Heart failure (HF) is a growing epidemiological challenge in the US and worldwide. In the US, it is estimated that at least 5 million people are affected by it, with more than a half-million new diagnosis being made each year.1 Despite major improvements in the treatment of HF, almost 30 % of patients treated for acute HF require readmission within 30 days.2,3 These sobering statistics underscore the need for continuing developments in the treatment of HF. One strategy for improving outcomes in HF is to better address the treatment of co-existing conditions that are linked with it.4
Sleep disordered breathing (SDB) is now recognized as a significant category of co-morbidities associated with HF. Contained within the realm of SDB are both obstructive sleep apnea (OSA) and central sleep apnea (CSA). Clinically, CSA in HF is associated with a higher degree of mortality than OSA.5–8 The overall prevalence of SDB in HF is estimated between 60 and 75 %, but ranges considerably; specifically, OSA is thought to occur in about 35 % of HF patients and CSA in up to 66 %.9–13 Common to both forms of sleep apnea are periods of abnormal breathing, including changes in frequency and depth of breaths as well as apnea.14 However, the two forms differ in their cause; OSA is primarily a mechanical issue arising from upper airway obstruction, whereas CSA stems from neurological responses to changing partial pressure of carbon dioxide (PaCO2) levels.15,16 Breathing patterns between the two disorders are similar but unique to each process (see Figure 1). Good detailed reviews of OSA and its treatment have been published elsewhere.17 In this article, we will review the impact of CSA on HF and novel therapeutic options in development.
Central Sleep Apnea and Heart Failure
As mentioned, HF causes changes in PaCO2 levels resulting in CSA. The initial imbalance is thought to be hyperventilation, which may occur when filling pressures within the left side of the heart lead to pulmonary congestion.18–20 Hyperventilation in turn leads to hypocapnea; when PaCO2 levels drop below a certain threshold (apnea threshold), central chemoreceptors trigger compensatory apnea.21 As a consequence, relative hypoxemia and recurrent hypercapnea occur.15 The fluctuations in PaCO2 levels around the apnea threshold create a continuous cycle of hyperventilation and apnea periods. Clinically, this cycle manifests as Cheyne-Stokes breathing.22
Cheyne-Stokes breathing can occur at any time and is not limited to periods of sleep. However, the effect of increased left-sided filling pressures on breathing regulation is more pronounced in the recumbent position, which may help explain the higher frequency of breathing disturbances at night.23 Furthermore, HF is associated with a reduced ability of central chemoreceptors to appropriately adjust for changes in PaCO2 levels, leading to greater swings in levels that increasingly cross the apnea threshold.24 These tendencies give rise to the belief that CSA may be under-recognized and underdiagnosed in the HF population.10
Physiological consequences of CSA extend beyond fluctuations in PaCO2 levels. A detailed review, by Khayat et al., of CSA and its pathophysiological pathways can be found in this edition, with major points summarized here. Cardiac preload is directly affected by changes in intrathoracic pressures during both apneic and hyperventilatory phases, leading to increased myocardial strain and increased ventricular afterload, particularly of the right ventricle.25,26 Additionally, hypoxemia from apnea can lead to an increase in inflammatory markers, which are now thought to have deleterious effects in HF.27 But perhaps the most important consequence of SDB in HF is the effect on neurohormonal balance. Recurrent apneic cycles can lead to an increase in systemic sympathetic tone, with worsening apnea leading to higher degrees of sympathetic activation.25,28,29 Growing evidence suggests that SDB creates an internal environment in which circulating neurohormonal levels are elevated during sleep and remain high throughout the day.30,31 Downstream consequences include the upregulation of the renin–angiotensin–aldosterone system.16 Pharmacological neurohormonal blockade is the foundation of HF treatment;1,32 therefore, it is not surprising that CSA’s counteraction against this goal makes it an independent predictor of poor outcomes.1,5,8,33 Given these strong physiological links between CSA and HF, an important need exists for proven therapies.34
Current Treatment Evidence
The treatment of OSA is well established and aimed at improving mechanical airway obstruction causing cyclic apnea through the use of continuous or bi-level positive airway pressure (CPAP or BiPAP) support.15 Several investigations have noted improvements in cardiovascular function, ejection fraction, and autonomic tone as a result of CPAP therapy in OSA.35–39 The role of CPAP in CSA, however, is less certain. It is accepted that patients can present with combined forms of sleep apnea, and therefore CPAP has been used to treat CSA as well as OSA. Several small studies of CSA treatment suggested optimistic results of the use of CPAP on cardiovascular function, similar to those seen in the treatment of OSA.40-42 Unfortunately, the larger Canadian continuous positive airway pressure (CANPAP) trial did not show significant improvements with CPAP therapy in HF-associated CSA with regards to mortality, ejection fraction, and quality of life.43 As a result, CPAP is not considered part of the standard therapy for CSA.44 Additional controversy exists concerning the appropriate therapy for CSA, since the treatment of the underlying HF can result in sleep improvements.45,46
A newer method of ventilation, adaptive servo-ventilation (ASV), has shown early promise in the treatment of CSA. ASV attempts to mechanically mimic a patient’s minute ventilation through breaths of variable rate and tidal volume.47,48 Compared with CPAP, ASV resulted in improved markers of sleep and cardiac function.47,49 These initial small studies have led to the development of the Serve-HF study (Treatment of predominant central sleep apnoea by adaptive servo ventilation in patients with heart failure, ClinicalTrials.gov identifier: NCT00733343), a large, randomized, multicenter trial currently under way aimed at investigating improvements in morbidity and mortality through the use of ASV in patients with HF and CSA. Results are expected in 2014.50 Importantly, ASV attempts to externally correct the consequences of CSA (i.e., air flow hypercapnea) through mechanical ventilation, but does not attempt to correct the underlying loss of neuromuscular stimulation. Treatment focused on eliminating the occurrence of apneic episodes by stimulating natural breathing patterns may provide an even more unique and effective therapy.
Phrenic Nerve Stimulation—A Novel Approach
Historically used to treat patients with diaphragmatic paralysis from cervical spine injuries or central alveolar hypoventilation syndrome, phrenic nerve stimulation has been shown to be a safe, effective, and tolerated treatment modality.51–53 In CSA, stimulation of the phrenic nerve results in a reduction of apnea and the restoration of physiologic respiration.54 It has been hypothesized that implantable phrenic nerve stimulation devices may provide a novel treatment approach that could effectively be used to treat HF-associated CSA.
Traditional phrenic nerve stimulation systems have been surgically implanted, an approach that has inherent technical limitations and surgical complication risks. These risks may be increased in HF patients because of their underlying cardiac dysfunction.55 As an alternative, transvenous approaches to device implantation have been developed. The remedē™ System (Respicardia, Inc., Minnetonka, MN) is a fully implantable, unilateral transvenous phrenic nerve stimulation system that has been developed to treat CSA in HF patients. The system, shown in Figure 2, delivers an electrical stimulus to the phrenic nerve through a stimulation lead in order to prevent apneas. Compliance with therapy, which is a major limitation of CPAP and ASV, is not a concern with an implanted system;48,56 therefore, transvenous phrenic nerve stimulation may provide an avenue of treatment superior to other modalities, such as CPAP and ASV.57–59
The implantation of the remede System is technically similar to that of cardiac resynchronization therapy (CRT) devices. Venous access is obtained via either the right or left axillary or subclavian vein, through which the stimulation lead is passed until it reaches an appropriate location to provide unilateral phrenic nerve stimulation. Custom stimulation leads have been designed to deliver electrical stimulation through the vessel wall.60 Cannulation into the left pericardiophrenic vein is feasible from both the left and right axillary and/or subclavian regions, making it one of the two target sites for unilateral phrenic nerve stimulation.60 The other target site is the right phrenic nerve, which lies in an anterolateral orientation to the right brachiocephalic vein. The right pericardiophrenic vein, which is anatomically situated next to its respective phrenic nerve, is generally too small to allow the passage of a lead.61 A sensing lead is placed in the azygous vein.62 Both the sensing and stimulation leads are connected to a generator device, which is programmable so as to provide adequate phrenic nerve stimulation. The generator device is then implanted sub-pectorally on the chosen side.
It is anticipated that physicians proficient in implantation techniques of CRT devices would readily gain familiarity with the implantation of phrenic nerve stimulation systems. Contraindications to implantation include phrenic nerve paralysis, severe underlying lung conditions, primary pulmonary hypertension, and anatomical limitations that would prevent the accurate placement of the leads or generator.
Early studies evaluating the use of transvenous phrenic nerve stimulation to treat CSA in HF patients have yielded promising results. In one acute study, 16 patients underwent nocturnal polysomnography to assess the severity of CSA on consecutive nights following the implantation of a transvenous phrenic nerve stimulating lead. One night, phrenic nerve stimulation took place via the implanted lead, while on another night there was no stimulation. Comparison between the treatment night and the control night demonstrated a mean reduction of 88 % in CSA episodes as well as an improvement in related parameters—including a 48 % decrease in the apnea-hypopnea index (AHI)—during the treatment night.59 A significant decrease of 49 % in arousals was also noted. Such reductions carry particular clinical implications, given the link between arousals and norepinephrine levels.63 Furthermore, long-standing CSA can lead to PaCO2 chemoreceptor desensitization and dysfunction and thus to a worsening of the negative cycle in HF. Treatment with phrenic nerve stimulation aims to break the cycle of CSA in HF (see Figure 3) and allow for PaCO2 chemoreceptor re-equilibration. In patients using other implantable devices for the treatment of HF (notably a CRT device, with or without an implantable cardioverter defibrillator), the implantation of a phrenic nerve stimulation device did not affect the functioning of these other devices.64
A chronic study is now under way to evaluate the remede System (Safety and efficacy evaluation of Respicardia therapy for central sleep apnea, ClinicalTrials.gov identifier: NCT01124370), with results expected in 2012. Pending the results of this trial, ongoing research will be required to confirm benefits and clinical improvements in patients.
CSA is an underdiagnosed condition co-existent with HF and linked with an associated worsened prognosis. Evolving understanding of the underlying pathophysiology of CSA has strengthened the link between HF and CSA, and provided insights into the negative effects of CSA on cardiovascular function. Despite the large prevalence of CSA, treatment options are currently limited and clinically relevant outcomes have not yet been definitively demonstrated. New treatment modalities, including ASV and transvenous phrenic nerve stimulation, are being investigated. Transvenous phrenic nerve stimulation in particular, with its physiological mechanism and improved patient compliance, holds promise as a novel approach. Results of ongoing studies will be available soon to help clinicians provide adequate treatment to patients.