1

CSA is a devastating disease

Central sleep apnea (CSA) is a unique form of sleep apnea that occurs primarily in cardiovascular patients . CSA is of particular interest to the cardiac community because of its strong link to cardiac disease and cardiovascular outcomes. While there are some non-cardiovascular patient groups with CSA (stroke, idiopathic and opioid-induced), the majority of patients have underlying cardiovascular disease, primarily heart failure (HF) . CSA is found in up to 40% of HF patients, 30% of patients with atrial fibrillation, and in 50% of patients following a cerebrovascular accident . The five-year mortality in ambulatory, well-treated (NYHA Class II and III) HF patients with CSA is approximately 50% . CSA contributes to significant co-morbidities as well, including recurrent HF hospitalizations, atrial fibrillation, and reduced quality of life .

The pathophysiology and patient population of CSA differ from the more commonly known obstructive sleep apnea (OSA). While OSA occurs when the airway closes, CSA occurs when the brain fails to signal the diaphragm to contract and the patient quietly stops breathing (apnea). This apnea event is associated with hypoxia, hypercapnia, and arousal episodes. In the most common form of CSA, Cheyne–Stokes respiration, the patient experiences periodic apneas followed by periods of rapid breathing . Rapid breathing is the compensatory physiologic response to high levels of carbon dioxide (CO ₂ ). The brain is unable to sense and respond to changes in CO ₂ levels until they become significantly elevated, leading to an over-response to hypercapnia and rapid breathing. This over-response lowers CO ₂ levels to inappropriately low concentrations causing breathing to once pause again in an attempt to restore CO ₂ to the normal range. Instead, an oscillatory pattern develops between hyperventilation and apnea that propagates the vicious cycle. This characteristic, abnormal breathing pattern (often referred to as Cheyne–Stokes respiration) seen in CSA is starkly different from that seen in obstructive sleep apnea ( Fig. 1 ). The Cheyne–Stokes breathing pattern was first described in end-stage heart failure patients but also can be observed in patients with minimal or no heart failure symptoms . Due to the expanded use of sleep testing, CSA is usually diagnosed during sleep, but it can be seen in persons while awake, or even during exercise testing .

Examples of Breathing Patterns (Showing Abdominal Belt Movement and Nasal Flow).

While CSA is a significant co-morbidity of many cardiac conditions including heart failure, atrial fibrillation, and ventricular arrhythmias, it often goes unrecognized . Symptoms are often attributed to the underlying cardiac disease rather than the sleep disorder. Whilst healthy patients with OSA complain of excessive daytime sleepiness, patients with heart failure, rarely report this. Instead, they complain of “fatigue” or “low energy”. The other common symptoms include frequent nighttime arousals for no apparent reason or for nocturia, irritability, inability to focus, memory loss, shortness of breath or weight gain. Accordingly, patients with CSA usually have much lower scores on the Epworth Sleepiness Scale than would be expected for the severity of disease . The most common risk factors for CSA include heart failure, nocturia, and male gender . Multiple prevalence studies have shown that up to 40% of HF patients have CSA. Meanwhile, in a recent review of the Medicare database, it was noted that only 5% of hospitalized HF patients were tested for any type of sleep apnea. Of those tested, 97% had sleep apnea . Under-diagnosis of CSA may primarily be due to a lack of awareness of patients who are at risk and the significant health impact of the disease on future morbidity and mortality. Until recently, screening has required an in-lab sleep study for most cardiac patients, but positive data regarding outpatient sleep studies are increasing utilization of these studies .

Every episode of apnea or hypopnea causes discrete toxicity in the body. Each episode results in both hypoxia and hyperoxia and is associated with ischemic damage to both the cardiac and brain tissue . Cyclical episodes of hypercapnia have been shown to increase the likelihood of atrial fibrillation . In addition, each arousal stresses the myocardium and causes vasoconstriction due to the release of nor-epinephrine . This cyclical pattern of nor-epinephrine release, hypoxia, and hypercapnia contributes to cardiac strain and worsening cardiac function contributing to heart failure, and atrial and ventricular arrhythmias .

The long term effects of CSA are devastating. While the associated mortality has long been recognized, the increased morbidity has become clearer only with recent studies. Even with optimal care, the 5 year mortality remains at 50% in heart failure patients without CSA treatment . The 6-month re-hospitalization rate in heart failure patients with CSA is twice as high as for those without CSA . CSA patients have a higher rate of ventricular arrhythmias compared to patients without sleep apnea or with OSA .

2

Diagnosis of CSA

Diagnosis of CSA is typically assessed by a sleep study. These diagnostic studies have typically been performed in sleep labs in an outpatient setting. Central sleep apnea is diagnosed when the majority of apneic events are central as opposed to obstructive in origin ( Fig. 1 ) . Recently, there has been advocacy to move to outpatient sleep studies allowing patients to be tested at home with equipment similar to a Holter monitor. While in home studies have been shown to be highly predictive of sleep apnea, they assume that the patient is asleep throughout the study and may underestimate the severity of sleep apnea due to poor sleep .

Severity of the disease is measured in terms of the total number of apneas (no air movement) and hypopneas (reduced air movement) per hour. This measurement is known as the apnea–hypopnea index. Typically an apnea hypopnea index (AHI) of < 5 events/h is categorized as normal, 5–14 events/h mild, 15–30 events/h moderate, and > 30 events/h severe. In the US, some sleep laboratories will classify hypopneas, while others will rely only on apneas to determine whether a patient has CSA or OSA. Variations in how the sleep test is scored may lead to an underestimation of the CSA burden if the majority of hypopneas are considered obstructive in nature.

2

Diagnosis of CSA

Diagnosis of CSA is typically assessed by a sleep study. These diagnostic studies have typically been performed in sleep labs in an outpatient setting. Central sleep apnea is diagnosed when the majority of apneic events are central as opposed to obstructive in origin ( Fig. 1 ) . Recently, there has been advocacy to move to outpatient sleep studies allowing patients to be tested at home with equipment similar to a Holter monitor. While in home studies have been shown to be highly predictive of sleep apnea, they assume that the patient is asleep throughout the study and may underestimate the severity of sleep apnea due to poor sleep .

Severity of the disease is measured in terms of the total number of apneas (no air movement) and hypopneas (reduced air movement) per hour. This measurement is known as the apnea–hypopnea index. Typically an apnea hypopnea index (AHI) of < 5 events/h is categorized as normal, 5–14 events/h mild, 15–30 events/h moderate, and > 30 events/h severe. In the US, some sleep laboratories will classify hypopneas, while others will rely only on apneas to determine whether a patient has CSA or OSA. Variations in how the sleep test is scored may lead to an underestimation of the CSA burden if the majority of hypopneas are considered obstructive in nature.

3

Current therapeutic approach to CSA

It has long been thought that treatment of the underlying disease would treat or cure CSA, but there is a growing understanding that this therapeutic approach, effective for select patients, remains therapeutically inadequate in many others. Reversal or improvement of CSA has been seen following cardiac transplantation , but not in all patients . The prevalence of CSA however has remained stable even with current therapeutic regimens . Thus, it may be that treating underlying, but progressive diseases may delay, but not cure, CSA onset and severity. Following treatment of the underlying disease, a number of treatment approaches have been tried for CSA including medications, oxygen, and most commonly, mask-based therapies . Mask-based therapies present a number of limitations in this patient population. Continuous positive airway pressure (CPAP) may worsen or “unmask” CSA (in patients with predominantly OSA) . While initially this therapy was shown to reduce the number of apneic events per hour, further study failed to show long term improvement in morbidity or mortality except in patients who had optimum treatment of CSA with CPAP . A number of theories have been postulated as the cause for failure of this therapy including increased right ventricular pressure and substantial rates of patient non-compliance . This has led to the development of a “smarter” or “better” mask.

A newer mask-based therapy with a focus on CSA has been developed; it is designed to decrease overall pressure delivered and improve tolerance. This therapy, adaptive servo ventilation (ASV), decreases the amount of pressure blown into the patient’s airway unless additional pressure is needed to open the airway . In addition, a backup breathing rate is delivered if the patient has a prolonged apnea. This is important as patients with CSA can have high numbers of apneic events (often close to 1/min) as well as prolonged apneas (> 30 s) . Small studies have found these newer mask-based therapies may improve ejection fraction, heart failure symptoms, and mortality; larger randomized studies are still ongoing . In a cardiac population with multiple co-morbidities, however, patient adherence continues to be very poor with trials reporting long term use in fewer than 50% of indicated patients .