A five week, polysomnographic assessment of zaleplon 10 mg for the treatment of primary insomnia

Sleep Medicine 1 (2000) 41±49

Original article

www.elsevier.com/locate/sleep

A ®ve week, polysomnographic assessment of zaleplon 10 mg for the treatment of primary insomnia James K. Walsh a, b,*, Gerald W. Vogel c, Martin Scharf d, Milton Erman e, C. William Erwin f, Paula K. Schweitzer a, Richard M. Mangano g, Thomas Roth h b

a Sleep Medicine and Research Center, St. Luke's Hospital, Chester®eld, MO, USA Department of Psychiatry, St. Louis University School of Medicine, St. Louis, MO, USA c Sleep Research Laboratory, Atlanta, GA, USA d The Center for Research in Sleep Disorders, Cincinnati, OH, USA e Paci®c Sleep Medicine, La Jolla, CA, USA f Department of Psychiatry, Duke University Medical Center, Durham, NC, USA g Wyeth-Ayerst Research, Radnor, PA, USA h Sleep Disorders and Research Center, Henry Ford Hospital, Detroit, MI, USA

Received 13 October 1999; received in revised form 19 November 1999; accepted 19 November 1999

Abstract Objective: To examine the hypnotic ef®cacy of zaleplon 10 mg, a selective benzodiazepine receptor agonist, over a period of 35 nights in primary insomniacs. Methods: A double-blind, parallel-group, placebo-controlled design was employed. Subjects were 113 men and women, ages 18 to 65 years. Polysomnographic and subjective sleep data were collected during baseline, on two nights during each of ®ve treatment weeks, and on the ®rst two nights after discontinuation of active medication. Results: Sleep latency was signi®cantly shortened with zaleplon 10 mg for all 5 weeks of treatment as assessed by polysomnography and by subjective sleep measures. Total sleep time, whether evaluated with polysomnography or with subjective estimates, was inconsistently affected. Sleep architecture was similar with zaleplon and placebo. There was no evidence of tolerance to the sleep promoting effects of zaleplon during the ®ve weeks of administration, and there was no rebound insomnia upon discontinuation. Adverse events occurred with equal frequency in the zaleplon and placebo groups. Conclusions: Zaleplon 10 mg is effective in the treatment of sleep onset insomnia over a period of 35 nights, with minimal evidence of undesired effects. q 2000 Elsevier Science B.V. All rights reserved. Keywords: Zaleplon; Primary insomnia; Polysomnography; Ef®cacy; Tolerance

1. Introduction Zaleplon, a pyrazolopyrimidine hypnotic, binds * Corresponding author. Sleep Medicine and Research Center, St. Luke's Hospital, 232 S. Woods Mill Road, Chester®eld, MO 63017, USA. Tel.: 11-314-205-6030; fax: 11-314-205-6025. E-mail address: [email protected] (J.K. Walsh)

selectively to the benzodiazepine type I site on the type A, gamma-aminobutyric acid (GABA) receptor complex. Pharmacokinetic studies indicate a plasma tmax and an elimination t1/2, both, of about 1 h [1]. This pharmacokinetic pro®le predicts a rapid onset and a relatively short duration of sedative activity. Clinical studies of zaleplon, administered for 2±14 days to either adult [2] or elderly [3±5] insomnia patients,

1389-9457/00/$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 1389-945 7(99)00006-4

42

J.K. Walsh et al. / Sleep Medicine 1 (2000) 41±49

demonstrate sleep-promoting effects which are quite consistent with this prediction. That is, both objective and subjective measures of sleep latency are reduced with zaleplon doses ranging from 2 to 10 mg. Total sleep time is less consistently increased, showing positive results in studies involving older patients [3,5]. The current investigation assessed the ef®cacy of zaleplon during 5 weeks of nightly administration for primary insomnia. We hypothesized that both polysomnographic (PSG) and subjective sleep latency would be signi®cantly shorter with zaleplon 10 mg (Z10) than with placebo (PBO) over the entire study duration. Additionally, PSG and subjective total sleep time were hypothesized to be increased with Z10 as compared to PBO at each time point.

2. Methods 2.1. Experimental design A double-blind, parallel group design was used to compare Z10 with PBO at eleven study sites. Early in the study, zaleplon 20 mg was a third treatment condition; however, the sponsor discontinued randomization to that condition because subject enrollment was slower than expected and because the 20 mg dose has not been shown to be substantially superior to 10 mg (data on ®le, Wyeth-Ayerst Research). Thus, the present paper includes data only from those subjects randomly assigned to PBO or Z10. The study involved a medical screening phase, a 4night PSG screening phase, a 2-night PBO-PSG screening phase (used as baseline), a 35-night treatment phase, a two night single-blind, PBO-discontinuation phase, and a post-study medical assessment. The study protocol, informed consent forms, and all patient recruitment materials were approved by each institution's human research committee. Randomization schedules were produced by the study sponsor and not revealed to any study site or investigator until the data analysis was completed. Beginning on the ®rst night of the PBO-screening phase, and terminating on the second discontinuation night, all patients took two capsules (either two PBO, or one Z10 and one PBO) orally with 200 ml of water

every night. All capsules were identical in appearance. 2.2. Subjects and screening procedures Men and non-pregnant women between ages 18 and 65 years were recruited principally via media advertisements. After telephone screening, all potential participants were scheduled for an initial visit at which time written informed consent was obtained and further screening was conducted. Screening included medical and psychiatric history, physical examination, 12-lead electrocardiogram, and clinical laboratory tests including beta-chorionic gonadotropin, pregnancy testing and a urine drug screen. Zung Anxiety and Depression scales were administered and a thorough sleep and medication history was taken. A sleep diary was completed for a minimum of one week. These screening procedures, performed or reviewed by the study physician, assured that all subjects met the following criteria: (1) a DSM-IV diagnosis of primary insomnia [6] characterized by a typical subjective sleep latency (SSL) .30 min, usual subjective total sleep time (STST) between 4 and 7 h, and insomnia-related daytime complaints (e.g. fatigue, irritability, dif®culty concentrating), (2) a time in bed between 7 and 8.5 h on 4 of 7 nights; bedtime varying by ,2 h during the week, no routine napping, (3) women must not be lactating and, if of child-bearing potential, must use oral, injectable, implant, intrauterine, or barrier contraception, (4) absence of a current medical condition, or current or past major psychiatric illness which may in¯uence the study [5], ,50 on either Zung Anxiety or Depression scales, (5) no illicit drug use or excessive alcohol use or abuse in the past 12 months, (6) urine drug screen negative for any illicit drug, or positive for any psychotropic medication, (7) no current use of cimetidine, rifampin, or an investigational drug (within 30 days), or intake of excessive caffeine or alcohol, (8) no prior participation in a study including zaleplon. Six hundred and seventy seven subjects meeting these initial screening criteria were scheduled for PSG screening, conducted with standard procedures [7], on four consecutive nights. Bedtime and duration of time in bed were held constant on these four nights (and all PSG nights) for each subject. Respiratory and leg electromyograms were conducted on the ®rst PSG

J.K. Walsh et al. / Sleep Medicine 1 (2000) 41±49

night to exclude patients with an apnea/hypopnea index .10/h or a periodic leg movement arousal index .10/h. Subjects having a latency to persistent sleep (LPS) ,15 min on night 1 were also excluded. LPS was de®ned as the time in minutes from the beginning of the recording to the ®rst of 20 consecutive epochs (i.e. 10 min) scored as sleep. PSG entry criteria assessed on nights 2±4 were a LPS .20 min on at least two of three nights (with LPS not ,15 min on either night) and a total sleep time (TST) between 240 and 420 min on the same two nights. Approximately 7 days later, 204 subjects meeting criteria for initial PSG screening, underwent two additional PSGs after taking PBO 30 min before bedtime. This additional single-blind screening was an attempt to reduce the in¯uence of PBO responders in the double-blind, randomized treatment phase of the study. On these PSG screening nights, PSG inclusion criteria were a mean LPS .20 min (not ,15 min on either night), and a TST between 240 and 420 min on both nights. All PSGs were scored initially for entry criteria at the study site and con®rmed by a central scoring laboratory. To reduce inter-scorer variability, all PSGs were scored by a single central laboratory for data analysis purposes. There were a small number of subjects who quali®ed based upon the scoring of the local site, but were found not to meet PSG entry criteria when scored by the central scoring site. These subjects were included in the present analysis. A total of 137 subjects were randomly assigned to treatment group following the PBO-screening nights (PBO ˆ 57, Z10 ˆ 56, Z20 ˆ 24). The 24 subjects assigned to the Z20 group are not included in this report. In the PBO group, 11 of 57 subjects did not complete the study for the following reasons: one reported an adverse reaction, ®ve withdrew for lack of ef®cacy or unspeci®ed request, and ®ve violated the

43

protocol. Of the 56 subjects randomized to Z10, seven did not complete the study: three reported an adverse reaction, and four had protocol violations. Table 1 contains demographic and baseline characteristics for the Z10 and PBO treatment groups. No statistical differences exist between groups. 2.3. Experimental procedures After randomization subjects participated for 37 days, with PSG being performed on the ®rst two nights for each of six consecutive weeks. Doubleblind medication was taken on each of the ®rst 35 nights and single-blind PBO was taken on nights 36 and 37, to assess discontinuation effects. Medication was taken 30 min before bedtime in the laboratory and subjects were instructed to take the medication `at bedtime' on home nights. PSG recordings were 480 min in duration and followed standard procedures. [7] Subjects completed postsleep questionnaires on all laboratory and home study days on which estimates of subjective sleep latency (SSL), subjective total sleep time (STST), and subjective number of awakenings were recorded. Periodic urine drug screens were completed to ensure compliance with the protocol restrictions on substances. All subjects were instructed to maintain a regular sleep schedule on non-laboratory nights. Compliance with the medication regimen was monitored with daily diaries maintained by the subjects and by examination of medication packages, both of which were returned to the study staff each week. Adverse events were recorded by subjects on daily diaries. Study staff inquired about all recorded adverse events, and asked about any problems experienced by subjects since their last visit, upon the subjects' arrival on each laboratory night.

Table 1 Demographic and baseline characteristics of the zaleplon 10 mg and placebo treatment groups Group

Mean age, years (SD)

Male/ female

Mean weight, kg (SD)

Mean Zung anxiety (range)

Mean Zung depression (range)

Mean latency to persistent sleep, min (range)

Placebo, N ˆ 57 Zaleplon 10 mg, N ˆ 56

42.4 (11.1) 41.8 (10.8)

13/44 16/40

74.4 (17.7) 74.8 (18.4)

31 (21±45) 31 (22±47)

35 (24±49) 35 (22±49)

57 (21±184) 56 (12±145)

44

J.K. Walsh et al. / Sleep Medicine 1 (2000) 41±49

2.4. Data analyses Data from this multi-center study were analyzed by the Biostatistics Section of Wyeth-Ayerst. All subjects who received one or more doses of randomized medication are included in the primary statistical analyses. Demographic data were compared between groups by using analysis of variance (ANOVA) for quantitative data, and Fisher's Exact Test for categorical data. Polysomnographic and subjective sleep (from laboratory nights only) data were averaged so that each subject had a single value per variable for each study week (baseline, randomization weeks 1, 2, 3, 4, and 5). Data for discontinuation nights 1 and 2 were examined separately. Data from non-laboratory nights were not included in the present analysis. Data for LPS, the primary ef®cacy variable, were entered into an analysis of covariance (ANCOVA) with study site, treatment group, and site-by-treatment interaction as factors, and the baseline as covariate. Because statistical assumptions underlying the ANCOVA model (e.g. normality and homoscedasticity) were not satis®ed, a non-parametric method of using rank-transformed data (with baseline data transformed separately) were used in the ANCOVA. The assumption of parallelism of treatment baselines was satis®ed so that the baseline covariate could be included in the analysis to control for baseline differences among patients. Z10 and PBO were compared

Fig. 1. Median latency to persistent sleep (LPS) for placebo and zaleplon 10 mg groups during each study week and on the ®rst night after discontinuation of active drug. Zaleplon values are signi®cantly shorter than placebo during weeks 1±5, (P , 0:03 or better).

at each study time point. Data for other sleep variables were analyzed using the same methods. Data from the two discontinuation nights were analyzed in two different ways to determine if there was evidence of rebound insomnia. First, an ANCOVA similar to that described above was used to compare variables between treatment groups on each discontinuation night. Second, a Wilcoxon Signed Rank test was performed separately for Z10 and PBO on the difference between discontinuation night values and baseline values to test whether the median change from baseline was signi®cantly different from zero for either treatment group.

3. Results 3.1. Polysomnographic ef®cacy measures Fig. 1 displays median LPS for both treatment groups during baseline, each week of double-blind treatment, and on the ®rst night following discontinuation of active drug. LPS for Z10 was signi®cantly shorter than for PBO during all ®ve treatment weeks. Median LPS with Z10 was 22±25 min shorter than baseline values during each of the treatment weeks. In comparison, median LPS for the PBO group was 9.5 to 16.0 min shorter than baseline values. Medians, means, and results of ANCOVAs are presented in Table 2. There was no consistent increase in TST with Z10 (see Table 3). TST was signi®cantly greater with Z10 only during week 1, although median and mean values for Z10 during each treatment week were consistently higher than those for PBO. The number of PSG awakenings and the minutes of wake time after sleep onset also showed no consistent difference between treatment groups. The minutes of wake time in the last quarter of the 8-h recording period were not signi®cantly different between the Z10 and PBO groups for any treatment week (see Table 4). The median difference from baseline at each week for wake time in the last quarter ranged from 22.0 to 0.3 min for PBO and from 21.9 to 2.5 min for Z10. Sleep architecture variables were compared between groups at each week. No signi®cant differences in the percentages of stage 1, stage 2, or stages 3/4 were found. During week 2, REM percent was

J.K. Walsh et al. / Sleep Medicine 1 (2000) 41±49

45

Table 2 Median and mean (SD) latency to persistent sleep (min) for each group during baseline and treatment weeks Group

Placebo

Zaleplon 10 mg

F df P

Week

N Median Mean SD N Median Mean SD

Baseline

1

2

3

4

5

57 54.8 57.2 28.9 56 44.3 55.9 31.9

56 35.1 48.2 34.3 55 27.0 32.5 21.2 8.49 1, 96 0.005

49 42.5 49.8 33.6 52 25.0 30.8 25.2 15.89 1, 86 ,0.001

49 36.8 38.2 23.5 52 23.8 28.7 19.1 5.23 1, 86 0.027

47 31.5 36.1 24.9 50 23.6 30.4 30.1 4.11 1, 82 0.031

46 31.1 38.0 23.6 48 23.0 28.5 22.4 6.20 1, 79 0.027

lower with Z10 than with PBO (18.7% vs. 21.7%; P ˆ 0:035) and during week 4, REM latency was signi®cantly longer with Z10, by 13 min, as compared to PBO (P ˆ 0:002). No differences were noted at any other week of treatment. 3.2. Subjective ef®cacy measures Postsleep questionnaire variables included SSL, STST, and subjective number of awakenings. Median SSL with Z10 was consistently lower than with PBO, usually 15±30 min shorter. ANOVAs indicate a signi®cant difference during all treatment weeks (see Table 5).

Median STST was 30.0±38.8 min greater during the ®ve treatment weeks as compared with baseline in the Z10 group. Median STST values were much more variable in the PBO group (see Table 6). Signi®cant increases in STST were found for Z10 during weeks 1 and 3, as compared with PBO. There was no consistent trend in subjective number of awakenings between groups. Z10 subjects reported signi®cantly fewer awakenings than did PBO subjects only during week 3 of treatment. 3.3. Discontinuation effects Data for the two discontinuation nights were

Table 3 Median and mean (SD) total sleep time (min) for each group during baseline and treatment weeks Group

Placebo

Zaleplon 10 mg

F df P

Week

N Median Mean SD N Median Mean SD

Baseline

1

2

3

4

5

57 371.3 368.8 38.4 55 376.8 374.9 38.5

56 384.9 376.4 51.2 54 406.5 398.3 38.4 4.69 1, 96 0.026

49 380.3 381.8 38.7 51 402.0 393.5 53.9 1.94 1, 86 0.116

49 403.3 391.3 47.0 51 411.8 400.2 49.0 1.23 1, 86 0.237

47 397.8 390.1 41.2 49 406.3 403.4 42.2 2.94 1, 82 0.078

45 396.5 392.5 43.3 47 413.5 403.9 43.8 2.35 1, 78 0.155

46

J.K. Walsh et al. / Sleep Medicine 1 (2000) 41±49

Table 4 Median and mean (SD) wake time (min) in the fourth quarter of the night for each group during baseline and treatment weeks Group

Placebo

Zaleplon 10 mg

Week

N Median Mean SD N Median Mean SD

Baseline

1

2

3

4

5

57 12.8 21.4 20.6 55 14.4 21.8 18.8

56 16.8 22.7 20.7 54 21.0 23.7 17.2

49 12.3 18.9 18.8 51 16.5 24.6 24.5

49 11.8 20.9 20.3 51 14.4 21.2 21.9

47 16.5 22.9 19.9 49 14.9 20.4 18.3

45 16.0 22.6 20.0 47 14.3 20.5 19.8

analyzed separately for evidence of rebound insomnia. Comparison between the treatment groups showed no signi®cant differences for the PSG measures of LPS, TST, or number of awakenings, on either discontinuation night. Similarly, subjective variables SSL, STST, and subjective number of awakenings did not differ between groups on either discontinuation night. Table 7 contains median PSG and subjective sleep values for the PBO and Z10 groups at baseline and on the ®rst discontinuation night. Examination of these data shows that all variables for the Z10 group on discontinuation night 1 are in the direction of better sleep relative to baseline values. Comparison of median change-from-baseline values on the discontinuation nights to zero, with the Wilcoxon test, showed

some signi®cant differences, but always in the direction of better sleep on the discontinuation night with zaleplon. 3.4. Adverse events The percentage of subjects reporting an adverse event which was new or which worsened after double-blind medication was begun was similar in the PBO (84%) and Z10 (79%) treatment groups. The most common adverse event was headache for both groups (53% of PBO subjects and 48% of Z10 subjects). Central nervous system adverse events (e.g. dizziness, somnolence) occurred in 22% of PBO and 14% of Z10 subjects. Examination of the frequency of occurrence of adverse events indicated no group

Table 5 Median and mean (SD) subjective sleep latency (min) for each group during baseline and treatment weeks Group

Placebo

Zaleplon 10 mg

F df P

Week

N Median Mean SD N Median Mean SD

Baseline

1

2

3

4

5

57 75.0 80.9 47.6 56 67.5 77.6 57.3

56 60.0 74.2 51.3 55 37.5 45.3 40.6 21.06 1, 96 ,0.001

49 60.0 65.6 40.3 52 30.0 47.1 47.7 15.50 1, 86 ,0.001

49 52.5 64.1 48.1 49 31.3 38.3 28.6 16.73 1, 86 ,0.001

47 45.0 59.3 44.1 47 30.0 41.2 38.5 7.24 1, 82 0.020

46 50.0 57.1 34.8 46 31.3 45.7 42.9 4.93 1, 79 0.036

J.K. Walsh et al. / Sleep Medicine 1 (2000) 41±49

47

Table 6 Median and mean (SD) subjective total sleep time (min) for each group during baseline and treatment weeks Group

Placebo

Zaleplon 10 mg

F df P

Week

N Median Mean SD N Median Mean SD

Baseline

1

2

3

4

5

57 345.0 334.9 71.5 56 352.5 342.7 70.5

56 345.0 334.9 79.5 55 375.0 372.2 64.2 7.58 1, 96 0.011

49 360.0 347.2 74.3 52 382.5 373.0 69.9 3.36 1, 86 0.071

49 360.0 347.9 94.4 52 393.8 383.3 61.9 5.60 1, 86 0.048

47 375.0 356.9 78.7 50 388.8 382.4 64.5 1.19 1, 82 0.378

46 367.5 360.9 66.0 48 390.0 382.4 67.3 2.11 1, 79 0.179

to PBO. These ®ndings are consistent with the shorter term studies of zaleplon [2±5]. This study also extends to ®ve weeks the duration of nightly use for which sleep latency is signi®cantly reduced by Z10 in adult primary insomnia patients. Daily diaries and medication packaging inspection indicated a very high rate of compliance for taking medication. For the 46 Z10 subjects included in the tolerance analysis, 37 took all doses, 6 missed a single dose and 3 missed two doses. The ®nding of no tolerance is strengthened by these data. Undesirable effects of hypnotic medications discussed in the literature include tolerance, rebound insomnia, residual sedation, adverse events, and alterations in sleep architecture [8,9]. The present data indicate no evidence of tolerance to the sleepinducing effects of Z10 during ®ve weeks of nightly use. Additionally, no discontinuation effects sugges-

differences. For example, for those subjects reporting headache, PBO subjects averaged 3.60 occasions of headache and Z10 subjects averaged 3.21 occasions during the 5-week treatment period. Four subjects did not complete the study because of adverse events. One PBO subject withdrew because of substernal pain, dyspepsia, and hemorrhagic gastritis. Three Z10 subjects were withdrawn because of an accidental injury, vomiting, and agitation, respectively.

4. Discussion Z10 signi®cantly decreased both PSG and subjective measures of sleep latency during all ®ve weeks of treatment in this investigation. PSG and subjective variables that relate more to sleep maintenance were inconsistently improved or unaffected by Z10 relative

Table 7 Median values for polysomnography and subjective measures comparing baseline and discontinuation night 1 PSG

Subjective

LPS (min)

TST (min)

No. of wakes

SSL (min)

STST (min)

No. of wakes

PBO-baseline PBO-disc. 1

54.8 25.0

371.3 400.3

19.5 18.5

75.0 30.0

345.0 360.0

3.0 2.5

Z10-baseline Z10-disc. 1

44.3 33.5

376.8 404.3

23.5 21.0

67.5 56.3

352.5 360.0

3.0 3.0

48

J.K. Walsh et al. / Sleep Medicine 1 (2000) 41±49

tive of rebound insomnia were found in this study, nor were there changes in sleep stage percentages or other consistent changes in sleep architecture. Further, these results showed no signi®cant increase in time awake during the last quarter of the night, a phenomenon that has been reported to occur with rapidly eliminated benzodiazepine hypnotic drugs [10]. It should be mentioned that although there was no evidence of rebound insomnia in this study, in the clinical situation patients must be warned that sleep dif®culty may return upon discontinuation of a hypnotic. For example, median LPS increased about 13 min from week 5 to the discontinuation night. Tapering the dose over a night or two may reduce the perceived abruptness of this change in the clinical environment. Residual sedation has not been detected even when assessed only 5 h after administration of Z10 in sleep maintenance insomniacs [11]. In two studies with normal volunteers, no sedation or performance impairment has been documented as little as 2 h after ingestion of Z10, whether administered during the day [12], or at night [13]. Overall, the frequency of adverse events with Z10 has not differed from PBO either in the present study or others [2,3]. Thus, available data suggest that these undesirable effects are uncommon with Z10. The present study is one of few in which nightly hypnotic administration, with a parallel PBO group, has been objectively examined for more than two weeks. Scharf et al. [14] demonstrated continued ef®cacy of zolpidem during ®ve weeks of nightly use. Monti et al. [15] administered zolpidem or triazolam for 27 consecutive nights with a parallel PBO group; however, the primary purpose of the investigation was to examine rebound insomnia and the study was not powered for ef®cacy analysis. Further, the dose of triazolam was twice the recommended dose (0.5 mg). Despite these methodological restrictions, both active drugs showed trends toward persistent hypnotic effects over the 27 nights of administration. Additionally, there are a number of studies, without parallel PBO groups, using either PSG or subjective sleep measures which indicate no signi®cant loss of hypnotic ef®cacy over periods of nightly use ranging from 6 to 24 weeks [16±19]. The current ®ndings with zaleplon indicate further that tolerance to the hypnotic effects of benzodiazepine receptor agonists does not develop within a few weeks of nightly use.

Two laboratory screening periods, the second with PBO administered, were used to reduce the improvement from baseline in the PBO group commonly seen in similar studies. Despite this design strategy, the sleep of the PBO group improved considerably during the ®ve weeks of double-blind treatment. A portion of the change from baseline in the PBO group can be accounted for by regression to the mean, after subjects qualify according to speci®c sleep criteria. Improvement in sleep associated with adherence to protocol restrictions (e.g. regular sleep schedule, limitations on caffeine and alcohol, etc.), spontaneous remission of insomnia, and a true placebo effect are other factors likely to contribute to this phenomenon. In conclusion zaleplon, a selective benzodiazepine receptor agonist, at a dose of 10 mg, was found to be effective in the treatment of sleep onset insomnia over a period of 35 nights, with no evidence of tolerance or discontinuation effects. Clearly, zaleplon has consistent sleep inducing properties which can be very useful in a number of clinical situations. The degree to which zaleplon's limited affect on sleep maintenance measures in research studies will affect its use in clinical medicine is yet to be determined. However, it is possible that some patients will require a longer-acting hypnotic to improve ability to maintain sleep. Acknowledgements The authors thank Gail Koshorek of Henry Ford Hospital, Detroit, and Mary Kay McPherson, Noreen Scherer, Timothy Whitaker, MD, and Carmen Wickland, Ph.D. of Wyeth-Ayerst Research, Radnor, PA, for their contributions to the study. This research was supported by Wyeth-Ayerst Research. References [1] Beer B, Clody DE, Mangano R, et al. A review of the preclinical development of zaleplon, a novel non-benzodiazepine hypnotic for the treatment of insomnia. CNS Drug Rev 1997;3:207±224. [2] Walsh JK, Fry J, Erwin CW, Scharf M, et al. Ef®cacy and safety of 14-day administration of zaleplon 5 mg and 10 mg for the treatment of primary insomnia. Clin Drug Invest 1998;16:347±354. [3] Ancoli-Israel S, Walsh JK, Mangano RM, Fujimori M. Zale-

J.K. Walsh et al. / Sleep Medicine 1 (2000) 41±49

[4] [5] [6] [7]

[8]

[9] [10] [11]

plon, a novel non-benzodiazepine hypnotic, effectively treats insomnia in elderly patients without causing rebound effects. Prim Care Companion J Clin Psychiat 1999;1:114±120. Hedner J, Emilien G, Salinas E. Improvement in sleep latency and sleep quality with zaleplon in elderly patients with primary insomnia. J Sleep Res 1998;7(Suppl 2):115. Erwin CW. Zaleplon effectively and safely treats chronic insomnia in elderly patients. J Am Geriatrics Soc 1999;47(9):S87. American Psychiatric Association. Diagnostic and statisical manual of mental disorders, 4th ed. Washington, DC: American Psychiatric Association, 1994. Rechtschaffen A, Kales A. A manual of standardized terminology, techniques and scoring systems for sleep stages of human subjects, Los Angeles, CA: Brain Information Service/Brain Research Institute, 1968. Mendelson WB. Medications in the treatment of sleep disorders. In: Meltzer HY, editor. Psychopharmacology: the third generation of progress, New York: Raven Press, 1987. pp. 1305±1311. Roth T, Roehrs T. Issues in the use of benzodiazepine therapy. J Clin Psychiat 1992;53:S14±S18. Kales A, Soldatos CR, Bixler EO, Kales JD. Early morning insomnia with rapidly eliminated benzodiazepines. Science 1983;220:95±97. Walsh JK, Pollak CP, Scharf MB, Schweitzer PK, et al. Lack of residual sedation following middle-of-the-night zaleplon

[12]

[13] [14] [15]

[16] [17] [18] [19]

49

administration in sleep maintenance insomnia. Clin Neuropharm 2000; in press. Danjou P, Paty I, Fruncillo R, Worthington P, et al. A comparison of the residual effects of zaleplon and zolpidem following administration ®ve to two hours before awakening. Br J Clin Phamacol 1999;48:367±374. Greenblatt DJ, Harmatz JS, von Moltke LL, Ehrenberg BL, et al. Comparative kinetics and dynamics of zaleplon, zolpidem and placebo. Clin Pharmacol Ther 1998;64:553561. Scharf MB, Roth T, Vogel GW, Walsh JK. A multicenter, placebo-controlled study evaluating zolpidem in the treatment of chronic insomnia. J Clin Psychiat 1994;55:192±199. Monti JM, Attali P, Monti D, Zipfel A, et al. Zolpidem and rebound insomnia±a double-blind, controlled polysomnographic study in chronic insomnia patients. Pharmacopsychiatry 1994;27:166±175. Oswald I, French C, Adam K, Gilham J. Benzodiazepine hypnotics remain effective for 24 weeks. Br Med J 1982;284:860±863. Leibowitz M, Sunshine A. Long-term hypnotic ef®cacy and safety of triazolam and ¯urazepam. J Clin Pharmacol 1978;18:302±309. Pierce MW, Shu VS. Ef®cacy of estazolam: the United States clinical experience. Am J Med 1990;88(Suppl 3A):65±115. Kummer J, Guendel L, Linden J, Eich FX, et al. Long-term polysomnographic study of the ef®cacy and safety of zolpidem in elderly psychiatric in-patients with insomnia. J Int Med Res 1993;21:171±184.