tablet.

Acetazolamide

Rationale :

Acetazolamide may be beneficial in treating CSA in HF by two mechanisms: 1) it decreases pulmonary venous congestion in HF (which contributes to CSA) through diuresis, and 2) it causes a metabolic acidosis that stimulates respiration. This effect increases the difference between prevailing PCO2 and apneic threshold PCO2, thus reducing propensity for central apnea.

Evidence :

Acetazolamide treatment for 6 days reduced CSA and daytime sleepiness and improved sleep quality in HF with reduced EF [5••].

Dosage :

3.5 mg/kg oral daily for 6 days [5••].

Contraindications :

Hypersensitivity to the drug, sulfa drug allergy, pregnancy, sickle cell anemia, liver, or kidney disease.

Adverse effects/complications :

Metabolic acidosis, taste alteration, numbness and tingling sensation, blurred vision.

Cost :

$0.50 per 125-mg tablet.

Interventional procedures

Atrial overdrive pacing

Rationale :

AOP can improve EF in HF, thus reducing pulmonary congestion and improving CSA.

Evidence :

AOP reduced CSA in CHF in one study [56], whereas there was no benefit in another study [57]. There was only modest improvement in OSA after AOP, although its effectiveness at reducing OSA is less than that of CPAP [58].

Contraindications :

Atrioventricular nodal block, refractory atrial tachyarrhythmias.

Complications :

Infection, air embolism, pneumothorax, myocardial perforation, vascular or nerve damage, thrombophlebitis, bleeding, premature atrial contractions, atrial tachycardia, premature ventricular contractions, ventricular tachycardia.

Left ventricular assist device

Evidence :

LVAD implantation improves CSA and exercise capacity in HF patients [59].

Cardiac resynchronization therapy

Evidence :

CRT has been shown to reduce CSA [60] as well as OSA [61] in HF patients with reduced EF. There is minor improvement in CSA when CRT is combined with AOP compared with CRT alone [62]. CRT also improves clinical outcome in HF patients in whom CSB persists despite modern optimal HF therapy [63].

Physical/speech therapy and exercise

Rationale :

Exercise may be useful by reducing body weight and thus improving OSA. It may also improve CSA by normalizing chemosensitivity [64].

Evidence :

A 6-month aerobic exercise program reduced CSA but not OSA in HF [65]. A 4-month exercise program improved OSA in HF patients with reduced EF [66]. Oropharyngeal exercises derived from speech therapy can improve OSA [67].

Surgery

Cardiac transplant surgery

Cardiac transplantation for end-stage HF may impact SDB, although it is not recommended solely to treat SDB in HF.

Evidence :

Cardiac transplantation may be useful in reducing CSA in HF, although CSA may still persist in some patients [68]. On the other hand, some HF patients may develop OSA after the surgery due to increased weight gain due to use of steroids and other immunosuppressants.

Contraindications :

Advanced liver or renal failure, insulin-dependent diabetes, peripheral vascular disease, recent thromboembolism, alcohol/drug/tobacco abuse, pulmonary hypertension.

Complications :

Infections, graft rejection, renal failure, complications (such as osteoporosis, hyperglycemia, hypertension and so forth) due to steroid therapy, allograft coronary artery disease, increased incidence of malignancies.

Pediatric considerations

  • The general algorithm for diagnosis and treatment of SDB in the pediatric population is similar to adults.