Sleep laboratory studies of flurazepam: A model for evaluating hypnotic drugs The results from six separate evaluations of jlurazepam 30 mg in the sleep laboratory were combined to determine the effectiveness of the drug in inducing and maintaining sleep and its effects on sleep stages in a large sample of insomniac subjects. The combined studies provide a model from which a detailed profile of the effects of a hypnotic drug over short-, intermediate-, and long-term conditions can be thoroughly evaluated. Although sleep was significantly improved on the first night of jlurazepam administration, peak effectiveness of the drug did not result until the second and third consecutive drug nights. Flurazepam continued to be effective in inducing and maintaining sleep with intermediate- and long-term drug use with only a slight loss of effectiveness with long-term use. Sleep was also significantly improved on the first and second nights of drug withdrawal. Carryover effectiveness of active metabolites of jlurazepam from one drug night to the next drug night and to withdrawal nights is discussed. The clinical implications are discussed with regard to the time of peak effectiveness of the drug, dosage recommendations and schedule, minimizing possible effects of the drug on daytime performance, and the rationale and method for using drug holidays in the treatment regimen. With this comprehensive profile of the drug's actions, the physician is able to more rationally and effectively utilize the drug in treating the insomniac patient. With short-term administration, jlurazepam produced a slight decrease in rapid eye movement (REM) sleep and an increase in REM latency. These effects were much more pronounced with intermediate-term drug administration, again possibly due to the accumulation of active metabolites. After withdrawal there was no rebound in REM sleep. Stages 3 and 4 sleep decreased progressively through short and intermediate drug administration. With initial withdrawal, there was a slight recovery in both sleep stages. Anthony Kales, M.D., Edward O. Bixler, Ph.D., Martin Scharf, B.A., and Joyce D. Kales, M.D. Hershey, Pa. Sleep Research and Treatment Center and Departments of Psychiatry and Pharmacology,
Pennsylvania State University, Milton S. Hershey Medical Center
Supported in part by Grant No. NGR39-009 and Contract No. 9-10835 from the National Aeronautics and Space Administration (NASA) and by grants from Hoffmann-La Roche Inc., Nutley, N. J. Received for publication Nov. 13, 1975. Accepted for publication Feb. 9, 1976. Reprint requests to: Or. Anthony Kales, Sleep Research and Treatment Center, Departments of Psychiatry and Pharmacology, Milton S. Hershey Medical Center, Hershey, Pa. 17m3.
576
Over the last several years, sleep laboratory studies have been increasingly used in the evaluation of the effectiveness of hypnotic drugs. The primary advantages of these studies are that they permit precise and objective measurements of effectiveness parameters
Flurazepam
Volume 19 Number 5. Part 1
577
Table I. Protocol for evaluating short-, intermediate-, and long-term hypnotic drug effectiveness Night
P
D
x 2 to 4 5 to 7
Lab
Home
X
X X
8 to 15
X
16 to 18
X
19 to 29
X
30 to 32
X
X X X X X X
throughout the night. Their major limitation, the small number of subjects who can be evaluated, is due to the cost of monitoring subjects throughout the night and the limited number of subjects who can be accommodated at one time in the sleep laboratory. 2. 6 In the last eight years, we have extensively evaluated fiurazepam (Dalmane 30 mg) in the sleep laboratory with several different protocols. 2 • 6 Initially, the efficacy of the drug was evaluated over periods of short-term drug administration (3 days).lo The results were positive, and the efficacy of the drug was then evaluated for an intermediate term (2 wk). 1 Since in clinical practice hypnotic drugs are frequently prescribed for long periods ,4 we evaluated the efficacy of the drug over a 4-wk period of administration. 7 These periods are substantially longer than those in the usual trials of hypnotic drugs, which generally include only I to 3 consecutive nights. In our study, by combining the data from 6 of our sleep laboratory evaluations of fiurazepam,l, 7, 8, 10, 11 we were able to determine the effectiveness of the drug in a large sample of insominac sUbjects. Methods
Subject selection. Twenty-three insomniac patients were selected from the Sleep Disorders Clinic of our Sleep Research and Treatment Center. The patients selected were not psychotic. Each patient had a history of difficulty either in falling asleep, staying asleep, or both.
Condition Adaptation to sleep lab environment Baseline measurements Initial and short-term drug effectiveness Evaluation in home surroundings Intermediate-term (2 wk) drug effectiveness Evaluation in home surroundings Long-term (4 wk) drug effectiveness
To qualify for the study, patients had to have a reported history of an average of longer than 45 min to fall asleep, or less than 6V2 hr of total sleep time. Design of studies. The short-term protocols (N = 12) included only a period of short-term drug administration. Automatically, the intermediate-term protocols (N = 6) included both short- and intermediate-term drug adrrumstration, and the long-term protocol (N = 5) included short-, intermediate-, and long-term drug administration. Since all of the protocols included a short-term period of drug administration, there were 23 short-term evaluations and in the same way there were 11 intermediate-term evaluations. The basic design of the different studies is described in Table I. In all three designs an initial placebo period was followed by drug and a subsequent placebo period. The initial series of placebo nights (1 to 4) allowed for adaptation to the sleep laboratory and baseline measurements. The first series of drug nights (5 to 7) permitted the monitoring of short-term drug effects. In the short-term protocol, the subjects were withdrawn from the drug by substituting placebo after the first series of drug nights. In the intermediate- and long-term protocols, the patients were continued on drug while at home (8 to 15), and then the intermediate-term drug effects were evaluated in the laboratory (16 to 18). For the subjects on the intermediate-term protocol, placebo was substituted for drug, while those on the long-term protocol were
578
Kales et al.
Clinical Pharmacology and Therapeutics
Table 11. Short-term drug administration: Effectiveness for sleep induction and maintenance and effects on sleep stages (N = 23) Short-term drug administration Parameters measured Sleep latency Wake time after sleep onset Total wake time No. of wakes Percent sleep stages REM % and min I
2 3 4
Baseline 2-4
5
I
6
I
7
I
5-7
45.4 32.4
37.2 16.2*
21.7* 8.9*
21.7* 13.5*
26.8* 12.9*
77.8 10.6
53.4* 8.0*
30.6* 6.0*
35.2* 6.9*
39.7* 6.9*
23.8(99.3) 5.0 58.6 8.5 4.2
21.9(97.7) 3.4* 63.8* 7.2 3.5
21.3t(99.9) 2.6* 65.5* 8.2 2.4
22.1(102.5) 3.3* 66.9* 6.3 1.4
21.8t(lOO.0) 3.1 * 65.4* 7.2 2.5
*p < 0.01. tp < 0.05.
continued on drug at home (19 to 29). Longterm drug effects were then evaluated in the laboratory (30 to 32) and placebo was again substituted for drug. In the short-term protocols, the placebo withdrawal period extended for 2 or 3 nights. In the intermediate- and long-term protocols, there were 4 initial placebo withdrawal nights. In the long-term protocol there was an extended placebo withdrawal period and the subjects were also monitored for 2 wk after drug withdrawal. Throughout the study, the subjects were instructed not to nap, not to significantly alter their level of daily physical activity, and not to use any drugs. On each night, subjects were administered single gelatin capsules. These capsules contained either flurazepam 30 mg or a matching placebo. On the laboratory nights, the subjects were continuously monitored for 8 hr with electroencephalogram (EEG), electromyogram (EMG) , and electrooculogram (EOG). These all-night sleep recordings were scored according to standardized criteria independent of any knowledge of the experimental conditions. 13 Parameters evaluated in the study included sleep induction (sleep latency), sleep maintenance (wake time after sleep onset and number of awakenings), sleep stages including stages REM, 1,2,3, and 4, number of REM periods, and interval from sleep onset to the first REM period (REM latency).
In comparing the data for each of the three treatment and two withdrawal conditions (short-, intermediate-, and long-term drug administration and short-term and extended withdrawal) to baseline, the baseline for each comparison consisted of data from only those patients who were included in the treatment condition being contrasted. Each of the comparisons to baseline were evaluated using a correlated t test. In analyzing the short-term withdrawal effects of flurazepam, the data were obtained from those studies that included at least three consecutive nights of placebo withdrawal. The withdrawal data from two studies was not included since each of these studies used only two consecutive withdrawal nights. The withdrawal data that were then analyzed were derived from four different studies (N = 4 in each study). In two of these studies, the withdrawal period was preceded by five consecutive nights of drug administration, while in the other two studies it was preceded by two and four consecutive weeks of drug administration, respectively. Results
Short-term jlurazepam administration. Effectiveness for sleep induction and maintenance. The mean results for the key parameters for sleep induction and sleep maintenance are recorded in Table 11. In each of the four parameters, a significant improvement was
Flurazepam
Volume 19 Number 5, Part I
Table Ill. Intermediate-term drug administration: Effectiveness for sleep induction and maintenance and effects on sleep stages (N = 11)
Parameters measured Sleep latency Wake time after sleep onset Total wake time No. of wakes
Baseline
lntermediateterm drug
2-4
16-18
39.7 30.6
16.7* 13.9t
70.3 13.3
30.6* 11.7
24.6(102.1) 4.0 58.0 8.6 4.8
19.7*(89.8) 2.3* 73.8t 4.0* 0.2
Percent sleep stages REM % and min
1 2 3 4 < om. tp < 0.05.
*p
noted with short-term administration. In addition, analysis by thirds of the night showed a decrease for percent of wake time for each of the last two thirds of the night (p < 0.05). When data for each of the three drug nights were examined separately, improvement was progressively greater from night 5 to night 7; thus, total wake time was decreased by 30% on night 5 while it was decreased 61 % and 55% on nights 6 and 7. Effects on sleep stages. The mean results for sleep stages are also listed in Table 11. Percent REM sleep was significantly decreased while the total REM sleep was similar to baseline due to an increase in total sleep time with short-term drug administration. Analysis of REM percent by thirds of the night showed that the decrease in the first third was significant (p < 0.05). REM latency increased 87.6 to 92.8 min after drug (not significant). Slow-wave sleep (stages 3 and 4 combined) was decreased from 12.7 to 9.7 percent (p < 0.01). The changes in REM sleep were relatively constant for the three drug nights while the decrease in slow-wave sleep was greatest on the third drug night, night 7. Intermediate-term administration. Effectiveness for sleep induction and maintenance. The mean results for the key parameters for sleep induction and sleep maintenance are recorded in Table Ill. All of the parameters
579
Table IV. Long-term drug administration: Effectiveness for sleep induction and maintenance and effects on sleep stages (N = 5)
Long-term drug
Parameters measured Sleep latency Wake time after sleep onset Total wake time No. of wakes
Percent sleep stages REM % and min 1 2 3 4 *p
30-32
39.0 29.8
25.5* 16.2
68.8 16.7
41.7t 15.5
26.4(108.6) 4.3 56.2 7.9 5.2
22.9*(100.1)
2.lt
70.6t 3.8 0.7
< 0.05.
tp < 0.01.
except the number of wakes showed significant improvement during intermediate-drug administration. Analysis by thirds of the night showed a decrease in the percent of wake time for each third of the night, but only the decrease during the last third was significant (p < 0.01). Effects on sleep stages. The mean results for sleep stages are recorded in Table Ill. Because of the established effect on sleep stages of readaptation to the sleep laboratory, 14 the first night upon returning to the sleep laboratory (night 16) was not included in the comparisons with baseline. With intermediate-term drug administration, percent REM sleep was significantly decreased from baseline. This decrease was considerably greater than during short-term administration. Analysis of REM percent by thirds of the night showed a marked decrease in each of the first two thirds (p < 0.01) and a nonsignificant decrease in the last third. REM latency was increased from 74.9 to 103.5 min (p < 0.01). Slow-wave sleep was decreased from 13.4% to 4.1 % (p < 0.01). This decrease was considerably greater than that during short-term administration. Long-term administration. Effectiveness for sleep induction and maintenance. The mean results for the key parameters for sleep induction and sleep maintenance are
580
Kales et al.
Clinical Pharmacology and Therapeutics
Table V. Initial drug withdrawal: Effectiveness for sleep induction and maintenance and effects on sleep stages (N = 16) Drug withdrawal nights Parameters measured Sleep latency Wake time after sleep onset Total wake time No. of wakes Percent sleep stages REM % and min 1 2 3 4
Baseline 2-4
I
1
2
I
3
44.3 33.5
21.5* 16.7
21.8* 19.7
40.2 40.0
77.8 10.0
38.2t 9.6
41.5t 12.2
80.2 10.6
23.8(99.2)
24.0(111.8)
23.8(109.3)
23.5(97.4)
5.8 59.1 6.9 4.3
4.6 68.5 2.6t 0.0
5.4 66.5 4.2t 0.1
6.5 65.1 4.6t 0.3
*p < 0.05. tp
Pennsylvania State University, Milton S. Hershey Medical Center
Supported in part by Grant No. NGR39-009 and Contract No. 9-10835 from the National Aeronautics and Space Administration (NASA) and by grants from Hoffmann-La Roche Inc., Nutley, N. J. Received for publication Nov. 13, 1975. Accepted for publication Feb. 9, 1976. Reprint requests to: Or. Anthony Kales, Sleep Research and Treatment Center, Departments of Psychiatry and Pharmacology, Milton S. Hershey Medical Center, Hershey, Pa. 17m3.
576
Over the last several years, sleep laboratory studies have been increasingly used in the evaluation of the effectiveness of hypnotic drugs. The primary advantages of these studies are that they permit precise and objective measurements of effectiveness parameters
Flurazepam
Volume 19 Number 5. Part 1
577
Table I. Protocol for evaluating short-, intermediate-, and long-term hypnotic drug effectiveness Night
P
D
x 2 to 4 5 to 7
Lab
Home
X
X X
8 to 15
X
16 to 18
X
19 to 29
X
30 to 32
X
X X X X X X
throughout the night. Their major limitation, the small number of subjects who can be evaluated, is due to the cost of monitoring subjects throughout the night and the limited number of subjects who can be accommodated at one time in the sleep laboratory. 2. 6 In the last eight years, we have extensively evaluated fiurazepam (Dalmane 30 mg) in the sleep laboratory with several different protocols. 2 • 6 Initially, the efficacy of the drug was evaluated over periods of short-term drug administration (3 days).lo The results were positive, and the efficacy of the drug was then evaluated for an intermediate term (2 wk). 1 Since in clinical practice hypnotic drugs are frequently prescribed for long periods ,4 we evaluated the efficacy of the drug over a 4-wk period of administration. 7 These periods are substantially longer than those in the usual trials of hypnotic drugs, which generally include only I to 3 consecutive nights. In our study, by combining the data from 6 of our sleep laboratory evaluations of fiurazepam,l, 7, 8, 10, 11 we were able to determine the effectiveness of the drug in a large sample of insominac sUbjects. Methods
Subject selection. Twenty-three insomniac patients were selected from the Sleep Disorders Clinic of our Sleep Research and Treatment Center. The patients selected were not psychotic. Each patient had a history of difficulty either in falling asleep, staying asleep, or both.
Condition Adaptation to sleep lab environment Baseline measurements Initial and short-term drug effectiveness Evaluation in home surroundings Intermediate-term (2 wk) drug effectiveness Evaluation in home surroundings Long-term (4 wk) drug effectiveness
To qualify for the study, patients had to have a reported history of an average of longer than 45 min to fall asleep, or less than 6V2 hr of total sleep time. Design of studies. The short-term protocols (N = 12) included only a period of short-term drug administration. Automatically, the intermediate-term protocols (N = 6) included both short- and intermediate-term drug adrrumstration, and the long-term protocol (N = 5) included short-, intermediate-, and long-term drug administration. Since all of the protocols included a short-term period of drug administration, there were 23 short-term evaluations and in the same way there were 11 intermediate-term evaluations. The basic design of the different studies is described in Table I. In all three designs an initial placebo period was followed by drug and a subsequent placebo period. The initial series of placebo nights (1 to 4) allowed for adaptation to the sleep laboratory and baseline measurements. The first series of drug nights (5 to 7) permitted the monitoring of short-term drug effects. In the short-term protocol, the subjects were withdrawn from the drug by substituting placebo after the first series of drug nights. In the intermediate- and long-term protocols, the patients were continued on drug while at home (8 to 15), and then the intermediate-term drug effects were evaluated in the laboratory (16 to 18). For the subjects on the intermediate-term protocol, placebo was substituted for drug, while those on the long-term protocol were
578
Kales et al.
Clinical Pharmacology and Therapeutics
Table 11. Short-term drug administration: Effectiveness for sleep induction and maintenance and effects on sleep stages (N = 23) Short-term drug administration Parameters measured Sleep latency Wake time after sleep onset Total wake time No. of wakes Percent sleep stages REM % and min I
2 3 4
Baseline 2-4
5
I
6
I
7
I
5-7
45.4 32.4
37.2 16.2*
21.7* 8.9*
21.7* 13.5*
26.8* 12.9*
77.8 10.6
53.4* 8.0*
30.6* 6.0*
35.2* 6.9*
39.7* 6.9*
23.8(99.3) 5.0 58.6 8.5 4.2
21.9(97.7) 3.4* 63.8* 7.2 3.5
21.3t(99.9) 2.6* 65.5* 8.2 2.4
22.1(102.5) 3.3* 66.9* 6.3 1.4
21.8t(lOO.0) 3.1 * 65.4* 7.2 2.5
*p < 0.01. tp < 0.05.
continued on drug at home (19 to 29). Longterm drug effects were then evaluated in the laboratory (30 to 32) and placebo was again substituted for drug. In the short-term protocols, the placebo withdrawal period extended for 2 or 3 nights. In the intermediate- and long-term protocols, there were 4 initial placebo withdrawal nights. In the long-term protocol there was an extended placebo withdrawal period and the subjects were also monitored for 2 wk after drug withdrawal. Throughout the study, the subjects were instructed not to nap, not to significantly alter their level of daily physical activity, and not to use any drugs. On each night, subjects were administered single gelatin capsules. These capsules contained either flurazepam 30 mg or a matching placebo. On the laboratory nights, the subjects were continuously monitored for 8 hr with electroencephalogram (EEG), electromyogram (EMG) , and electrooculogram (EOG). These all-night sleep recordings were scored according to standardized criteria independent of any knowledge of the experimental conditions. 13 Parameters evaluated in the study included sleep induction (sleep latency), sleep maintenance (wake time after sleep onset and number of awakenings), sleep stages including stages REM, 1,2,3, and 4, number of REM periods, and interval from sleep onset to the first REM period (REM latency).
In comparing the data for each of the three treatment and two withdrawal conditions (short-, intermediate-, and long-term drug administration and short-term and extended withdrawal) to baseline, the baseline for each comparison consisted of data from only those patients who were included in the treatment condition being contrasted. Each of the comparisons to baseline were evaluated using a correlated t test. In analyzing the short-term withdrawal effects of flurazepam, the data were obtained from those studies that included at least three consecutive nights of placebo withdrawal. The withdrawal data from two studies was not included since each of these studies used only two consecutive withdrawal nights. The withdrawal data that were then analyzed were derived from four different studies (N = 4 in each study). In two of these studies, the withdrawal period was preceded by five consecutive nights of drug administration, while in the other two studies it was preceded by two and four consecutive weeks of drug administration, respectively. Results
Short-term jlurazepam administration. Effectiveness for sleep induction and maintenance. The mean results for the key parameters for sleep induction and sleep maintenance are recorded in Table 11. In each of the four parameters, a significant improvement was
Flurazepam
Volume 19 Number 5, Part I
Table Ill. Intermediate-term drug administration: Effectiveness for sleep induction and maintenance and effects on sleep stages (N = 11)
Parameters measured Sleep latency Wake time after sleep onset Total wake time No. of wakes
Baseline
lntermediateterm drug
2-4
16-18
39.7 30.6
16.7* 13.9t
70.3 13.3
30.6* 11.7
24.6(102.1) 4.0 58.0 8.6 4.8
19.7*(89.8) 2.3* 73.8t 4.0* 0.2
Percent sleep stages REM % and min
1 2 3 4 < om. tp < 0.05.
*p
noted with short-term administration. In addition, analysis by thirds of the night showed a decrease for percent of wake time for each of the last two thirds of the night (p < 0.05). When data for each of the three drug nights were examined separately, improvement was progressively greater from night 5 to night 7; thus, total wake time was decreased by 30% on night 5 while it was decreased 61 % and 55% on nights 6 and 7. Effects on sleep stages. The mean results for sleep stages are also listed in Table 11. Percent REM sleep was significantly decreased while the total REM sleep was similar to baseline due to an increase in total sleep time with short-term drug administration. Analysis of REM percent by thirds of the night showed that the decrease in the first third was significant (p < 0.05). REM latency increased 87.6 to 92.8 min after drug (not significant). Slow-wave sleep (stages 3 and 4 combined) was decreased from 12.7 to 9.7 percent (p < 0.01). The changes in REM sleep were relatively constant for the three drug nights while the decrease in slow-wave sleep was greatest on the third drug night, night 7. Intermediate-term administration. Effectiveness for sleep induction and maintenance. The mean results for the key parameters for sleep induction and sleep maintenance are recorded in Table Ill. All of the parameters
579
Table IV. Long-term drug administration: Effectiveness for sleep induction and maintenance and effects on sleep stages (N = 5)
Long-term drug
Parameters measured Sleep latency Wake time after sleep onset Total wake time No. of wakes
Percent sleep stages REM % and min 1 2 3 4 *p
30-32
39.0 29.8
25.5* 16.2
68.8 16.7
41.7t 15.5
26.4(108.6) 4.3 56.2 7.9 5.2
22.9*(100.1)
2.lt
70.6t 3.8 0.7
< 0.05.
tp < 0.01.
except the number of wakes showed significant improvement during intermediate-drug administration. Analysis by thirds of the night showed a decrease in the percent of wake time for each third of the night, but only the decrease during the last third was significant (p < 0.01). Effects on sleep stages. The mean results for sleep stages are recorded in Table Ill. Because of the established effect on sleep stages of readaptation to the sleep laboratory, 14 the first night upon returning to the sleep laboratory (night 16) was not included in the comparisons with baseline. With intermediate-term drug administration, percent REM sleep was significantly decreased from baseline. This decrease was considerably greater than during short-term administration. Analysis of REM percent by thirds of the night showed a marked decrease in each of the first two thirds (p < 0.01) and a nonsignificant decrease in the last third. REM latency was increased from 74.9 to 103.5 min (p < 0.01). Slow-wave sleep was decreased from 13.4% to 4.1 % (p < 0.01). This decrease was considerably greater than that during short-term administration. Long-term administration. Effectiveness for sleep induction and maintenance. The mean results for the key parameters for sleep induction and sleep maintenance are
580
Kales et al.
Clinical Pharmacology and Therapeutics
Table V. Initial drug withdrawal: Effectiveness for sleep induction and maintenance and effects on sleep stages (N = 16) Drug withdrawal nights Parameters measured Sleep latency Wake time after sleep onset Total wake time No. of wakes Percent sleep stages REM % and min 1 2 3 4
Baseline 2-4
I
1
2
I
3
44.3 33.5
21.5* 16.7
21.8* 19.7
40.2 40.0
77.8 10.0
38.2t 9.6
41.5t 12.2
80.2 10.6
23.8(99.2)
24.0(111.8)
23.8(109.3)
23.5(97.4)
5.8 59.1 6.9 4.3
4.6 68.5 2.6t 0.0
5.4 66.5 4.2t 0.1
6.5 65.1 4.6t 0.3
*p < 0.05. tp
