Published ahead of print on September 4, 2003, doi:10.1164/rccm.200211-1344OC
© 2003 American Thoracic Society Reduction of Sleep-disordered Breathing after PhysostigmineSleep Laboratory, Departments of Pulmonary Medicine and Clinical Pharmacology; and Sahlgrenska University Hospital, Göteborg, Sweden Correspondence and requests for reprints should be addressed to Jan Hedner, M.D., Ph.D., Department of Pulmonary Medicine and Allergology, Sahlgrenska University Hospital, S-413 45 Göteborg, Sweden. E-mail: jan.hedner{at}lungall.gu.se
The cholinesterase inhibitor physostigmine (PHYS) was investigated in a double-blind, placebo-controlled, randomized, crossover trial of 10 male patients with moderate to severe obstructive sleep apnea. PHYS (0.12 µg/minute/kg, 7-hour infusion) reduced mean apnea/hypopnea index (AHI) by 13.6 (95% confidence interval [CI], 2.2–25.1) corresponding to 21.4% (95% CI, –5.5 to 47.9) and increased minimum SaO2 by 8.7% (95% CI, –0.3 to 17.7) corresponding to 23.2% (95% CI, 4.8–41.3). During the last third of the night, coinciding with predicted plasma concentration steady state, non-REM sleep AHI decreased by 19.2 (95% CI, 0.1–38.3) or 14.9% (95% CI, -43.6 to 77.7) and REM AHI by 33.8 (95% CI, 13.7–54.0) or 67.5% (95% CI, 49.7–85.3). Mean total sleep time was reduced by 74 minutes (95% CI, 33.9–114.9), but patients with the least pronounced sleep shortening had the largest reduction of AHI (r = 0.73, p < 0.02). The nocturnal decline in heart rate was reduced by PHYS. Moreover, resting (early-night placebo heart rate) was positively correlated with proportional reduction of REM apnea index (r = 0.69, p < 0.02). Body mass index was negatively correlated with reduction of REM AHI (r = 0.77, p < 0.02). This, predominantly REM-related, reduction of obstructive sleep apnea after PHYS may provide a new treatment option if the effects are maintained in long-term studies.
Key Words: sleep-disordered breathing • apnea • treatment • acetylcholine • physostigmine Sleep-disordered breathing encompasses a wide spectrum of disturbances of breathing extending from modest and intermittent airflow restriction to complete upper airway collapse (obstructive sleep apnea [OSA]) during the sleeping period. The underlying pathophysiology is poorly understood, but a range of factors including altered upper airway structure and functional dynamics, reduced arousability from sleep, and destabilization of the respiratory control system have been proposed (1, 2). Thus, reduced upper airway apertures (3) as well as acquired functional changes including attenuated upper airway sensory function (4) and altered muscle structure (5) appear to be common in OSA. Electromyographic recordings from upper airway dilatory musculature during sleep have demonstrated reduced tonic and phasic activity coinciding with airway collapse (6). Other more recent work has focused on ventilatory oscillations arising from the feedback nature of the respiratory chemical control system that may provide a possibility for altered "loop gain" as a cause of repetitive upper airway collapse (7, 8). Physostigmine (PHYS), an inhibitor of acetylcholinesterase (AChE) activity, exhibits a number of pharmacologic effects including muscle contraction via muscle end-plate depolarization at the neuromuscular junction, increased salivary secretion as a result of cholinomimetic effects in major and intramucosal salivary glands (9). PHYS also increases sympathetic autonomic activity by a central mechanism of action (10). Other central effects include induction of REM sleep by shortening the time to REM sleep onset and increasing the relative amount of REM sleep (11). A recent study of patients with multiple system atrophy demonstrated an association between reduced thalamic nerve terminal density and severity of OSA, suggesting that decreased pontine cholinergic projections may contribute to sleep-disordered breathing in certain degenerative conditions (12). In this study we hypothesized that PHYS would reduce the severity of sleep-disordered breathing and tested this by performing a placebo-controlled trial of PHYS, administered by continuous overnight infusion, in patients with moderate to severe OSA. Some of the results of this study have previously been published as an abstract (13)
Subjects Patients referred to the sleep laboratory for an in-depth diagnostic classification of OSA were asked to participate in the study. Ten male patients (Table 1) consented to a study protocol approved of by the Sahlgrenska University Hospital Human Research Ethics Committee. The protocol included a single-center, randomized, double-blind, crossover study of PHYS and matching placebo (saline). Inclusion criteria were male sex, age 25 to 65 years inclusive, and previously diagnosed OSA defined as an oxygen desaturation (4%) index of 10 or more obtained from a preceding routine outpatient screening recording (EdenTec, Eden-Praire, MN). The screening study had to encompass at least 6 hours overnight recording of SaO2, respiratory movements (impedance), heart rate, nasal and oral airflow, breathing sounds, and body position. Exclusion criteria were coronary heart disease, previously known or treated cardiac arrhythmia of any type, myocardial infarction or stroke within 12 months before study enrolment, previous or present known or treated psychiatric disease, regular intake of bensodiazepines or related compounds, previously known intolerance to cholinesterase inhibitory agents, myasthenia gravis, a body weight exceeding 120% of ideal (Metropolitan scale), previously known renal or hepatic disease, and current alcohol or drug abuse. Previous exposure to continuous positive airway pressure was allowed but not closer than 1 month before the first study night.
Polysomnography Patients underwent a standard all-night polysomnography study on two occasions separated by a washout period of at least 2 but no more than 7 days. The monitored sleep studies included a two-channel EEG (C4/A1, C3/A2), a two-channel electrooculogram, a one-channel submental (EMG), leg EMG, SaO2, breathing movements (Respitrace; Ambulatory Monitoring, Ardsley, NY), nasal and oral air flow (Protech, Woodinvill, WA) and body position. All signals were preamplified, fed into a polysomnography system (Nightingale; Judex Datasystemer, Alborg, Denmark), stored and subsequently scored (30-second epochs) according to the Rechtshaffen and Kales scoring manual (14) by an experienced polysomnography technician before any unblinding of the study code. Arousals were documented and classified (15). Total sleep time (TST) was defined as the time elapsed between the first and last registered sleep epoch. An ECG signal was continuously recorded from standard leads, stored on the computer, and also displayed on a separate monitor during the study nights. An obstructive apnea was defined as loss of nasal flow accompanied by markedly reduced or paradoxic respiratory fluctuations in the Respitrace signals for more than 10 seconds. An obstructive hypopnea was defined as a more than 50% reduction in both the rib cage and abdominal Respitrace signals or a more than 50% reduction in only one of these channels but accompanied by chest wall paradoxic motion through most of inspiration, for more than 10 seconds. Heart rate was continuously calculated from the R–R interval and calculated as an average value for each of the three thirds of the night in each patient. Apneic (index; AI) and apneic–hypopneic event intensity (apnea/hypopnea index [AHI]) were expressed as number per hour spent asleep as well as asleep in a particular sleep stage (non-REM or REM). The ratio of REM to non-REM sleep AI was determined to identify subjects with REM sleep–predominant sleep disorder. The duration of the sleep study after commencement of drug or placebo infusion (420 minutes) was also divided into three parts (each 140 minutes) and in an ad hoc analysis data were analyzed separately for the first, second, and third parts of the infusion period (see below for application).
Study Procedure In four subjects, a plasma sample (10 ml whole blood) was withdrawn from an indwelling cannula placed in a superficial forearm vein in the contra lateral arm immediately before termination of the infusion (6:00 A.M.). Samples were immediately cold centrifuged (+4°C), plasma pipetted off, and stored at -70°C. The two sample sets obtained were subsequently analyzed for plasma concentration of PHYS using a previously described HPLC technique (17).
Statistical Analysis
Ten patients with moderate to severe OSA (Table 1) completed both study nights. There were no complications with the infusion procedure. Specific questioning on nausea, central nervous system effects and sleep disturbances, as well as open questioning revealed no side effects. Sleep quality assessed by a 5-degree scale ranging from worst to best possible did not differ between the study and control nights (data not shown). The average reduction in AHI after PHYS was 13.6 events/hour (95% confidence interval [CI], 2.2–25.1) corresponding to 21.4% (95% CI, –5.5 to 47.9) and average increase in minimum SaO2 by 8.7% (95% CI, –0.3–17.7) corresponding to 23.2% (95% CI, 4.8–41.3). The difference between the two study nights in total AHI remained within ± 10% in three patients, whereas in six patients the proportional reduction in total AHI ranged between 20 and 98%, and in one subject AHI increased (Figure 1) . However, in clinical terms only one patient restored AHI to a value below 10 per hour. PHYS appeared to have a more pronounced effect during REM sleep when AHI decreased from 54 ± 36 to 30 ± 26 compared with the reduction during non-REM sleep from 55 ± 25 to 44 ± 25 (Figure 2 , Table 2) .
TST was reduced by 74 minutes (95% CI, 33.9–114.9) during the PHYS night compared with the placebo night (Table 2). However, the reduction in TST was negatively correlated (r = 0.73, p < 0.02) with the reduction of AHI, and the greatest reduction in TST (exceeding 110 minutes) was seen in three patients who did not alter or had slightly increased AHI with PHYS (Figure 3) .
An ad hoc analysis, undertaken to account for the kinetic aspects of the continuous PHYS infusion, separated the full study night into three equal time portions. This analysis was based on the assumption that steady-state concentration of PHYS was not reached until after approximately 2 to 3 hours after the start of infusion. Indeed, there was a more pronounced effect in essentially all the evaluated variables during the last third of the night that would be expected to coincide with the kinetic steady-state condition. During this period the non-REM sleep AHI decreased by 19.2 events/hour (n = 10; 95% CI, 0.1–38.3) or 14.9% (95% CI, –43.6 to 77.7), whereas REM AHI decreased by 33.8 events/hour (n = 8, 95% CI, 13.7–54.0) or 67.5% (95% CI, 49.7–85.3) (Figures 4A and 4B) .
As expected, the mean heart rate decreased during the course of the night but less so after PHYS compared with placebo. The placebo to PHYS differences during the three thirds of the infusion period were 2.1 (95% CI, –0.04 to 4.2), 2.7 (95% CI, –0.03 to 5.4), and 4.2 (95% CI, 1.7–5.9), respectively. The relative REM apnea index reduction during treatment with PHYS was most pronounced (r = 0.69, p < 0.02) in subjects with the lowest baseline heart rate (defined as the mean value of the first third of the placebo night). There was no correlation between age and change of any of the breathing parameters. In contrast, the proportional reduction of REM AHI was negatively correlated with body mass index (r = 0.77, p < 0.02) suggesting that the more pronounced reduction of sleep-disordered breathing after PHYS was found in leaner patients (Figure 5) .
Samples for determination of plasma concentration were obtained at the end of infusion in four of the patients (BH, UC, BM, and FS) studied. PHYS was nondetectable on the placebo study night, whereas on the infusion night plasma PHYS concentrations was 3.06, 3.11, 3.71, and 3.50 nmol/L, respectively, in the four subjects.
This randomized, placebo-controlled study suggests that the AChE inhibitor PHYS reduces sleep-disordered breathing, most prominently during REM sleep, in patients with moderate to severe OSA. Although the exact mechanism of action remains unclear, the PHYS effect appeared to be independent of the relative severity of the breathing disorder but possibly more pronounced in leaner patients. PHYS, an alkaloid from the West African perennial shrub Physostigma venenosum, is the oldest known AChE inhibitor. The dominant pharmacology of the tertiary amine PHYS is due to a short-acting inhibition of the enzymes AChE and butyrylcholinesterase. This effect of PHYS underlies the clinical use in the treatment of glaucoma and atropine and organophosphate intoxication. PHYS has also, due to its effects on central cholinergic mechanisms, generated considerable attention as a memory- and cognition-enhancing agent in geriatric memory dysfunction and Alzheimer's disease (18, 19). Moreover, a recently demonstrated association between reduced thalamic nerve terminal density and severity of OSA in patients with multiple system atrophy suggests that decreased pontine cholinergic projections may contribute to sleep-disordered breathing at least in certain degenerative conditions (12). The pharmacokinetic and pharmacodynamic properties of PHYS include rapid and complete absorption from the gastrointestinal tract, fairly extensive hepatic clearance, moderate binding to plasma proteins, widespread distribution including the central nervous system, and almost complete metabolic elimination (20, 21). In consideration of the short mean elimination half-life, 21.7 minutes (22), the current study used intravenous administration of the drug. The target plasma concentration was 3 ng/ml based on a previous study by Hartvig and coworkers (16), who reported a total plasma clearance of 40 ml/minute kg after intravenous administration in patients who underwent surgery. This concentration was reached in at least the four patients with plasma samples available for analysis. Assuming an elimination half-life of slightly less than 30 minutes and a constant infusion, it may be expected that plasma concentration steady-state conditions were reached approximately 2 to 2.5 hours after onset of infusion. For this reason data obtained in this study were analyzed after subdividing the study period (infusion period) into three equal parts, each 140 minutes long, whereby the first period would be characterized by a continuous concentration build-up and the two last periods by steady-state conditions. The finding that not only reduction of breathing abnormalities but also effects on heart rate were more pronounced toward the end of the night further supports the proposed pharmacologic effect of PHYS and indirectly suggests a dose-dependency of the effect. It should however be recognized that factors such as natural biological increase in REM sleep toward the end of the sleeping period may have confounded these findings. AChEs are associated primarily with nerve and muscle, typically at synaptic contacts. Inhibition of AChE by PHYS has been shown to elevate brain (23, 24) and peripheral acetylcholine content in several animal species as well as in humans in a study of brain slices from patients with Alzheimer's disease (25). As a substrate competitor of cholinesterase, the dominant pharmacologic effects of PHYS may be anticipated to mimic those of the physiologic cholinergic transmitter, acetylcholine. However, in a pharmacokinetic compartmental model (26) the distribution of PHYS was shown to vary between different tissues as well as between different parts of the brain. This and other factors such as the extent to which a particular cholinergic pathway is active and thus susceptible to the potentiating action of AChE inhibition will influence the pharmacodynamic effects of PHYS. Generation of behavioral and bioelectric REM sleep for example has been associated with cholinergic mechanisms in discrete areas of the pontine tegmentum (27). Not unexpectedly, and in line with previous experiments showing an increased time spent in REM sleep and a shortening of REM sleep latency in humans (11), we found that the proportion of REM sleep was increased after PHYS. The proportionally greater effect of PHYS on sleep-disordered breathing during REM sleep could be explained by the increased cholinergic activity occurring in that sleep state. REM-predominant apnea is a relatively common finding (28), and this observation also raises the possibility that AChE inhibitors may be specifically effective in this group of patients. PHYS was well tolerated in this study, but it was evident that sleep time was dramatically shortened in some patients. AChE inhibitors have been demonstrated to have proportionally small effects on TST in animals and humans (29), but the general analeptic properties of PHYS have been explored in postanesthesia patients who underwent surgery (15). Interestingly, we were able to demonstrate a negative correlation between sleep time and reduction of AHI in the present study, suggesting that effects on sleep on the one hand and breathing on the other may be dissociated. It is also possible that these effects require differential concentration thresholds or that the reduction of AHI (and presumably less fragmented sleep) achieved in some patients masked a potential sleep-time shortening effect of PHYS. Nevertheless, altered sleep amount or loss of Stage 3 + 4 non-REM sleep as noted in this study may limit any potential clinical usefulness of PHYS in patients with OSA. Further studies focusing on the dose–response relationship for these mechanisms are likely to shed further light on this complex association. This small exploratory study did not permit us to apply a multivariate analysis exploring potential factors that possibly may have influenced the proneness to respond with a reduction of AHI after PHYS. The proportion of sleep in the back position was slightly higher on the placebo treatment night, but the change of time in this position was unrelated to all evaluated respiratory indexes. Likewise time spent, or change of time spent, in REM sleep did not correlate with the effect of PHYS on breathing. In a univariate analysis, however, PHYS appeared to be particularly effective in leaner patients. This finding may raise several speculations. First, PHYS may in this context be regarded as a pharmacologic tool to identify the presumably differential key pathophysiologic mechanisms that may prevail in lean and obese patients with OSA. The leaner patients in this study also included the subjects with the most REM sleep–predominant OSA. It is also possible that the more obese patients represented a group with smaller upper airway aperture and higher critical closing pressures (30). Indirectly, this implies that PHYS acted via mechanisms other than stabilization of upper airway muscles during sleep. Acetylcholine transmission is involved in the central regulatory control of respiration in the medulla oblongata (31, 32) as well in chemosensory signaling in the carotid body (33). PHYS administered locally to anesthetized, vagotomized, and artificially ventilated cats resulted in a phase shifting of the hypoglossal-to-phrenic nerve temporal firing properties (34), and it remains possible that alterations in these mechanisms may have accounted for the differences between lean and more obese subjects. Another potential mechanism includes salivary secretion that follows PHYS administration (35). Salivation may have altered upper airway surface tension and thereby increased airway stability as described for compounds with surface tension lowering effects (36). Finally, a weak statistical relationship suggested that subjects with lower resting heart rate responded better to PHYS. Animal experiments employing intrapontine administration of PHYS have demonstrated an increased efferent sympathetic autonomic activity, tachycardia, and a pressor effect by a central mechanism of action (10). Presumably this effect was mirrored by the increase in heart rate seen across the night in the present study. It cannot be excluded that alteration of autonomic activity resulted in autonomic changes that, as recently implied in other studies of sleep apnea using cardiac pacing (37), may have stabilized the respiratory pattern and pharyngeal caliber by interaction with the ventilatory loop gain (7). The present study investigated sleep and breathing during single nights only, and no attempts were made to evaluate the impact of drug treatment on daytime function. Although REM sleep–dominant apnea is a common phenomenon (28), many patients will have sleep apnea across all sleep stages. As time spent in REM sleep is generally less than 25% of TST, elimination of apneas during REM sleep will have a proportionally smaller impact on total AHI across the night. However, a recent clinical study (38) has suggested that significant daytime symptoms were present in patients with OSA with selective REM sleep–related disordered breathing. This would mean that a drug with REM-predominant effects would have specific benefits in OSA. Nevertheless, this remains a controversial point as a larger and cross-sectional study (28) found that daytime sleepiness as assessed by the Multiple Sleep Latency Test was independently and strongly associated with disordered breathing during non-REM sleep. This placebo-controlled exploratory trial has demonstrated a predominantly REM sleep–related apnea alleviating effect of PHYS administered by steady-state infusion. Further studies need to address whether this effect of PHYS also encompasses other compounds of the AChE inhibitor drug class, if dose optimization is possible and if the effect may be maintained using schemes of repeated dosing.
Supported by grants from The Swedish Heart and Lung Foundation and Faculty grants from the University of Göteborg. Conflict of Interest Statement: J.H. is an owner in part of a patent addressing the use of physostigmine in sleep-disordered breathing; H.K. has no declared conflict of interest; Y.P. is an owner in part of a patent addressing the use of physostigmine in sleep-disordered breathing; P.M. is an owner in part of a patent addressing the use of physostigmine in sleep-disordered breathing. Received in original form November 17, 2002; accepted in final form September 1, 2003
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
