| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
Dr. Javaheri is affiliated with the Sleep Disorders Laboratory, Department of Veterans Affairs Medical Center, Divisions of Pulmonary and Critical Care Medicine, and Cardiology, University of Cincinnati College of Medicine.
Correspondence to: Shahrokh Javaheri, MD, Professor of Medicine, Pulmonary Section (111F), VA Medical Center, 3200 Vine St, Cincinnati, OH 45220; e-mail: shahrokh.javaheri{at}med.va.gov
In 1907, Pembrey1 showed that supplemental nasal oxygen improved Cheyne-Stokes respiration in patients with congestive heart failure (CHF). Almost 100 years have passed, and there has been no long-term trial to determine whether treatment of central sleep apnea (CSA) with oxygen improves the natural history of CHF! Yet, CHF is highly prevalent and carries a poor prognosis,2 and CSA could be a potential contributory cause.3 Meanwhile, there are a number of studies showing that oxygen improves CSA and its pathophysiologic consequences.
Based on studies4
5
6
7
8
9
10
11
12
from different countries, sleep-related breathing disorders occur in about 40 to 80% of subjects with systolic heart failure. This wide range in part reflects the varied definitions of hypopnea. In our prospective study,4
requiring either a 4% drop in arterial oxyhemoglobin saturation and/or an arousal to define hypopnea for entrance, 49% of patients had an apnea-hypopnea index of 
15 episodes per hour. Using arbitrary polysomnographic criteria, 40% had CSA and 11% had obstructive sleep apnea.4
This distribution, however, is also quite variable4
5
6
7
8
9
10
11
12
and depends on a number of factors, including accuracy in scoring respiratory events (which at times is difficult) and the inclusion of obese subjects.
Studies have reported various effects of supplemental nasal oxygen on CSA in patients with systolic heart failure. Oxygen decreases periodic breathing and CSA, improves the hypnogram, virtually eliminates desaturation, decreases sympathetic activity, and improves exercise capacity.
Oxygen therapy improves CSA in patients with systolic heart failure.13 14 15 16 17 18 19 Hanly and associates14 should be credited for the first randomized, placebo-controlled study. In nine subjects with systolic heart failure, the authors showed that the administration of nasal oxygen for one night (when compared to nasal air) improved CSA, sleep architecture (ie, decreased arousals and shifted sleep to deep stages), and arterial oxyhemoglobin desaturation.
We studied 36 consecutive patients with systolic heart failure whose initial polysomnograms showed an apnea-hypopnea index of 15 episodes per hour.19
Nasal oxygen was administered at 2 L/min and was raised progressively to 4 L/min if disordered breathing resulted in desaturation < 90%. Subjects in this study had relatively severe periodic breathing with an average apnea-hypopnea index of 49 episodes per hour. Oxygen administration decreased periodic breathing with the most significant effect on CSA. While the patients were receiving oxygen, desaturation was virtually eliminated.
Although one night of oxygen administration decreased the apnea-hypopnea index of most subjects, in 40% of patients the index decreased to < 15 episodes per hour, a threshold we used to define clinically significant sleep apnea. We observed that the awake arterial PCO2 was significantly lower in patients who did not respond fully to oxygen than those who did (PaCO2, 36 ± 4 vs 39 ± 5 mm Hg, respectively; p = 0.03). In the only study15 that reported the reversal of sleep apnea with oxygen, the PaCO2 values of the subjects were within the normal range. In the studies13 14 16 17 in which oxygen therapy resulted in only partial response, the patients were hypocapnic. We found a similar trend for responsiveness with use of one night of therapy with continuous positive airway pressure (CPAP).20
In patients with CSA, the increase in sympathetic activity is multifactorial. Hypoxemia, hypercapnia, altered sleep stages (ie, lighter sleep), and arousals collectively contribute to increased sympathetic activity. Increased sympathetic activity is particularly relevant in systolic heart failure as it heralds poor prognosis.
In healthy subjects, induced hypoxemia and hypercapnia, and voluntary central apnea increase sympathetic activity.21 22 In this issue of CHEST (see page 366), Andreas and associates show that voluntary central apnea in subjects with heart failure also increases sympathetic activity, as measured by peroneal microneurography. However, the most important observation is that oxygen decreases sympathetic activity even though the central apnea becomes longer.
Staniforth et al23 measured plasma and urinary catecholamine levels in subjects with systolic heart failure and CSA. In this double-blind, placebo-controlled, crossover trial, subjects were randomized to 4 weeks of therapy with either nocturnal oxygen or air. Oxygen therapy resulted in a modest reduction in apnea-hypopnea index and the elimination of desaturation. Oxygen significantly decreased overnight urinary catecholamine excretion by about 50%. Plasma catecholamine levels did not change significantly. Urinary catecholamine levels may serve as a more representative indicator of overall nocturnal sympathetic activity than spot serum sampling in the morning. However, the latter has an independent prognostic value for survival in systolic heart failure. There are no data on urinary catecholamines.
Taken together, the studies of Andreas and colleagues and Staniforth et al23 indicate that in systolic heart failure, oxygen decreases sympathetic activity due to CSA.
In a double-blind, placebo-controlled, crossover trial,16 Andreas and associates randomized 22 subjects with systolic heart failure to 1 week each of nocturnal oxygen therapy or room air. After 1 week of treatment with oxygen, periodic breathing, desaturation, sleep architecture, and exercise capacity improved significantly (ie, rise in peak oxygen consumption, 835 to 960 mL/min). This finding is important since peak oxygen consumption is an independent prognostic variable in patients with systolic heart failure24 and coronary artery disease.25
The mechanisms of the therapeutic effects of oxygen on CSA are multifactorial. These include a rise in PCO2 and, presumably, a widening of the difference between the prevailing PCO2 and the PCO2 at the apneic threshold, the suppression of ventilatory response to hypercapnia, and increasing the body stores of oxygen.
An important determinant for the genesis of CSA is the difference between two PCO2 set points, the baseline prevailing PCO2 and the PCO2 at the apneic threshold.26 When the difference between these two PCO2 set points is wide, the likelihood of the occurrence of CSA decreases because a large ventilatory overshoot is necessary to drive PCO2 below a low apneic threshold. In contrast, when this difference is narrow, a small ventilatory overshoot, for example, occurring due to an arousal, may induce CSA.
In subjects with systolic heart failure, oxygen administration increases PCO2.15 27 In one study,27 PCO2 increased because tidal volume decreased. However, PCO2 may also increase if oxygen relieves hypoxic pulmonary arteriolar vasoconstriction, which may be present because of ventilation-perfusion mismatch.
Since a reduction in ventilation and a rise in PCO2 should increase the plant gain (which per se should increase the likelihood of developing periodic breathing), the overall therapeutic effect of oxygen in improving CSA should relate to a widening of the difference between the prevailing PCO2 and the apneic threshold PCO2. In this context, the studies of Xie and colleagues28 in humans and Nakayama and associates29 in dogs show that in hypoxemia-induced periodic breathing during sleep, the difference between the prevailing PCO2 and the PCO2 at the apneic threshold narrows. This occurs primarily because of a reduction in baseline PCO2 due to hypoxemia-induced hyperventilation. Therefore, if oxygen administration increases the baseline PCO2,16 27 with the PCO2 at the apneic threshold remaining relatively unchanged (an assumption), the difference between the two set points should increase.
As I noted earlier, in our study,19 hypocapnic subjects failed to fully respond to oxygen. I speculate that in these subjects, due to intense nonchemical ventilatory stimuli that are present in heart failure, oxygen failed to raise their baseline PCO2 in the transition from wakefulness to sleep.
Another reason for why oxygen may improve CSA is the suppression of the hypercapnic ventilatory drive 30 and the hypoxic ventilatory drive (which has not been systematically studied). Subjects with CSA have enhanced CO2 chemosensitivity.31 32 33 During arousals, enhanced CO2 (and hypoxic) chemosensitivity above eucapnia, will tend to lower the prevailing PCO2 and increase the likelihood of the occurrence of apnea during subsequent sleep.
Finally, oxygen administration increases the oxygen store in the body (ie, the oxygen content in the lungs and blood). This should dampen the change in PaO2 (for a given change in ventilation) and therefore decrease the likelihood of ventilatory instability.
Because oxygen decreases sympathetic activity and eliminates desaturation, long-term therapy has the potential to decrease the morbidity and mortality of subjects with CHF. Careful, randomized, placebo-controlled, multicenter studies with mortality as the end point require large numbers of patients, are labor-intensive, and are expensive,34 yet they need to be conducted. In this context, I am pleased to see the progress of the Canadian study35 to determine the effect of nasal CPAP on the natural history of systolic heart failure. Compared with oxygen, CPAP has the additional advantage of increasing the intrathoracic pressure and decreasing the transmural pressure of the intrathoracic structures. On the other hand, oxygen may have the advantage of improved long-term compliance, although this is an assumption. Furthermore, although afterload reduction is important, it has a small effect on survival in systolic heart failure. In the Studies of Left Ventricular Dysfunction study,34 the overall mortality rate of the placebo arm was 39.7% compared to 35.2% with enalapril (absolute risk reduction, 4.5%). In contrast, therapy with ß-blockers increases survival considerably.36 This improved survival, however, in part might have been due to an improvement in sleep-related breathing disorders. Because of various cardiorespiratory effects, such as increasing ejection fraction, decreasing pulmonary capillary pressure, and sympathetic activity, therapy with ß-blockers may improve CSA. However, with continued myocyte loss and fibrosis, as left ventricular function deteriorates, sleep-related breathing should recur.
The time has come to enact Pembrey’s dream.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
