Introduction

Top Abstract Introduction Results Discussion References

Insomnia is among the complaints most frequently expressed by patients when visiting a physician (Kupfer and Reynolds, 1997). Although hypnotic drugs should not be used on a chronic basis, consumption of benzodiazepines and related drugs such as zopiclone and zolpidem is still considerable (Langer et al., 1999). In the last decades, attention has been focused on negative consequences of hypnotic drug use such as impairment of psychomotor performance and memory, abuse and dependence potential, and rebound insomnia (Langer et al., 1999). This is of particular relevance because hypnotics are often prescribed to relatively healthy people suffering from sleep disturbances related to circadian rhythms (e.g., jet lag, shift work) and who may drive automobiles (Dingemanse, 1995). Therefore, hypnotic drugs not prone to the disadvantages of full agonists at the benzodiazepine receptor are still being sought (Martin et al., 1994).

Ro 41-3696, (S)-1-[(10-chloro-6,7-dihydro-4-oxo-3-phenyl-4H-benzo[a]quinolizin-1-yl)-carbonyl]-3-ethoxy-pyrrolidine (Fig. 1), is a partial agonist at the benzodiazepine receptor that, based on preclinical data, offers the potential for developing a hypnotic without effects on performance and memory and with a low physical dependence liability (Martin et al., 1994; Scherschlicht 1994; Tsuboi et al., 1994). In contrast to zolpidem, a full BZ1 receptor agonist, compounds belonging to the quinolizinone class are partial, nonselective benzodiazepine receptor agonists (Jenck et al., 1992; Holm and Goa, 2000). The main behavioral differences between partial and full benzodiazepine receptor agonists are in their side effect profile (Podhorna and Krsik, 2000). The clinical pharmacology profile of Ro 41-3696 after single and multiple doses has been studied in young and elderly subjects, respectively (Dingemanse et al., 1995a,b, 2000). A wide range of single oral doses (0.1-30 mg) was well tolerated and central nervous system depressant activity was observed only at doses of 10 and 30 mg (Dingemanse et al., 1995b). The drug was quickly absorbed, had an elimination half-life of about 4 h, and was rapidly metabolized to its desethylated derivative Ro 41-3290 (Fig. 1) (Dingemanse et al., 1995b). At 1.5 h after night-time administration, Ro 41-3696 (1-10 mg) induced smaller effects on psychomotor performance and memory than 10 mg of zolpidem (Dingemanse et al., 1995a). Small but statistically significant effects were observed at 8 h after administration of doses of 10 and 30 mg (Dingemanse et al., 1995a,b). In view of the pharmacokinetic profile of the parent compound, this suggests the presence of a long-lived active metabolite. Since maximum plasma concentrations of Ro 41-3290 are more than 10-fold greater than those of Ro 41-3696 and the elimination half-life of the metabolite is approximately 8 h, it is likely that Ro 41-3290 is at least partly responsible for the overall pharmacodynamic profile after administration of Ro 41-3696. In general, hypnotics with a short elimination half-life show a favorable efficacy versus adverse effects profile with regard to functioning on awakening and daytime functioning (Monti and Monti, 1995). Active metabolites play an important role in the effects of many psychotropic drugs (Garattini, 1985). Whenever possible, their pharmacokinetics and pharmacodynamics are to be investigated after administration per se in a clinical pharmacology study (Dingemanse et al., 1988).

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Chemical structure of Ro 41-3696 and its O-desethyl-metabolite Ro 41-3290.

The objectives of the present study were to assess the tolerability, safety, pharmacodynamics, and pharmacokinetics of Ro 41-3290 in comparison with zolpidem after single-dose administration to healthy young male subjects. Zolpidem was chosen as comparator drug because in many countries it is a leading hypnotic, and extensive experience now exists with this drug (Langtry and Benfield, 1990; Darcourt et al., 1999; Holm and Goa, 2000). Zolpidem also fulfills several of the requirements set for further development of Ro 41-3696, e.g., not prone to hangover effects. Also, in pharmacokinetic terms, zolpidem and Ro 41-3696 are similar (Dingemanse et al., 1995b; Holm and Goa, 2000). However, zolpidem still clearly impairs psychomotor performance, which should not be the case for an ideal hypnotic.

Experimental Procedures

Subjects. Thirty male Caucasian subjects in the age range 20 to 30 years and within 20/+15% of their ideal body weight participated in this study. Ethics Committee approval was obtained from the Toegepast Natuurwetenschappelijk Onderzoek (TNO) Institutional Review Board, Leiden, The Netherlands, and all subjects gave their written informed consent before any screening procedures were performed. The entire study was conducted in full conformity with the principles of the Declaration of Helsinki and its amendments. Subjects were selected who were healthy on the basis of a medical history, physical and neurological examination, and clinical laboratory determinations. No concomitant medication was allowed during the study, and restrictions were applied regarding the intake of methylxanthine-containing beverages and food. At screening and admission into the clinic, a urine drug screen for drugs of abuse (including barbiturates and benzodiazepines) was performed.

Design. This entry-into-humans study with Ro 41-3290 was a double-blind, randomized, placebo- and zolpidem-controlled, ascending single-dose study of orally administered Ro 41-3290. The dose levels studied were 5, 10, and 30 mg of Ro 41-3290 and 10 mg of zolpidem. Ro 41-3290 had to be investigated in an ascending dose design because it had not yet been administered to humans. Bias was avoided to the greatest extent possible by inclusion of four control subjects into each group. The starting dose of 5 mg was chosen based on the following considerations. The potency of Ro 41-3290 in several animal experiments reflecting hypnotic properties is about one-tenth to one-fourth that of the parent compound. The further doses of 10 and 30 mg were selected to enable a comparison with the results obtained following 10 and 30 mg of Ro 41-3696 (Dingemanse et al., 1995a,b). The subjects were hospitalized from about 22 h before until 52 h after drug administration. Each treatment group consisted of 10 subjects who received a single dose of Ro 41-3290 (n = 6), 10 mg of zolpidem (n = 2), or placebo (n = 2). The decision to proceed to the next dose level was made on the basis of tolerability results at the previous dose level. After an overnight fast, the treatment was administered as one capsule in the morning with 150 ml of tap water, after which fasting continued for 4 h.

Subject Assessments.

Tolerability and safety Adverse events were assessed by spontaneous reports, observations, and questioning at regular intervals. In line with the requirements of regulatory authorities, an adverse event was any adverse change from the subject's baseline (pretreatment) condition that occurred during the course of the study after treatment had started, whether considered related to treatment or not. The intensity of the adverse events was rated on a three-point scale (mild, moderate, severe), and the potential relationship to drug was assessed by the investigator before breaking the code. Sitting blood pressure, pulse rate, and body temperature were measured at frequent intervals. A 12-lead electrocardiogram was recorded just before and at 1 h after drug administration. At discharge from the clinic, a physical examination and routine clinical laboratory tests were performed.

Pharmacodynamics. Psychomotor performance tests were conducted to measure information processing under the influence of the drug, and a visual serial learning test was conducted for examination of possible long-term memory effects. A battery of psychomotor tests that covered a spectrum of central nervous system functions, impairment of which is considered critical for a potential new hypnotic, were conducted. Tracking and memory search tests were performed as part of the standardized task battery, Taskomat (Boer and Wientjes, 1988; Gaillard et al., 1988; Dingemanse et al., 1995b). For training purposes, the tests were conducted four times on the day before drug administration. Just before drug administration and at 1.5, 4, 8, and 24 h after drug intake psychomotor performance was recorded. The tracking test requires both perceptual and motor processing and is sensitive to lapses of attention. The subject's task was to move a line with a small gap in a horizontal direction on a computer screen, such that a vertically moving track passed through the middle of the gap (Gaillard et al., 1988). The distance of the track to the middle of the gap was sampled at 200-ms intervals and was averaged for 14 periods of 30 s each.

Pharmacokinetics. Blood samples of 5 ml were collected into tubes containing heparin as anticoagulant via a catheter inserted into a forearm vein, just before and 0.25, 0.5, 1, 1.5, 2, 3, 4, 6, 8, 10, 15, 20, 25, 30, 35, 40, 45, and 50 h after drug administration. Blood samples were centrifuged, and plasma was separated and stored in glass tubes at 20°C pending analysis.

Evaluation.

Tolerability and safety The adverse events and clinical laboratory data were evaluated descriptively. For analysis, data from all subjects receiving placebo were combined. The same was done for subjects receiving zolpidem. Individual vital sign data were screened for values outside the predetermined normal ranges, i.e., systolic blood pressure 80 to 160 mm Hg, diastolic blood pressure 50 to 95 mm Hg, pulse rate 40 to 120 beats/minute, body temperature 35.0-37.5°C. Means of vital sign data were screened for trends. Clinical laboratory values were compared with the normal ranges supplied by the analyzing laboratory.

Pharmacodynamics. In the tracking test, the dependent measure was the root mean square error (RMS)¹ given by the following equation: RMS = (e_i²/n), in which e_i is the distance (=error) between the middle of the gap and the track at each measurement i, and n is the number of samplings. RMS values are presented as the mean RMS across periods 2 to 13, i.e., the first and the last period were omitted from the analysis.

Pharmacokinetics. Pharmacokinetic parameters (mean ± S.D.) were determined for Ro 41-3290 by model-independent methods. The maximum plasma concentration (C_max) and the time of its occurrence (t_max) relative to dosing were read directly from the concentration-time data. The terminal elimination rate constant (_z) was obtained by log linear regression analysis of the terminal portion of the curve. The elimination half-life (t_1/2) was calculated using ln(2)/_z. The area under the concentration-time curve (AUC_0-) was calculated by linear-trapezoidal summation and extrapolation to infinity.

Statistics. A one-way ANOVA was conducted on the pharmacodynamic parameters for the three groups of two placebo subjects and the three groups of two zolpidem subjects to search for a period effect.

0-

	Results

Top Abstract Introduction Results Discussion References

Tolerability. All 30 subjects completed the study according to the protocol. The total and most frequently reported adverse events are presented in Table 1. All treatments were well tolerated, there were no serious adverse events, and all events resolved without sequelae. Adverse events were reported by three, six, two, four, and six subjects treated with placebo, 5 mg of Ro 41-3290, 10 mg of Ro 41-3290, 30 mg of Ro 41-3290, and 10 mg of zolpidem, respectively. Most adverse events were judged to be of mild or moderate intensity. Two subjects reported severe somnolence (one on 10 mg of Ro 41-3290 and one on zolpidem), and one subject reported severe ataxia (zolpidem). No differences were apparent between the groups treated with any dose of Ro 41-3290 or with placebo, whereas there was a trend for more adverse events related to the central and peripheral nervous system in the subjects treated with zolpidem. There was no adverse pattern of abnormal laboratory values or vital signs observed during the study, and the abnormalities were not judged to be clinically relevant. There were no electrocardiogram changes compared with baseline or placebo. Results of investigations performed at discharge from the clinic did not reveal relevant differences from baseline values.

Pharmacodynamics. Training in the psychometric tests was considered sufficient based on the differences in the outcome of the pharmacodynamic variables between the last training session and the session just before drug intake. The placebo data indicate that subjects showed stable performance over the study day. Results of ANOVA performed on the data for the six subjects receiving placebo or zolpidem indicated the absence of significant period effects, allowing these subjects to be combined in two groups. The resulting statistical power was previously shown to be sufficient (Dingemanse et al., 1995b). For both the tracking and the memory search tests the baseline data of the placebo and active treatment groups were similar, indicating the homogeneity of the population studied.

Tracking test. RMS is presented in Fig. 2 as a function of time after the different treatments. Zolpidem induced clear effects at 1.5 h after administration that had disappeared by 4 h. The Ro 41-3290 treatments elicited moderate effects peaking at 4 to 8 h, whereas performance at 24 h was again back to baseline. Statistical analysis of RMS results with both mean and maximum differences to baseline yielded no significant differences between the treatments.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Median RMS score in the tracking test as a function of time for the different treatments. *, placebo; , 5 mg of Ro 41-3290; , 10 mg of Ro 41-3290; , 30 mg of Ro 41-3290; ×, 10 mg of zolpidem.

Memory search test. The slope of the linear regression of the relationship between reaction time and the number of comparisons is given in Fig. 3 as a function of time and treatment. The magnitude of effect was similar for all active treatments, but the maximum effect for zolpidem was reached at 1.5 h and for Ro 41-3290 at 4 to 8 h after administration. There were no differences between the three doses of the latter drug. Statistical analysis of the slope results with both mean and maximum differences to baseline did not yield significant differences between the treatments. Table 2 presents the variability of the pharmacodynamic variables for the tracking and the memory search test, both for baseline and maximum effect.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Median slope in the memory search test as a function of time for the different treatments. *, placebo; , 5 mg of Ro 41-3290; , 10 mg of Ro 41-3290; , 30 mg of Ro 41-3290; ×, 10 mg of zolpidem.

Long-term memory test. A summary of the parameters of the 15-words test is given in Table 3. There was a clear trend toward an increase in memory impairment with zolpidem, whereas Ro 41-3290 did not differ from placebo. Figure 4 depicts the difference between the number of correct words in the learning and recall trials.

View larger version (45K): [in this window] [in a new window]   Fig. 4.   Difference in number of correct words between the learning and the recall trial as a function of treatment. Data are presented as means ± S.E. (n = 6).

Pharmacokinetics. Figure 5 presents the mean plasma concentration-time profile of Ro 41-3290 following the different doses. A summary of the pharmacokinetic parameters in each treatment group is given in Table 4. The drug was relatively slowly absorbed at all dose levels with a t_max of approximately 2.5 h. C_max and AUC increased proportionally with dose. No statistically significant differences were observed when subjecting dose-normalized C_max and AUC values to ANOVA. The disposition phase was biphasic with a mean terminal elimination half-life of approximately 8 h.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Plasma concentration-time profiles of Ro 41-3290 following 5, 10, and 30 mg of Ro 41-3290.  Data are presented as the mean ± S.E. (n = 6). , 5 mg; , 10 mg; , 30 mg.

	Discussion

Top Abstract Introduction Results Discussion References

The aim of the development of the quinolizinone Ro 41-3696 is to identify a hypnotic that has a sleep-inducing and -maintaining effect but that does not lead to any psychomotor or memory impairment. The pharmacokinetics and pharmacodynamics of Ro 41-3696 have been investigated in detail after single- and multiple-dose administration to young and elderly subjects, respectively (Dingemanse et al., 1995a,b, 2000). These studies provided indications that the pharmacodynamic effects observed after administration of Ro 41-3696 could at least partly be due to the presence of the long-lived metabolite Ro 41-3290. This compound is a partial agonist at the benzodiazepine receptor, inhibiting the in vitro binding of flumazenil with a K_i value of 0.77 nM. Virtually no binding activity (receptor occupancy <20%) was detected after oral administration of doses up to 100 mg/kg in rats and mice (data on file, F. Hoffmann-La Roche). The ED₅₀ for in vivo binding after i.v. administration in both species is approximately 1.6 mg/kg. Data suggest that the drug penetrates the blood-brain barrier relatively slowly. Plasma concentrations increased disproportionately with dose, which may be related to the low water solubility of Ro 41-3290. The potential of Ro 41-3290 to replace its parent compound as an investigational hypnotic drug was explored in an innovative study design, namely inclusion of a comparator drug into an entry-into-humans study.

The adverse events reported by subjects treated with Ro 41-3290 did not show any consistent dose-dependent phenomenon, and the differences with subjects treated with placebo were minor. The similarity in type and incidence of adverse events following intake of Ro 41-3290 is in accordance with a shallow dose-response relationship in this dose range, as is to be expected for a partial agonist at the benzodiazepine receptor. A dose of 10 mg of zolpidem, which is the recommended initial dosage in adult patients (Lorizio et al., 1990), clearly induced more adverse events than Ro 41-3290. The events reported were in accordance with data published previously. The most common adverse events in larger clinical trials were dizziness and lightheadedness (5.2%), somnolence (5.2%), headache (3%), and gastrointestinal complaints (3.6%) (Langtry and Benfield, 1990; Holm and Goa, 2000). In agreement with results of preclinical studies (Podhorna and Krsik, 2000), partial benzodiazepine receptor agonists appear to be better tolerated than full agonists. However, the clinical relevance of this difference can only be determined when results of efficacy studies with Ro 41-3290 in insomniacs are available.

Drug effects were studied at 1.5 h, the time of expected maximum exposure to Ro 41-3290, at 4 h, and at 8 h after drug administration, i.e., the normal duration of one night's sleep (Hindmarch, 1991). The pharmacodynamic results obtained in the tracking and memory search test were similar. Moderate but consistent effects were obtained with the three doses of Ro 41-3290. The effects were most pronounced at 4 h after drug intake and indistinguishable among the doses. The maximum pharmacodynamic effects appear to be delayed when compared with the time of the maximum plasma concentrations, suggesting a slow penetration through the blood-brain barrier. Zolpidem induced a marked effect at 1.5 h after intake, in particular in the tracking task, whereas impairment was marginal at 4 h and had disappeared at 8 h. The quick onset and short duration of effect induced by zolpidem are in accordance with its fast absorption, absence of pharmacologically active metabolites, and elimination half-life of approximately 2.5 h (Langtry and Benfield, 1990; Holm and Goa, 2000). The effects in both the tracking and the memory search test at 1.5 h after administration of zolpidem in this study were nearly 2-fold less pronounced than after night-time administration (Dingemanse et al., 1995a). This could be due to the fact that subjects are more sensitive to central depressant effects of drugs during the night. However, this phenomenon was not apparent when comparing the results following intake of Ro 41-3696 during daytime and at night (Dingemanse et al., 1995a,b). This illustrates that comparing results obtained in different studies and in different subjects should be performed with extreme caution, even when the same methodology in the same center was applied.

In the long-term memory test, learning, recall, recognition, and relearning were affected to a marginal extent by Ro 41-3290 and zolpidem. The results with the latter drug were in close agreement with those obtained after night-time administration of the same dose (Dingemanse et al., 1995a). Subjects who received the highest dose of Ro 41-3290 (30 mg) showed some improvement of memory, and those who received zolpidem showed some impairment. Ro 41-3290 induced smaller effects on memory than equivalent doses of the parent compound Ro 41-3696 (Dingemanse et al., 1995a,b). Because of a possible time delay between plasma concentration and pharmacodynamic effect (see above), it cannot be excluded that the effects of Ro 41-3290 on long-term memory would have been more pronounced when presenting the list of words at 3 h instead of 2 h after administration. However, when viewing the results obtained with the parent compound Ro 41-3696 (Dingemanse et al., 1995b), a drug which is quickly absorbed and whose pharmacodynamic effects are also quickly attained, there appear to be differences in the influence on memory between partial agonists and zolpidem.

The psychomotor and memory-impairing effects of high doses of Ro 41-3290 are much less pronounced than those of Ro 41-3696, suggesting that the maximum effects observed around 1.5 h after administration of Ro 41-3696 are mainly caused by the parent drug. However, the small residual effects at 8 h after intake of Ro 41-3696 may be due to the presence of Ro 41-3290.

The pharmacokinetics of Ro 41-3290 were also extensively investigated in this study because an integrated pharmacokinetic-pharmacodynamic approach is of great value in early drug development (Greenblatt et al., 1987). Furthermore, it has been demonstrated that onset and duration of action of hypnosedatives primarily depend on their pharmacokinetic profile (Nishino and Mignot, 1999). Ro 41-3290 is more slowly absorbed than Ro 41-3696 with t_max values of around 2.5 and 1 h, respectively (Dingemanse et al., 1995b). This may be related to the poor water solubility of Ro 41-3290. To ensure a rapid onset of action after oral dosing, a hypnotic should rapidly reach the systemic circulation (Jochemsen et al., 1983). The maximum concentrations of Ro 41-3290 when administered per se are similar to those reached upon administration of Ro 41-3696. This suggests that the extent of absorption of both compounds is similar and that indeed Ro 41-3290 is quantitatively the most important metabolite of Ro 41-3696. The elimination half-life of Ro 41-3290 found in this study is similar to the one following administration of Ro 41-3696 (Dingemanse et al., 1995b). The pharmacokinetic parameters of zolpidem have been well characterized and shown to be consistent for the population in the present study, i.e., young, healthy male subjects who were fasting until 4 h after drug administration (Langtry and Benfield, 1990; Fraisse et al., 1996). It is very unlikely that the determination of zolpidem plasma levels would have changed any conclusion from this study because no circumstances were identifiable that could have exerted an important influence on the pharmacokinetics of zolpidem, such as disease characteristics, comedication, food intake, etc. Furthermore, the pharmacodynamic results obtained with zolpidem did not provide any indication that absorption of the drug had been irregular or incomplete.

It can be concluded from the results of the present study that the pharmacodynamic and pharmacokinetic properties of Ro 41-3290 are not appropriate to replace Ro 41-3696 as an investigational hypnotic drug. Absorption of Ro 41-3290 from the gastrointestinal tract is relatively slow and variable and, based on the concentration-effect time delay, the compound probably also passes the blood-brain barrier slowly as had been suggested on the basis of animal experiments. Following intake of Ro 41-3696, the pharmacodynamic effects, in particular those early after administration (in clinical terms important for sleep induction), are mainly caused by Ro 41-3696 itself and not by Ro 41-3290. Some effects seen at later time points (in clinical terms important for hangover) may be caused by Ro 41-3290 because of its longer elimination half-life and concentration-effect time delay. These data suggest the limited utility of Ro 41-3290 as a treatment for insomnia.