Effect of zopiclone on sleep, night-time ventilation, and daytime
ABSTRACT We have assessed the effects of zopiclone (7.5 mg), a new cyclopyrrolone hypnotic drug, on ventilation, sleep parameters, and daytime vigilance in.
Printed in UK - ERS Journals Ltd 1997 European Respiratory Journal ISSN 0903 - 1936 Effect of zopiclone on sleep, night-time ventilation, and daytime vigilance in upper airway resistance syndrome F. Hypnotics, especially long-acting benzodiazepines, are known to blunt defence mechanisms such as arous- ability and to depress the respiratory centres and upper airway dilator muscles. Zopiclone is a short-acting cyclopyrrolone that ap- pears to differ clinically from benzodiazepines in sev- eral ways: it preserves slow wave sleep (SWS), it does not impair daytime performance in normal sub- jects, it has no significant effect on the control of breathing during quiet sleep, quiet wakefulness, or induced hyperventilation in patients with pulmonary disease, and it has only minimal deleterious effects on respiration in patients with OSAS. It has also been suggested that benzodiazepines may be asso- ciated with increased mortality in the elderly. ABSTRACT: We have assessed the effects of zopiclone (7.5 mg), a new cyclopy- rrolone hypnotic drug, on ventilation, sleep parameters, and daytime vigilance in snorers with upper airway resistance syndrome (UARS). The prevalence of this new syndrome is proba- bly high in a population of heavy snorers, and sleep fragmentation has been found in more than half the non- apnoeic subjects referred to an otorhinolaryngology clinic for snoring with social impairment. Polysomnography followed by a multiple sleep latency test (MSLT) was performed during the last night of each treatment period. Page 2. Cau- tion should therefore be observed when prescribing hypnotic drugs to snorers with a complaint of insomnia. reported the case of an insomniac heavy snorer who developed OSAS after receiving flurazepam 30 mg. Correspondence: F. Lofaso, F. In conclusion, zopiclone has no adverse effects on sleep architecture, respirato- ry parameters during sleep, and daytime sleepiness in patients with UARS. For example, MENDELSON et al. Eur Respir J 1997; 10: 2573–2577 DOI: 10.1183/. Goldenberg*, C. Because insomnia is also a common problem in heavy snorers, many patients with UARS probably receive hyp- notic drugs as the first approach to their condition. Any of these effects could precipitate the transition from UARS to OSAS. Thebault, C. ERS Journals Ltd 1997. Because it is important to document the safety of this widely used hypnotic in UARS, a study was designed to evaluate the effects of 7.5 mg·day-1zopiclone during 1 week, versus a placebo, on ventilation, sleep parame- ters and daytime vigilance in snorers with UARS. **Laboratoires Rhône-Poulenc Rorer-Spécia, 92545 Montrouge, France. *Service de Physiologie-Explorations Fon- ctionnelles and Institut National de la Santé et de la Recherche Médicale INSERM U 296, Hôpital Henri Mondor, 94010 Cré- teil, France. Using a randomized double-blind design, eight male patients with UARS took either oral zopiclone or a placebo each evening for seven consecutive days and then crossed over to the other drug after a 7 day placebo period. Janus, A. Clinical experience and epidemiological studies indicate that OSAS is a progressive disease in which simple snoring without apnoeas gradually evolves into a disabling and even fatal condition. Similarly, DOLLY and BLOCK observed that flurazepam significantly modified the incidence and severity of breathing abnor- malities during sleep in asymptomatic patients. Janus**, A. Obstructive sleep apnoea syndrome (OSAS) is belie- ved to cause poor nocturnal sleep in snorers. Goldenberg, C. It had no effect on sleep architecture or on the arousal index (placebo 17±8 arousals·h-1versus zopi- clone 17±4 arousals·h-1). Zopiclone produced significant improvements in the sleep efficiency index (total sleep time/time in bed) (placebo 84±15% versus zopiclone 91±7%) and average MSLT (placebo 10.3±3.7 min versus zopiclone 14.9±2.8 min), as well as nonsignif- icant improvements in sleep onset latency and total sleep time. In this condition the arousal stimulus is the increased ventilatory effort in response to airway obstruction. Eur Respir J 1997; 10: 2573–2577. Lofaso, Service de Phy- siologie-Explorations Fonctionnelles, Hôpi- tal Henri Mondor, 94010 Créteil, France Keywords: Arousal hypnotic respiration during sleep sleepiness snoring upper airway resistance syndrome Received: March 20 1997 Accepted after revision August 24 1997 This study was supported by the Laboratories Rhône-Poulenc Rorer-Spécia, 92545 Mon- trouge, France. Lofaso*, F. Furthermore, none of the respiratory parameters were significantly affected by zopiclone. 97. Thebault**, C. Harf. The sleep fragmentation of UARS can cause daytime sleepiness and difficulty in maintaining sleep. It is therefore important to ascertain that these drugs are associated with minimal or no effects on respiration in the population. More recently, it has been de- monstrated that habitual snoring without apnoea, hy- popnoea, or hypoxic dips can be accompanied by subtle disturbances in nocturnal sleep and sleep fragmentation, a condition called upper airway resistance syndrome (UARS). Harf* Effect of zopiclone on sleep, night-time ventilation, and daytime vigilance in upper air- way resistance syndrome, F.
ABSTRACT We have assessed the effects of zopiclone (7.5 mg), a new cyclopyrrolone hypnotic drug, on ventilation, sleep parameters, and daytime vigilance in snorers with upper airway resistance syndrome (UARS). Page 1. In conclusion, zopiclone has no adverse effects on sleep architecture, respiratory parameters during sleep, and daytime sleepiness in patients with UARS. Zopiclone produced significant improvements in the sleep efficiency index (total sleep time/time in bed) (placebo 84+/-15% versus zopiclone 91+/-7%) and average MSLT (placebo 10.3+/-3.7 min versus zopiclone 14.9+/-2.8 min), as well as nonsignificant improvements in sleep onset latency and total sleep time. Polysomnography followed by a multiple sleep latency test (MSLT) was performed during the last night of each treatment period. Using a randomized double-blind design, eight male patients with UARS took either oral zopiclone or a placebo each evening for seven consecutive days and then crossed over to the other drug after a 7 day placebo period. Furthermore, none of the respiratory parameters were significantly affected by zopiclone. It had no effect on sleep architecture or on the arousal index (placebo 17+/-8 arousals x h(-1) versus zopiclone 17+/-4 arousals x h).
Statistical analysis Comparisons between the two treatments were made using paired Wilcoxon's tests. 90 ZopiclonePlacebo 5 0 15 10 Apnoea index 100 98 96 94 92 Mean Sa,O2 % Zopiclone Placebo 15 10 5 0 95 90 85 80 75 70 Apnoea-hypopnoea index 100 Minimal Sa,O2 % Page 4. the Epworth sleepiness scale (ESS) ; and 2) a sub- jective assessment of sleep by a French version of the Spiegel questionnaire. REM: rapid eye movement. During domiciliary polysomnog- raphy, their AI and AHI were 1±1 and 3±3 events·h-1, respectively, and their arousal index was 17±7 events·h-1 during sleep. 1), although one patient showed a 3% decrease with zopiclone. Figure 1 shows that individual respiratory variables were similar with zopiclone and the placebo. Mean and SEM values are shown by the open symbols and bars. 1. analysis 9.9±6.5 15.8±5.1 12.6±4.6 16.8±4.9 17.9±3.4 14.9±2.8 NS p=0.01 NS p<0.05 NS p<0.01 Fig. 1). During the laboratory polysomnography performed before study inclusion, the mean oesophageal pressure swing at the end of the nonapnoeic-nonhypop- noeic respiratory events was 27± 5 cmH2O. The AHI did not exceed 5 events·h-1during sleep on either of the two treatment nights, except in one patient, whose AHI was lower with zopiclone than with the placebo. The arousal index was similar with zopiclone and the placebo (table 1). – Multiple sleep latency test data Time of the Placebo Zopiclone Statistical latency test 09:00 h6.4±3.7 11:00 h6.2±5.1 12:00 h12.7±5.0 15:00 h10.6±6.9 17:00 h15.5±6.7 Mean 10.3±3.7 Values are presented as mean±SD in minutes. NS NS NS p<0.05 NS NS NS NS NS NS NS Table 2. – Sleep data Placebo Zopiclone Statistical analysis 34±67 11±11 81±55 97±33 399±75 421±39 84±15 91±7 37±24 34±24 12±6 12±4 53±6 51±9 7±2 7±3 12±4 12±9 16±10 17±4 17±8 17±10 Sleep-onset latency min REM sleep latency min Total sleep time min Total sleep time/time in bed % Wake after sleep onset min Stage I % of total sleep time Stage II % of total sleep time Stage III % of total sleep time Stage IV % of total sleep time REM % of total sleep time Arousal index event·h-1during sleep Values are presented as mean±SD. No differences were found in other indices of sleep architecture. Their mean body mass index (BMI) was 26.3±2.4 kg·m-2. Table 1 shows the sleep architecture data for the two treatment nights. In addi- tion, the mean total duration of apnoeas and hypopnoeas during the night was similar with zopiclone and place- bo (6±10 min versus 7±10 min, respectively). No patient reported a worsening of snoring with zopi- clone or the placebo. In this patient, minimal Sa,O2was higher during the zopiclone ZOPICLONE AND UPPER AIRWAY RESISTANCE SYNDROME 2575 Table 1. There were no differences seen between the two treatments regarding mean Sa,O2(fig. There were no statistically significant differences be- tween zopiclone and the placebo regarding mean scores on the ESS (13±5 versus 12±6, respectively) and the Spiegel questionnaire (20±4 versus 20±4, respectively). – Apnoea index, apnoea-hypopnoea index, mean Sa,O2and minimal Sa,O2in the eight study patients during the two polysomnog- raphy nights (placebo versus zoplicone). The index of respiratory events of (apnoea and hypopnoea included) was 21±4 events·h-1, of which 15±3 were terminated by an arousal according to the American Sleep Disorders Association (ASDA) criteria. As compared with the placebo, zopi- clone was associated with a significant improvement in the sleep efficiency index (total sleep time/time in bed) and with nonsignificant improvements in sleep onset latency and total sleep time. Results Eight patients aged 34–56 yrs (mean±SD 46±9 yrs) took part in the study. Minimal Sa,O2showed no differences between treatments, and zopiclone had no important individual effect on this index (fig. The level of significance was set at 5%. Table 2 shows that mean latency measured during MSLT was significantly increased by zopiclone as com- pared to the placebo.
A nonapnoeic-nonhypopnoeic respiratory event was defined as the concomitant occurrence, in the absence of hypopnoea (airflow drop <50% of the preceding baseline during 10 s and desaturation < 3% of the prece- ding baseline level) of both a gradual change in the shape of the inspiratory flow contour characterized by increasing limitation and an increase in the oesophageal pressure swing, with termination of the event as abrupt normalization of both the inspiratory flow contour and the oesophageal pressure swing. The total duration of the study was 3 weeks, with two 1 week periods during which either zopiclone or a placebo was given orally in the evening, and an intervening 1 week placebo period. Nocturnal polysomnography started at 23:00 h and stopped at 06:45 h. Patients were also required to have a habitual consumption of no more than four caffeine- containing beverages per day and to have no history of alcohol abuse. All sleep recordings were reviewed by the same independent examiner. Polysomnography was performed in the standard way, using a 14 channel paper recorder (Electroencephalograph; Nihon Kohden), and included electroencephalography (C4-A1, C3-A2), electro-oculography, chin electromyogra- phy, electromyography of the tibialis anterior of both legs, oronasal thermistors, thoracic and abnormal move- ments and arterial pulse oximetry (Nellcor BS; Nellcor Inc). In addition, respiratory effort was monitored by measuring oesophageal pressure (Gaeltec; Dunvegan, Skye, UK). Diagnosis of UARS Two polysomnographic evaluations were required for the diagnosis of UARS. Thus, intrusion of alpha activity of less than 3 s, alpha activity occur- ring less than 10 s after an arousal, and alpha activity during REM sleep, not accompanied by a concurrent increase in submental EMG amplitude do not meet the criteria of arousal. The domiciliary study included recordings of electro- encephalogram (C4-A1, C3-A2), electro-oculogram, chin electromyogram, electromyogram of the tibialis anteri- or muscle of both legs, oronasal airflow and rib cage movement (Multi-Parameter Analysis recorder 2 and Medilog 9200; Oxford Medical Instrument, Abingdon, UK), and arterial pulse oximetry (Nellcor BS; Nellcor Inc, Hayward, USA). Electroencephalogra- phic (EEG) arousals were detected based on an abrupt shift in EEG frequency, including alpha and/or frequ- encies greater than 16 Hz but not spindles, and were scor- ed according to standard criteria. The laboratory polysomnographic study included re- cordings of an electroencephalogram (C4-A1, C3-A2), electro-oculogram, a chin electromyogram, thoracic and abdominal movement, electromyogram of the tibialis anterior of both legs and arterial pulse oximetry (Nellcor BS; Nellcor Inc). A second polysomnographic study was then per- formed in the laboratory to demonstrate that the sleep fragmentation was due to an increase in respiratory ef- fort in response to an increase in upper airway resis- tance. An abnormal breathing event during objectively measured sleep was defined according to commonly used clinical criteria as either complete ces- sation of airflow lasting 10 s or more (apnoea) or a reduction in oronasal airflow lasting 10 s or more with a drop in arterial oxygen saturation (Sa,O2) of at least 3% of baseline (hypopnoea). The apnoea index (AI) and hypopnoea+apnoea index (AHI), expressed as events· h-1during sleep, were used as summary measurements of sleep-disordered breathing. During the laboratory study night, oronasal airflow was quantified using a tight-fitting facial mask and a Fleisch No. During the last night of each treatment period, the study patients underwent all night polysomnography follow- ed by a multiple sleep latency test (MSLT). 2574 Page 3. Zopiclone versus placebo study The study was approved by the Research Ethics Com- mittee of our institution, and each patient gave consent in accordance with the requirements of the Committee. 2 pneumotachograph (Fleisch, Lausanne, Switzerland) connected to a differential pres- sure transducer (Validyne MP45 ±5 cmH2O; Northridge, CA, USA). The study was performed using a randomized, dou- ble-blind, cross-over design. The MSLT, a standardized method used to assess day- time sleepiness, comprises five consecutive tests conducted at 2 h intervals starting at 09:00 h, and is designed to evaluate the propensity to fall asleep. In addition, at the end of the MSLT, each patient com- pleted a questionnaire including: 1) a French version of F. The criterion for diagnosis of UARS and eligibility for the study was the presence of more than 10 nonapnoeic-nonhypopnoeic respiratory events·h-1 during sleep terminated by an aro- usal. Medication (zopiclone or placebo) was administered immediay before lights-out. LOFASO ET AL. Patients were excluded if physical examination, labora- tory tests (serum creatinine and hepatic enzymes) elec- trocardiograph (ECG), vital capacity, or forced expiratory volume in one second (FEV1)/forced vital capacity (FVC) were abnormal. Sleep staging was performed according to standard criteria. To be eligible for the laboratory study, patients had to have a domiciliary study that met the following criteria: apnoea index <5 events·h-1dur- ing sleep; apnoea-hypopnoea index <10 events·h-1dur- ing sleep; oxygen desaturation index <10 events·h-1 during sleep; and arousal index greater than 10 events·h-1dur- ing sleep (see below for the definitions of apnoea, hypopnoea, oxygen desaturation and arousal). All signals were recorded on a 14 channel paper recorder (Electroencephalograph; Nihon Kohden, Tokyo, Japan) for subsequent analysis. A home polysomnograph- ic study was performed first to identify snorers with abnormal sleep fragmentation but no obstructive sleep apnoea (OSA) or periodic leg movements to explain this finding. Methods Patients All included patients were subjects with UARS taken from a group of heavy snorers who complained of day- time tiredness and/or sleepiness. Beverages containing alcohol or caffeine were prohibited during the days of study. Subjects with a current medical illness, or a history of serious psychiatric disease or who were taking medication known to affect sleep or vigilance were excluded.
In other words, despite the fact that the patients were asked to consider only the preceding 7 days, since some of the situations described by the Epworth test were not encountered within these last 7 days it is possible that the patients were simply basing the answers on their experience over a longer period of time. Al- though the results from this patient may have biased our data somewhat, he was similar to the seven others in that neither his AHI nor his total duration of apnoeas and hypopnoeas increased with zopiclone. However, these methods have several limitations. In that study, slow wave sleep values were nor- mal in middle-aged nonapnoeic snorers with an arousal index of less than 10 events·h-1(mean 7±2 events·h-1), but were abnormal in those with an arousal index of more than 10 events·h-1 (mean 15±5 events·h-1). In addition, a subjective rating scale that uses a questionnaire based on the experience of specific conditions may not be well suited to the evaluation of a change in sleepiness dur- ing a period of only 7 days. With the placebo, the mean ESS score was above normal ranges and the mean MSLT score was similar to that observed in the original study on UARS, i.e., slightly above the pathological range. The ESS is limited by individual differences in the de- scription of subjective sleepiness. For instance, subjects have been observed to fall asleep during periods when they rated themselves as fully alert. However, no patients reported a worsening of snoring with zopiclone or placebo. However, he was the only patient whose mean Sa,O2decreased by more than 1% during the zopiclone night as compared to the placebo. F. The beneficial effect of zopiclone on daytime alertness documented by the MSLT may be a specific effect of zopiclone. LOFASO ET AL. Lending support to this hypothesis is the absence of any increase in the number or duration of apnoeas or hypopnoeas with zopiclone, with most disordered breathing events dur- ing sleep being terminated by arousal. Although zopiclone did not decrease sleep frag- mentation and had very few beneficial effects on sleep parameters, it was associated with a substantial improve- ment in MSLT. Sleep efficiency increased significantly with zopiclone and there was no decrease in SWS. These findings show that when we used the ASDA criteria in our laboratory, arousals occurring at a rate of more than 10 events·h-1 significantly affect the archi- tecture of nocturnal sleep. This reduction was probably due to a reduction in baseline sleep Sa,O2since the AHI and the duration of apnoea and hypopnoeas remained unchanged with zopiclone. However, it does not involve simulating the environment in which the subject may usually fall asleep. These results are consistent with studies in chronic insomniacs and normal sleepers demonstrating either an increase or no change in SWS. We found that the effects of zopi- clone in UARS patients were similar to those previously reported in normal subjects, in that no increase in daytime sleepiness was seen from 09:00 h to 17:00 h. This suggests that the sedation provided by 7.5 mg of zopiclone is not suffi- cient to increase the arousal threshold. All the present patients complained of tiredness and/ or sleepiness. Interestingly, no decrease in the arousal index was seen with zopiclone. We attempted to exclude subjects with obstructive sleep apnoea by preliminary domiciliary polysomnog- raphy. Before discussing the implications of our findings, we will address several methodological issues. 2576 Page 5. These important dif- ferences between the ESS and MSLT may explain why the results provided by these two methods were not cor- related in our study. We quantified sleepiness using both the ESS, a subjective rating scale, and the MSLT, the only electrophysiological test that has been validated as a tool for distinguishing different degrees of sleepiness. Although the ASDA has developed criteria for defining arousal, there is still room for subjectivity. The def- inition of sleep fragmentation as an arousal index of more than 10 events·h-1during sleep may seem arbi- trary. We did not make objective measurements of snoring and we are, therefore, unable to exclude an increase in snoring with zopiclone. In contrast to the ESS, MSLT is an objective measurement in which the functional conse- quences of a treatment are evaluated at 2 h intervals during the waking portion of the day. However, one of our patients met the criteria for mild SAS during the placebo study. Our selection of 10 arousals·h-1during sleep as the thresh- old defining sleep fragmentation was based on a recent study that we performed in 105 nonapnoeic heavy snor- ers. Discussion The main results of this study are that zopiclone cau- sed no deterioration in sleep architecture, respiratory mea- surements or daytime vigilance in patients with UARS, as compared with a placebo. The most likely explanation for this reduction is a fall in resting ventilation. Improvements in daytime alertness with zopiclone have also been observed in athletes without sleep abnormalities, although this study did not use objective measurement of daytime alertness. Thus, normal subjects and normal snorers had less than 10 short EEG arousals events·h-1 during sleep in some stud- ies, whereas in another study control subjects without OSA had an arousal index of 21 events·h-1. In addi- tion, in another recent study, we observed that suc- cessfully treated sleep apnoea syndrome (SAS) patients had less than 10 short EEG arousals·h-1during sleep. Another study in normal subjects found that 7.5 mg of zopiclone was not associated with any improve- ment in MSLT. This is in agreement with a previous study that found no worsening of mod- erate OSAS with zopiclone. This finding may be ascribable to the short half-life of zopiclone and its active metabolite (range 3.5–6 h) res- ponsible for an absence of any residual effects after 8 h for doses lower than 10 mg. The AHI has been shown to vary substantially from night to night in pati- ents with mild sleep-disordered breathing. On the other hand, it has been demonstrated that zopiclone has no significant effects on control of breathing in patients with severe chronic obstructive pulmonary disease. Nevertheless, in our seven other patients with UARS, none of the respiratory parameters were substantially affected by zopiclone (figure 1). night than during the placebo night (84% versus 72%, respectively), whereas no differences were seen for AHI (12 events·h-1versus 15 events·h-1, respectively) or total duration of apnoeas and hypopnoeas (28 min versus 31 min, respectively).