|
Background The role of atrial overdrive pacing (AOP) in sleep apnea remains uncertain. We prospectively evaluated the effect of AOP after 24 hours and after one month in patients with the obstructive sleep apneahypopnea syndrome and compared it with the use of nasal continuous positive airway pressure (n-CPAP).
Methods We studied 16 patients with a moderate or severe case of the obstructive sleep apneahypopnea syndrome (baseline mean apneahypopnea index, 49) and normal left ventricular systolic function in whom a dual-chamber pacemaker had been implanted. After 48 hours, the patients were randomly assigned to AOP (pacing at 15 bpm above the spontaneous mean nocturnal heart rate) or backup atrial pacing (pacing at a heart rate below 40 bpm); the latter group began n-CPAP therapy one day later. After one month, the two groups switched therapies and were followed for an additional month. Polysomnographic studies were performed at baseline, on the first night after randomization, at crossover, and at the end of the study.
Results During AOP, no significant changes were observed in any of the respiratory variables measured. The change in the apneahypopnea index at one month with AOP was +0.2 (95 percent confidence interval, 2.7 to +2.3; P=0.87). In contrast, all variables improved significantly after one month of n-CPAP (change in the apneahypopnea index, 46.3; 95 percent confidence interval, 56.2 to 36.5; P<0.001).
Conclusions Nasal continuous positive airway pressure therapy is highly effective for the treatment of the obstructive sleep apneahypopnea syndrome, whereas AOP has no significant effect.
Our understanding of the pathophysiology of the obstructive sleep apneahypopnea syndrome is somewhat limited. Anatomical narrowing of the airway, increased collapsibility of the airway tissues, a disturbance in the reflexes that affect the caliber of the upper airway, and abnormalities of pharyngeal-muscle function all contribute to upper-airway occlusion during sleep.3 However, it has also been suggested that the obstructive sleep apneahypopnea syndrome is a systemic illness associated with the metabolic syndrome, rather than simply a local airway abnormality.4
Treatments that have been used for the obstructive sleep apneahypopnea syndrome include upper-airway surgery, weight loss, oral appliances, and nasal continuous positive airway pressure (n-CPAP). Since 1981, n-CPAP has been considered first-line therapy and has proved to be highly effective in alleviating symptoms, reducing morbidity and mortality, and improving the quality of life.5,6,7 However, patients' compliance with n-CPAP is low (ranging from 65 percent to 80 percent) because of problems with the nasal-mask interface, limited acceptance of treatment, and inadequate educational programs before n-CPAP titration.8,9
Accordingly, all efforts to develop new approaches to treatment are welcome. In one study, atrial pacing at a rate 15 beats higher than the mean sinus rate (atrial overdrive pacing [AOP]) during sleep was found to reduce significantly the number of episodes of apnea and hypopnea in a cohort of patients in whom pacing for conventional indications was already being used and who also had sleep apnea.10 Although a tentative pathophysiological explanation was given for these results, further investigation is necessary to confirm them and to determine whether they also apply over longer periods.
Our aim was to evaluate prospectively the effect of AOP after 24 hours and after one month in patients with the obstructive sleep apneahypopnea syndrome. We also sought to compare AOP with n-CPAP, the established form of therapy.
Methods
Patients
The study population included patients with moderate or severe obstructive sleep apneahypopnea syndrome and sleep-related bradyarrhythmias (sinus pauses longer than six seconds or transient atrioventricular block) who on initial consultation reported more than two episodes of syncope during the preceding year (Figure 1). Patients with heart failure, left ventricular systolic dysfunction (ejection fraction, <50 percent), or sleep apnea of the central type were excluded from the study. Patients were enrolled in the trial after written informed consent had been obtained.
|
Study Design
In all patients, a dual-chamber pacemaker was implanted because of a history of syncope and the presence of bradyarrhythmias on polysomnography (Figure 1). Although these findings are not considered standard indications for pacing, it is our practice to suggest pacing in this select population. The consent form for participation in the study noted that pacemaker implantation would not be universally recommended under these circumstances.
All pacemakers in study participants were initially programmed to initiate atrial pacing only when the intrinsic heart rate fell below 40 bpm. Two days later, the patients were randomly assigned with equal probability to one of two groups by means of a computer algorithm. In the first group, the pacemakers were programmed for AOP (pacing at a rate 15 bpm greater than the spontaneous mean nocturnal heart rate); in the second group, the programming was not changed, such that backup atrial pacing at less than 40 bpm was continued (Figure 1).
Twenty-four hours after randomization, the patients in the backup-pacing group began n-CPAP therapy after n-CPAP titration. Those in the AOP group did not receive n-CPAP. One month later, the two groups of patients switched therapies (i.e., the patients in the AOP group began n-CPAP, and the patients in the n-CPAP plus backup-pacing group began AOP), and follow-up was continued for an additional month (Figure 1).
Before commencement of n-CPAP therapy, the patients underwent an educational session from a qualified nurse in the n-CPAP clinic, who provided detailed explanation of the use of the n-CPAP device, adequate fitting of one of several types of available mask, and familiarization with the sensation of receiving n-CPAP. All patients spent two additional nights in the sleep laboratory after the initial n-CPAP setup so that any problems could be quickly identified and solved. Patients were instructed to contact the n-CPAP clinic nurse daily by telephone within the first week. Subsequently, they were seen two weeks and four weeks after the initiation of n-CPAP therapy. The bed partner (e.g., spouse) was included in the n-CPAP educational sessions and training and in the follow-up visits. In addition, a nurse from a home health care agency visited patients at home when needed.
Polysomnographic studies were performed at baseline, the first night after randomization, at crossover, and at the end of the study period. Compliance with n-CPAP was examined subjectively, on the basis of patients' responses on questionnaires, and objectively, on the basis of data downloaded from the built-in timer on the n-CPAP machine (Auto-CPAP Respironics). Information about sleep history and the score on the Epworth Sleepiness Scale (an eight-item questionnaire to evaluate subjectively the severity of daytime sleepiness; scores can range from 0 to 24, with higher scores indicating excessive daytime sleepiness) were also recorded at baseline, at crossover, and at the time of the final polysomnographic studies. The study was approved by the hospital ethics committee.
Statistical Analysis
To compare the effects of the two treatments, we used a classic crossover design in which the patients served as their own controls. In this repeated-measures design, each patient received one of the two treatments (AOP or n-CPAP) during different one-month periods (Figure 1). The treatment order was randomly assigned (either AOP followed by n-CPAP or n-CPAP followed by AOP).
Descriptive statistics for continuous variables are given as means ±SD. Ninety-five percent confidence intervals around the difference between the values for each intervention, as well as between the values for each intervention and baseline, were constructed to provide additional information about the magnitude of true mean differences. Repeated-measures analysis of variance was used to determine whether there was any treatment effect or order effect. Where findings for any variable were significant, post hoc Bonferroni adjusted tests for multiple comparisons (baseline vs. AOP, baseline vs. n-CPAP, and AOP vs. n-CPAP) were performed to pinpoint the differences. In case the distribution of any of the variables violated analysis-of-variance assumptions, Friedman's nonparametric test was also applied to confirm the significance of the results. A P value of less than 0.05 was considered to indicate statistical significance.
Results
There were 2329 consecutive patients referred to our hospital between July 2002 and October 2004 for the evaluation of symptoms related to sleep apnea (Figure 1). Of these, 41 had had more than two episodes of syncope during the preceding year. Twenty-three were excluded for other reasons, including refusal to participate (2 patients); thus, 16 patients were enrolled in the trial. The patients' baseline characteristics are listed in Table 1. Twelve of the participants were men, and the patients' mean (±SD) age was 60±11 years. The mean apneahypopnea index (the mean number of episodes of apnea and hypopnea per hour of sleep) was 49±19. Twelve patients had sinus pauses (mean duration, 13±3 seconds), and four had episodes of complete heart block.
|
During AOP, consistent pacing was suggested by a significant increase in the mean heart rate on the polysomnographic electrocardiographic recordings. The mean nocturnal heart rate during AOP in each patient was at least 11 bpm higher (mean increase, 17 bpm; range, 11 to 21) than the mean nocturnal heart rate at baseline and at least 14 bpm higher (mean increase, 18 bpm; range, 14 to 25) than the corresponding heart rate during n-CPAP therapy. Patients' rate of compliance with n-CPAP therapy was 100 percent.
Short-Term Effects of AOP
During the second polysomnographic study, two nights after pacemaker implantation, no significant changes were observed in the apneahypopnea index, arousal index (the mean number of arousals per hour of sleep), mean or lowest arterial oxyhemoglobin saturation, total sleep time, or partial pressure of arterial carbon dioxide in either the AOP group or the backup-pacing group (Table 2). The desaturation index decreased significantly (P=0.02) in the latter group, and the duration of rapid-eye-movement (REM) sleep decreased in seven of the eight patients in each group. No differences in the apneahypopnea index, arousal index, desaturation index, mean or lowest arterial oxyhemoglobin saturation, total sleep time, or partial pressure of arterial carbon dioxide were observed between the two groups during the same period. Patients in the backup-pacing group, though, had significantly more REM sleep than those receiving AOP (P=0.03).
|
There was a striking long-term difference between AOP and n-CPAP in all the variables measured in the study except for total sleep time (Table 3 and Figure 2). With n-CPAP, the apneahypopnea index decreased from 49.0 to 2.7 (net change, 46.3; 95 percent confidence interval, 56.2 to 36.5; P<0.001). All measured variables except total sleep time improved in every patient undergoing n-CPAP. Moreover, significant clinical improvement was noticed at follow-up as measured by the patients' Epworth Sleepiness Scale score (net change, 10.2; 95 percent confidence interval, 11.7 to 8.8; P<0.001). In contrast, AOP had no measurable effect on any of the variables. With AOP, the apneahypopnea index increased from 49.0 to 49.2 (net change, +0.2; 95 percent confidence interval, 2.7 to +2.3; P=0.87). There also was no significant change with AOP in the patients' Epworth Sleepiness Scale score (net change, +0.1; 95 percent confidence interval, 0.24 to +0.36; P=0.67).
|
|
Discussion
The pathophysiology of the obstructive sleep apneahypopnea syndrome is complex and not completely understood. Treatment is aimed mainly at reducing the number of episodes of apnea and hypopnea, the number of arousals, and oxyhemoglobin desaturation during sleep. These reductions have been correlated with a decrease in the rate of motor-vehicle accidents among persons with this syndrome13 as well as with improvements in blood pressure,14,15 baroreceptor sensitivity,16 and nitric oxide derivative production17; and reductions in sympathetic nervous system activation18,19,20 and cardiac arrhythmias (especially bradyarrhythmias).12,21,22 In recent years, the mainstay of therapy for the obstructive sleep apneahypopnea syndrome has been n-CPAP, which maintains a patent airway during sleep, thereby decreasing apnea. However, although n-CPAP is highly effective, compliance with the treatment is a problem.
Recently, an alternative treatment, based on the increase and stabilization of heart rate during sleep, has been proposed. The proposal followed the observation that some patients who had received a pacemaker with AOP to reduce the incidence of atrial tachyarrhythmias reported a reduction in breathing disorders after implantation of the pacemaker. Garrigue et al.10 found that in 15 patients with the sleep apnea syndrome, AOP significantly reduced the number of episodes of central or obstructive sleep apnea, without reducing total sleep time. Similar results had previously been reported by Kato et al.,23 who found that after physiologic cardiac pacing in three patients with central or obstructive apneas (a mixed type of the syndrome), the apneahypopnea index decreased as the average heart rate increased. With regard to patients with central sleep apnea, these results are predictable: when the heart rate increases with the use of AOP, factors that have been related to the occurrence of central sleep apnea also improve: cardiac output increases, circulation time shortens, and pulmonary congestion decreases. Furthermore, Sinha et al.24 recently reported that, in association with the improvement in cardiac function, cardiac-resynchronization therapy improved central sleep apnea in patients with heart failure and left bundle-branch block.
In contrast, the beneficial effect of AOP among patients with obstructive sleep apnea was somewhat unexpected. Although Garrigue et al.10 suggested several possible pathophysiological explanations, such as the vagolytic effects of pacing and the reduction of the ventilatory-loop gain and consequent stabilization of the respiratory pattern, confusion still remains: why, for example, has there been little success with theophylline in the treatment of the obstructive sleep apneahypopnea syndrome, even though it acts in roughly the same way as AOP. In addition, the suggested vagolytic effect of pacing has not been proved.
In our study, we used the same increase in paced heart rate as that tested by Garrigue et al.10 (15 beats above the mean nocturnal rate). Although our study was small and had power comparable to that of the study by Garrigue et al., we were unable to show any beneficial effect of pacing in reducing the number of episodes of apnea or hypopnea per hour or in increasing arterial oxyhemoglobin saturation during sleep. Moreover, there was no difference between the pacing and nonpacing phases in the number of arousals per hour. Furthermore, not only was AOP ineffective on the first night; it also showed no benefit even after one month.
The only variable that was altered by AOP at 24 hours was the duration of REM sleep. However, the same effect was also observed in patients who were not receiving AOP, which means that this alteration could be attributed to the intervention (pacemaker implantation) two days before randomization or perhaps simply to the patients' participation in the study, rather than to any particular therapy. Our results show that increasing the heart rate or reducing the variations in rate during sleep has no effect on the pathophysiology of obstructive sleep apnea, at least among patients with the characteristics of our study population that is, middle-aged, obese patients with normal left ventricular systolic function and a purely obstructive type of the syndrome. Such persons constitute the majority of patients with obstructive sleep apnea.
In the study by Garrigue et al.,10 most of the patients were older and had mild impairment of left ventricular ejection fraction and the mixed type of the syndrome. It seems that in such patients, cardiac pacing could reduce the number of secondary obstructive events by improving cardiac function and preventing the periodic breathing that may have a role in the pathophysiology of some cases of the obstructive sleep apneahypopnea syndrome.25 It should be noted that in a more recent report, Pepin et al.26 state that they were unable to find beneficial effects of pacing in another, more representative population of patients who had moderate-to-severe, predominantly obstructive sleep apnea and a mean left ventricular ejection fraction of 64±13 percent and who had received a pacemaker for conventional indications. Furthermore, Luthje et al.27 found that in patients with normal or impaired left ventricular function, AOP had no effect on the apneahypopnea index or on oxygen desaturation. The difference between the results reported by Luthje et al. and those reported by Garrigue et al. in relation to patients with impaired left ventricular function might be attributed to the higher mean nocturnal heart rate in the former study.
In our study, we preferred to compare pacing therapy with the more established n-CPAP therapy, rather than comparing pacing with no pacing in the long-term phase (one month), both for ethical reasons and because we were seeking a comparative standard of efficacy. In contrast to AOP, n-CPAP therapy was highly effective in reducing episodes of apnea and hypopnea, arousals, and oxyhemoglobin desaturation (while increasing the mean and lowest arterial oxyhemoglobin saturation), without reducing total sleep time. These findings are consistent with the benefits of n-CPAP noted in previous studies.5,6,7 It is likely, however, that the results of our study cannot be applied to all patients with sleep-related breathing disorders. Rather, they indicate the necessity of further investigating the possible therapeutic effect of AOP in other subgroups of patients.
In conclusion, n-CPAP therapy was highly effective in the treatment of patients with obstructive sleep apnea, whereas AOP had no effect on a representative population of such patients. The findings of our study may not apply in general to all patients with the obstructive sleep apneahypopnea syndrome. The data suggest that further investigation is needed in order to determine in which groups of patients AOP may have some therapeutic effect.
Source Information
From the Departments of Cardiology (E.N.S., S.I.C., G.I.C., N.C.K., P.E.V.) and Pneumonology (S.E.S., N.M.S.), Heraklion University Hospital, Heraklion, Crete, Greece.
Address reprint requests to Dr. Vardas at the Department of Cardiology, Heraklion University Hospital, P.O. Box 1352, Stavrakia, Heraklion, Greece, or at cardio{at}med.uoc.gr.
References
| |||||||||||||||||||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |