Figures
Abstract
Background and Aims
Compounds that act on GABA-receptors produce anti-hyperalgesia in animal models, but little is known on their effects in humans. The aim of this study was to explore the potential usefulness of GABA-agonism for the control of pain in humans. Two agonists at the benzodiazepine-binding site of GABAA-receptors (clobazam and clonazepam) were studied using multiple experimental pain tests. Positive results would support further investigation of GABA agonism for the control of clinical pain.
Methods
In a randomized double-blind crossover design, 16 healthy male volunteers received clobazam 20 mg, clonazepam 1 mg and tolterodine 1 mg (active placebo). The area of static hyperalgesia after intradermal capsaicin injection was the primary endpoint. Secondary endpoints were: area of dynamic hyperalgesia, response to von Frey hair stimulation, pressure pain thresholds, conditioned pain modulation, cutaneous and intramuscular electrical pain thresholds (1, 5 and 20 repeated stimulation), and pain during cuff algometry.
Results
For the primary endpoint, an increase in the area of static hyperalgesia was observed after administration of placebo (p<0.001), but not after clobazam and clonazepam. Results suggestive for an anti-hyperalgesic effect of the benzodiazepines were obtained with all three intramuscular pain models and with cuff algometry. No effect could be detected with the other pain models employed.
Conclusions
Collectively, the results are suggestive for a possible anti-hyperalgesic effect of drugs acting at the GABAA-receptors in humans, particularly in models of secondary hyperalgesia and deep pain. The findings are not conclusive, but support further clinical research on pain modulation by GABAergic drugs. Because of the partial results, future research should focus on compounds acting selectively on subunits of the GABA complex, which may allow the achievement of higher receptor occupancy than unselective drugs. Our data also provide information on the most suitable experimental models for future investigation of GABAergic compounds.
Citation: Vuilleumier PH, Besson M, Desmeules J, Arendt-Nielsen L, Curatolo M (2013) Evaluation of Anti-Hyperalgesic and Analgesic Effects of Two Benzodiazepines in Human Experimental Pain: A Randomized Placebo-Controlled Study. PLoS ONE 8(3): e43896. https://doi.org/10.1371/journal.pone.0043896
Editor: John E. Mendelson, California Pacific Medicial Center Research Institute, United States of America
Received: July 4, 2012; Accepted: November 12, 2012; Published: March 15, 2013
Copyright: © 2013 Vuilleumier et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The study was funded by the Swiss National Science Foundation (SPUM 33CM30-124117) and the Scientific Funds of the University Department of Anesthesiology and Pain Therapy of the University of Bern. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Chronic pain is still a largely unresolved problem. Effective pharmacological approaches need to target alterations in the nociceptive system that are responsible for pain and disability. Extensive animal research has demonstrated that pain conditions are associated with plastic changes of the central nervous system that alter the processing of the somatosensory and nociceptive input [1]. These changes result in reduced pain thresholds to sensory stimuli, enhanced pain after supra-threshold stimulation and enlargement of the areas of local and referred pain. All these manifestations have been consistently detected in human chronic pain [2] and are likely to be highly relevant in terms of amplification of pain and disability. Thus, pharmacological targeting of neuroplastic changes is one important aim of translational pain research.
A relevant aspect of neuroplastic changes in inflammatory and neuropathic conditions is the reduction in inhibitory glycinergic and GABAergic control of dorsal horn neurons: a reduction in the GABAA -mediated endogenous inhibitory control within the central nervous system leads to exaggerated pain and hyperalgesia [3]. Potentiation of GABAA receptor-mediated synaptic inhibition by benzodiazepines reverses pathologically increased pain sensitivity in animal studies [4], [5]. Subtype-selective compounds targeting the alpha2 and/or alpha3 subunit of the GABAA receptor produce antihyperalgesia in mice and rats without sedation and without tolerance induction [6]. These findings open new perspectives for a more selective targeting of pain pathways with GABAergic drugs.
Benzodiazepines produce anti-hyperalgesia in animal models by acting as agonists at the benzodiazepine-binding site of GABAA receptor [4]. These compounds can therefore be studied to explore the potential usefulness of GABA-agonism in human pain conditions. To date, little is known on the effects of GABAA receptor targeting drugs on nociceptive processes in humans. A review on the efficacy of non-opioid analgesics in human experimental pain models did not include these drugs [7]. We are not aware of comprehensive experimental human studies on the efficacy of drugs acting at GABA-receptors. This information is important to evaluate the potential clinical usefulness of these compounds in pain management and to guide future research.
The aim of the present study was to analyze the effect of two agonists at the benzodiazepine-binding site of GABAA receptors (shortly named as benzodiazepines) on different mechanisms of pain processing, using a multimodal experimental testing procedure in healthy volunteers. Clonazepam was studied because it is probably the most commonly prescribed benzodiazepine in chronic pain management. Clobazam is another benzodiazepine that may cause less sedation than clonazepam [8], [9] and might also have anti-hyperalgesic properties. Clobazam has been therefore selected as test compound, with clonazepam as positive control and tolterodine as active placebo.
The ultimate aim of the study was to explore the potential usefulness of GABA-agonism for the control of clinical pain, providing guide for future clinical research. The results were positive for a subgroup of experimental pain modalities, supporting further investigation of GABA-agonism in clinical pain.
Methods
The protocol for this trial and supporting CONSORT checklist are available as supporting information; see Checklist S1 and Protocol S1.
Ethic statement
The study was approved by the cantonal ethics committee (No. 152/09), registered in the Clinical Trials Protocol Registration System (NCT01011036), and performed in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants.
Design
This is a randomized double-blind crossover study, using pain assessment methods that explore different nociceptive mechanisms. Clobazam 20 mg, clonazepam 1 mg and tolterodine 1 mg (active placebo) were compared. At the end of each session the benzodiazepine antagonist flumazenil 0.2 mg was administered to evaluate whether the observed effects can be reversed.
Setting
The experiments were performed at the University Department of Anesthesiology and Pain Therapy, Inselspital Bern, Switzerland. Pharmacokinetic and pharmacogenomic investigations were performed at the Division of Clinical Pharmacology and Toxicology, University Hospital of Geneva.
Subjects
Volunteers were recruited by advertisement at the Inselspital and at the University of Bern by the first author (P.V.). Sixteen healthy volunteers were tested between December 2009 and July 2010 (figure 1). They received 200 Swiss Francs for each session and a total of 800 Swiss Francs if they completed all the three experiments.
Inclusion criteria were: european ethnicity (in order to minimize pharmacogenetic variations), male gender (in order to avoid possible variation due to hormonal changes during the menstrual cycle), age 18–55 years old, induction of static mechanical hyperalgesia by capsaicin as described below and status as non smoker or moderate smoker (≤10 cigarettes/day).
Exclusion criteria were: current or past history of drug or alcohol abuse, intake of any psychotropic drug currently or in the last month, chronic alcohol intake, any concomitant illness, current or regular intake of any drugs that might affect pain or nociception.
Study medication
Clobazam, clonazepam and tolterodine were given orally to the same volunteers in a randomized order on three different sessions, with a minimal washout interval between two consecutive sessions of two weeks. All subjects were tested in the morning and had been instructed to take a very light breakfast at home. The minimum interval between breakfast and drug intake was 3 h.
Clobazam (Urbanyl®, Sanofi-Aventis AG, Meyrin, Switzerland) is a benzodiazepine. To our knowledge, it has never been tested for analgesic or antihyperalgesic properties. Unlike 2 mg clonazepam, 20 mg clobazam did not affect significantly cognitive and psychomotor functions, possibly because of its unusual 1–5 chemical structure [9]. Clonazepam (Rivotril®, Roche Pharma AG, Reinach, Switzerland) is probably the benzodiazepine most widely used for the treatment of clinical pain, although the evidence behind this practice is weak [10]. It was therefore chosen as positive control.
The doses of clobazam and clonazepam were chosen based on the following considerations. Clonazepam is typically prescribed in chronic neuropathic pain at doses between 0.5 and 1 mg, and 2 mg a day is an average dose for this indication [11]. Typical anticonvulsant starting doses of clonazepam and clobazam are 0.5–2 mg and 10–20 mg, respectively [12]. A dose of 20 mg of clobazam has been shown to be less sedative than 1 mg of clonazepam and should be equipotent [8]. Therefore these doses have been chosen for the present study.
Tolterodine (Detrusitol®, Pharmacia Gmbh, Pfizer Group, Berlin, Germany) is an anticholinergic compound [13]. Anticholinergic compounds usually cause some sedation and dry mouth; to our knowledge, they are devoid of analgesic effects. Because of its sedative properties, tolterodine has been chosen as an active placebo in order to keep double blinding.
Flumazenil (Anexate®, Roche Pharma AG, Reinach, Switzerland) selectively antagonizes or attenuates the effects of benzodiazepines on GABAA receptors [14]. To minimize the risk of seizure, a dose of 0.2 mg iv was chosen.
The subjects received the study medications in a randomized order. A computer-generated random list was prepared by the hospital pharmacy and was known only by the hospital pharmacy. The drugs were enclosed by the hospital pharmacy in the same type of capsule to assure blinding. The pharmacy delivered the blinded capsules to the investigators, numbered according to the defined randomization order. The first author (P.V.) assigned participants to the experimental sessions.
Tests
Endpoints.
The primary endpoint was the area of static hyperalgesia induced by intradermal capsaicin injection. The additional measures were secondary endpoints.
General methodological aspects.
Because of the very limited information on which experimental pain models would be sensitive to the effects of benzodiazepines, we applied a multimodal testing procedure that is expected to explore different aspects of nociceptive processes. The rationale is summarized in table 1. In particular, we applied models of hyperalgesia, tissue-specific pain sensitivity (cutaneous vs. muscular), sensitivity to short and ongoing painful stimulation, temporal summation (increase in pain perception during stimulation of constant intensity), stimulus-response function (dependency of pain rating from different stimulation intensities) and conditioned pain modulation (exploring endogenous pain modulation).
Before starting each session, trainings of the pain tests were performed until the subjects were familiar with the testing procedures.
All the tests were applied on the dominant side, except the cold pressor test and cuff algometry (see below for explanations). All subjects but one were right-handed. Each subject was tested on the same side in all three experimental sessions.
Volunteers were not allowed to see the area tested and any read-outs from any instruments.
Intradermal capsaicin.
Intradermal injection of capsaicin causes a brief stinging/burning pain at the injection site followed by development of secondary hyperalgesia, i.e. hyperalgesia detected at surrounding (not injected) skin [15], [16]. Secondary hyperalgesia after intradermal capsaicin is the result of sensitization of the central nervous system, which is one of the most relevant aspects of neuroplastic changes [17]. This model was sensitive to the action of GABAergic compounds in animal studies [18].
The capsaicin solution was prepared by the Hospital Pharmacy and underwent sterile filtration into sterile septum-sealed vials. The sterile solution (1 mg/ml) was placed into sterile syringes and used for intradermal injection at room temperature. The skin temperature was measured using a digital thermometer, and was 32°C+/−1 in all subjects.
Subjects remained semi-supine during the experiment. Before capsaicin injection, the skin of the forearm was cleaned using an antiseptic wipe and allowed to air dry. The injection site was marked midway between the elbow and the wrist. Using a surgical skin marker, 4 lines were drawn through the injection site, so that they intersected the corners of a regular octagon (figure 2). These lines were marked every 0.5 cm from the center to the edges using a predefined grid. This resulted in the creation of triangles radiating from the intersection of the 4 lines outwards. We determined the area of hyperalgesia by adding the areas of each triangle in which hyperalgesia was recorded.
The dots represent the points where stimulation was applied. The circle in the center of the hectagon is the site of capsaicin injection.
Using a 1-ml tuberculin syringe fitted with a 27-gauge disposable needle, 100 µl capsaicin were injected epidermally into the skin, at the intersection of the 4 vectors previously drawn. A white skin elevation appeared during injection ensuring correct injection.
Static mechanical hyperalgesia was judged to be present when the subjects reported that applying a 512 mN von Frey Hair, 5 and 30 minutes after capsaicin injection, elicited pain with an intensity of at least 4 using a numerical rating scale (NRS, whereby 0 = no pain and 10 = worst pain). Only these subjects were included in the study. In order to document pain induction, the subject recorded pain intensity at injection and 5 minutes post-injection using a 10 cm visual analogue pain scale (VAS), whereby 0 = no pain and 10 = worst pain imaginable.
The area of secondary hyperalgesia was assessed using a calibrated 512 mN von Frey hair. The punctuated probe was moved along the 8 radial lines defined above, starting approximately 6 cm away from the site of injection, at each mark in steps of 0.5 cm. The weighted von Frey hair was placed gently on the skin and the load was applied for 2 seconds. The volunteers were asked to report when the pricking sensation changed to a pain sensation. At least 4 seconds elapsed between consecutive stimuli. This procedure was repeated for each vector. The number corresponding to the marker at which sensation changes as described above was noted and the individual numbers were used to calculate the area of hyperalgesia.
The area of dynamic mechanical hyperalgesia was determined by gently stroking a hand-held cotton wool tip on a 1 cm strip of the skin, at a rate of approximately 1 cm/s. Subjects were asked to report when the sensation changed from a non-painful to a painful sensation. The borders of dynamic hyperalgesia were delineated similarly to the determination of static hyperalgesia.
Additionally, mechanical pain sensitivity was assessed using a set of seven weighted pinprick stimuli to obtain a stimulus-response function. The test was applied 1 cm inside the outer border of the pinprick hyperalgesic area, using a set of 7 graded von Frey hair mechanical stimulators with fixed stimulus intensities (custom-made at Aalborg University, Denmark). The flat contact area of the stimulators has a 0.2 mm diameter, and the 7 stimulators exert forces of 8, 16, 32, 64, 128, 256 and 512 mN. The order of stimulation was defined randomly by computer. The stimulators were applied at a rate of 2 s on/2 s off. Subjects were asked to give a pain rating for each stimulus on a 10 cm VAS. Two assessments for each stimulation were made and the mean of these 2 measurements was used for the data analysis.
Pressure stimulation.
Pain detection and tolerance thresholds were measured with an electronic pressure algometer (Somedic, Hörby, Sweden) applied at the center of the pulp of the 2nd toe. The probe had a surface area of 1 cm2. The pressure was increased from 0 at a rate of 30 kPa/s to a maximum pressure of 1200 kPa. Pain detection threshold was defined as the point at which the pressure sensation turned to pain. Pain tolerance threshold was defined as the point at which the subject felt the pain as intolerable. If the threshold was not reached at 1200 kPa, this value was considered as threshold. Three assessments were made and the mean of these 3 measurements were used for the data analysis.
Conditioned pain modulation (CPM).
This method explores the endogenous modulation of nociceptive input. Under normal conditions, pain after application of a “test” nociceptive stimulus is attenuated by the application of an additional conditioning noxious stimulus to a remote body region, reflecting diffuse endogenous inhibition [19], [20]. In the present study, pressure pain detection threshold and cold pressure test (see below) were used as test and conditioning stimuli, respectively. An increase in pressure pain detection threshold immediately after cold pressure test was an indication of CPM.
Cold pressor test.
The subjects placed their hand into a container filled with ice water. In order to maximize heterotopic stimulation, the hand contralateral to the side of pressure stimulation was used. The water was regularly mixed to maintain the temperature near to 0°C. The temperature of the water near the hand was monitored by a thermometer with a digital display (±0.1°C). The subjects were asked to keep the hand in the water until they felt an intolerable sensation of pain and were forced to remove the hand from the container, with a maximum time of 2 min.
Pressure stimulation after cold pressor test.
Pressure pain detection threshold was measured again at the same time as the subject was withdrawing the hand from the water (one single measurement). CPM was measured as the difference in pressure pain detection threshold between measurements after and before the cold pressure test.
Cutaneous single electrical stimulation.
Electrical stimulation was performed through electrodes placed distal to the lateral malleolus. A 25 ms, train-of-five, 1 ms, square-wave impulse (perceived as one stimulus), was delivered by a computer-controlled constant current stimulator (Digitimer DS7A, Neurospec, Letchworth Garden City, UK). The current intensity was increased from 1 mA in steps of 0.5 mA until a pain sensation was evoked.
Three assessments were made and the mean of these 3 measurements was used for the data analysis.
Cutaneous repeated (5 stimuli) electrical stimulation.
The stimulus burst used for single stimulus was repeated 5 times at 2 Hz, at constant intensity. The current intensity of the 5 stimuli was increased from 1 mA in steps of 0.5 mA until the subjects felt pain during the last 2–3 of the 5 stimuli (indicating temporal summation) [21].
Three assessments were made and the mean of these 3 measurements was used for the data analysis.
Cutaneous repeated (20 stimuli) electrical stimulation.
The stimulus burst used for single stimulus was repeated for a train of 20 pulses at 2 Hz, at an intensity corresponding to the temporal summation threshold. During this 10 s stimulation, pain intensity was continuously rated by the subject with an electronic VAS, and the area under the curve (AUC) was computed. Additionally, the maximal VAS during stimulation was recorded.
Intramuscular electrical stimulation (1, 5 and 20 stimuli).
A needle was placed in the tibialis anterior muscle, 14 cm distal from the caudal end of the patella and 20 mm in depth. The same single and repeated stimulation patterns and the same procedure described for cutaneous stimulation were used.
Cuff algometry.
A tourniquet cuff was applied to the middle of the leg, at the level of the heads of the gastrocnemius and soleus muscle [22]. The cuff was applied at the side contralateral to the side of electrical stimulation in order not to interfere with the positioning of the intramuscular electrodes. During stimulation, the volunteer rated the pain intensity on an electronic VAS scale. The cuff was inflated with compressed air until VAS 6 was reached [22]. The maximum allowed inflating pressure was 200 kPa. The pressure was maintained for 10 min or until the subjects rated the pain as intolerable. The area under the curve VAS-time was computed. For those subjects who felt intolerable pain before 10 min, the time when the cuff was deflated was recorded and VAS 10 was extrapolated until 10 min. Additionally, the maximal VAS during inflation was recorded.
Psychomotor performance.
The psychomotor performance was assessed by the digit symbol substitution test (DSST), a subscale of the Wechsler Adult Intelligence Scale. The DSST evaluates the ability to concentrate and modifications in processing information [23]. It is a two-minute paper-and-pencil test. The subject was asked to replace digits with corresponding symbols according to a code given on the same sheet of paper. The score consists in the total number and the correct number of symbols drawn.
Timing of the experiment.
The flow of the experiment is illustrated in figure 3. After training, basal values for the pain tests were determined. The study medication was given immediately after the intradermal injection of capsaicin. Thus, basal values were recorded for all tests, except for those related to capsaicin injection. Because of the relatively short duration of hyperalgesia induced by capsaicin, the recordings of basal values before drug administration would yield a too short testing procedure [17].
Horizontal arrow: Testing time; ZZ: Resting time; EX: Medical examination, installation of the testing equipment and training measures; CT: Measures on the area of capsaicin hyperalgesia; ET: Cutaneous electrical stimulation; ED: Intramuscular electrical stimulation, pressure stimulation, cold pressor test, conditioned pain modulation and cuff algometry; BT: Psychomotor performance test; b: Blood sample.
The following tests were performed starting 30, 60, 90 and 120 min after medication: all assessments related to capsaicin injection and all cutaneous electrical pain measurements. Pressure pain, conditioned pain modulation (CPM), intramuscular electrical pain tests, cuff algometry and DSST were performed 120 min after drug intake.
Flumazenil was injected 2∶45 h after intake of the study medication. After 15 min, the whole battery of tests was performed.
The occurrence of any side effect was recorded every 30 min, starting at drug administration.
Four hours after intake of the study medication, the volunteers were discharged after control of blood pressure, heart rate and clinical evaluation of cognitive functions.
Pharmacokinetic and pharmacogenetic investigations
Pharmacokinetic and pharmacogenetic investigations were undertaken on clobazam as the test compound. For pharmacokinetics, blood samples (6 ml) were collected in heparinized tubes at 0.5, 1, 1.5, 2, 4 and 24 hours after drug administration. A specific LC-MS/MS assay was developed and validated in the laboratory of the Division of Clinical Pharmacology and Toxicology of the Geneva University Hospital.
The pharmacogenetic investigation was made in order to control for this factor in case of outlier concentrations of clobazam. It took place 1–4 weeks before the first testing session to evaluate the metabolic activity of cytocrhome P-450 involved in clobazam metabolization.
Phenotype: in vivo activities of 2C19 and 3A4/5 was assessed using micro-doses of omeprazole 2 mg (2C19) and midazolam 0.1 mg (CYP3A4/5), both administered orally at the same time. Two hours after the drugs intake, 2 venous blood samples (6 ml) were collected into heparinized tubes and centrifuged. According to a standardized protocol, the phenotype was determined by calculating the metabolic ratio between deconjugated midazolam and its metabolite 1′-OH-midazolam for CYP3A4 and the ratio of omeprazole and its metabolite OH-omeprazole for CYP2C19 2 h after the drug intake.
Genotype: DNA extracted from blood samples was used for genotype determination. Three single nucleotide polymorphisms (SNP) of the CYP2C19 gene were determined, CYP2C19*2 (c.681G>A), CYP2C19*3 (c.636G>A) and CYP2C19*17 (g.-806C>T) using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) as described previously.
Deviations from study protocol
No important changes to methods or outcomes after the trial commenced were done. The following minor deviations from the study protocol were made.
The body side to which the pain tests were applied was not selected by randomization, as originally planned; we decided to perform most tests on the dominant side in order to place the volunteer with the tested side close to the devices; see section 2.6, “general methodological aspects”.
The skin temperature before application of capsaicin was measured, but not recorded; since the skin temperature in the range of 32°C+/−1 was a condition for performing the experiment and no analysis on this parameter was planed, we did not record the individual temperatures; see section 2.6, “intradermal capsaicin”.
For the cold pressor test, we renounced to record the pain intensity continuously using an electronic VAS, because this measurement was not an outcome and in order to simplify the experiment.
The recordings of the DSST test were limited to 120 min and after flumazenil, in order to simplify the experiment.
Sample size calculation
The primary variable was the area of static hyperalgesia induced by capsaicin injection. The additional measures were secondary variables. No data on the drugs under investigations were available. In a previous study that employed the capsaicin model, pregabalin caused a reduction in the area of hyperalgesia of 10.91 cm2, with a standard deviation of 11.54 [24]. This results in a sample size of 11, adopting a 5% level for statistical significance and an 80% power, two-sided. In a conservative prediction of the effects of the drug that we investigated, we planned to study 16 subjects.
Statistical analysis
The main analysis was a two-way repeated measures ANOVA (SigmaStat 3.5, Systat Software Inc, Chicago, Il, USA), with medication and time as factors of repetition. Non normally distributed data were rank transformed before performing the ANOVA. The analysis tested primarily for the interaction of drug with time (timepoint of the measurements). In case of no significant interaction, we tested the overall effects of the “factors” drug and time. In models with significant interactions, the main effects of the factors drug and time are not interpretable and are therefore not presented. Pairwise comparisons were performed by the Holm-Sidak method.
Because of the high variability of the plasma levels of clobazam (see results), a pharmacokinetic-pharmacodynamic modeling was not feasible. Since the time at which clobazam reached the peak plasma concentration was very variable across subjects, we assumed that its effect might be better detected by considering only the highest drug effect before administration of flumazenil (referred to as “peak value analysis”). We therefore performed a repeated-measure ANOVA by considering as value after medication only the highest drug effect before administration of flumazenil (referred to as “peak value analysis”). This was done for all three drug sessions, in order to keep an unbiased placebo control. For threshold measurements, the peak was the highest value, reflecting the maximum analgesic effect; conversely, for area of hyperalgesia and VAS assessments, the peak analgesic effect corresponded to the lowest value of these parameters.
Finally, we also assumed that including subjects with delayed peak plasma level of clobazam or too low plasma levels during the phase of testing may prevent the detection of drug effects. We therefore repeated the ANOVA analyses after exclusion of subjects with clobazam plasma peaks beyond 2 h after study medication and plasma concentrations lower than 200 µg/ml during the same period (n = 8) (referred to as “subgroup analysis”). The limit of 200 µg/ml was chosen because this is the described mean peak concentration after an oral dose of 10 mg [25].
A p-value<0.05 was considered as statistically significant. Because of the explorative nature of the study, the p-values were not adjusted for multiple comparisons.
Results
The flowchart of the study is displayed in figure 1. In order to obtain 16 complete experiments, we assessed 20 subjects for eligibility, recruited 18 subjects and randomized 17 of them. In all participants capsaicin elicited pain with an intensity of at least 4 (NRS scale). One of these volunteers was excluded from the analysis because his area of hyperalgesia after capsaicin injection was larger than the vectors drawn on the skin, rendering the measure of the primary endpoint impossible. All the remaining 16 tested male subjects completed the study. Their mean age was 26.1 years (SD 4.2) and the BMI was 22.6 kg/m2 (SD 2.6).
The mean values and standard deviations of all performed tests are presented in table 2. For the sake of brevity and to avoid repetition, part of the results of the many statistical analysis are omitted in the following text: here we present the most relevant data, while all statistical analyses are reported in details in table 3. In table 4, an overview of the results is presented.
Area of static hyperalgesia after capsaicin
The pain rating on the VAS-scale immediately after injection of capsaicin was 9.6 (SD 0.7). After five minutes it decreased to 5.8 (SD 1.9). There was no statistically significant difference in the pain rating after among the 3 experimental sessions (p = 0.926 and 0.417 immediately after injection and at 5 min, respectively).
Main analysis.
For the area of static hyperalgesia, there was a statistically significant interaction between “factors” drug and time (p = 0.016). Pairwise multiple comparisons within each drug showed that under the placebo tolterodine a significant increase in the area of hyperalgesia over time was observed (p<0.001), whereas there was no statistically significant increase for clonazepam and clobazam (figure 4). No statistically significant difference among the three drug sessions after flumazenil was detected. Thus, capsaicin injection induced a progressive increase in the area of static hyperalgesia, which was prevented by clobazam and clonazepam.
The area of static hyperalgesia significantly increased with the active placebo tolterodine (p<0.001), but not with clobazam and clonazepam.
Peak value and subgroup analyses.
The peak value analysis did not yield statistically significant results. In the subgroup analysis, there was again a significant interaction between time and drug (p = 0.017). Pairwise multiple comparisons within each drug showed that clonazepam significantly decreases the area of hyperalgesia between the time-point 60 min and the measurement after flumazenil (p = 0.004), indicating that flumazenil could not abolish the clonazepam effect.
Pain intensity after von Frey hair stimulation
In the main analysis, the “factor” drug and interaction between time and drug were not statistically significant for any assessment. The peak value analysis did not yield statistically significant results. In the subgroup analysis, the “factor” interaction of drug with time was significant only for the measurement with 512 mN (p = 0.001). The pairwise multiple comparison within each drug showed an increase of pain rating from the measure at 30 minutes (VAS 3.2, SD 2.8) to the assessment at 120 minutes (VAS 5.1, SD 2.9) with clobazam (p<0.001). This isolated finding is difficult to explain and could be the result of chance.
Pressure stimulation
For both pain detection and tolerance threshold, the “factor” drug and the interaction between drug and time were not statistically significant for any analysis.
Cutaneous single and repeated electrical stimulation (1, 5 and 20 stimuli)
The “factor” drug and the interaction between drug and time were not statistically significant for any analysis.
Intramuscular single electrical stimulation
Main analysis.
The “factors” drug and interaction between drug and time were not statistically significant, whereas there was a significant overall time effect (p = 0.01). Pairwise comparisons within each drug revealed a significant increase in the pain threshold only within clonazepam from baseline (1.84, SD1.23) to the time-point after flumazenil (4.05, SD 3.74) (p<0.001).
Peak value and subgroup analyses.
In the peak value analysis, the “factors” drug and interaction between drug and time were not statistically significant, whereas there was a significant overall time effect (p = 0.015). Pairwise comparisons within each drug revealed clonazepam to significantly increase pain detection thresholds from baseline (9.7, SD 3.3) to peak value (11.0, SD 3.6) (p = 0.032).
The subgroup analysis did not yield statistically significant results.
Intramuscular repeated (5 stimuli) electrical stimulation
Main analysis.
The “factors” drug and interaction between drug and time were not statistically significant, whereas there was a significant overall time effect (p<0.001). The pairwise multiple comparison revealed a significant increase in pain thresholds for clonazepam (p = 0.021) and clobazam (p = 0.021) between baseline measures and 120 minutes, whereas no statistically significant change in pain threshold after tolterodine was observed (figure 5).
There was a statistically significant increase in the temporal summation thresholds for clonazepam (p = 0.021) and clobazam (p = 0.021) between baseline measures and 120 minutes, whereas no statistically significant change in threshold after the active placebo tolterodine was observed.
Peak value and subgroup analyses.
The peak value analysis confirmed the results of the main analysis. Pairwise comparisons within each drug revealed a significant increase in pain detection threshold from baseline to peak value after clonazepam (p = 0.017) and clobazam (p = 0.007), but not after tolterodine.
In the subgroup analysis, the interaction of drug with time (p = 0.042) was statistically significant. The pairwise multiple comparisons showed that clonazepam significantly increased the pain threshold from baseline (1.69 mA, SD 1.23) to 120 minutes (2.93 mA, SD 2.00) (p<0.001), confirming the above results. The difference persisted also after flumazenil administration (3.58 mA, SD 2.77) (p = 0.004).
Intramuscular repeated (20 stimuli) electrical stimulation
In the main and peak value analysis, no statistical significance for the “factor” drug and the interaction between drug and time was observed.
In the subgroup analysis the “factor” drug was statistically significant for the maximal VAS (p = 0.047). In the pairwise comparisons among drugs, clonazepam displayed lower values than tolterodine (p = 0.031) and clobazam (p = 0.031).
Cuff algometry
In the main and peak value analysis, the “factors” drug and the interaction of drug with time were not statistically significant.
In the subgroup analysis, the effect time was significant (p = 0.019), whereas “factors” drug and interaction of drug with time failed to reach statistical significance. In the pairwise multiple comparisons, the AUC after clonazepam was significantly lower than after tolterodine at 120 min (p = 0.024). The AUC after clonazepam significantly decreased from baseline at 120 min (p = 0.003) and after administration of flumazenil (p = 0.007), whereas no significant time effect for the other two drugs was detected.
Psychomotor performance (DSST-Score)
The mean (SD) of the DSST-score for tolterodine, clobazam und clonazepam 120 min after drug administration was 76.5 (11.7), 74.9 (14.7) and 69.3 (15.8), respectively. After flumazenil, the values for tolterodine, clobazam und clonazepam were 81.5 (10.7), 81.1 (12.1) and 79.1 (14.5), respectively.
In the main analysis, the effect time (p<0.001) was statistically significant whereas the effect drug and the interaction of drug with time failed to show statistical significance. In the pairwise comparison within the time-points, clonazepam (but not clobazam) was significantly associated to a lower score than tolterodine at 120 min (p = 0.031).
In the subgroup analysis, the interaction of drug and time (p = 0.04) was statistically significant. The pairwise multiple comparison showed a lower psychomotor performance with clonazepam at 120 minutes, compared with the other two drugs (p = 0.025). The score within clonazepam significantly increased after flumazenil injection (p = 0.001).
No side effect other than impairment of psychomotor performance was observed.
Pharmacokinetic and pharmacogenetic investigations
Table 5 presents the pharmacokinetic parameters of clobazam in all subjects (n = 16) and in the subgroup that included the volunteers with peak plasma levels of clobazam before administration of flumazenil (Tmax beyond 180 minutes) and plasma concentrations of at least 200 µg/ml during the same period (n = 8). Five of remaining subjects had a Tmax longer than 180 minutes, which was the time of flumazenil administration. Three more subjects had a maximal concentration (cmax) below 200 µg/ml at 120 minutes. We determined CYP 2C19 phenotype, and one of the 16 subjects was poor CYP2C19 metabolizer.
Discussion
This study analyzed the anti-hyperalgesic and analgesic effect of two benzodiazepines in a multimodal experimental pain procedure. The ultimate aim was to explore the potential of GABA-agonism for future clinical research. Interactions between time and drug, results of pairwise comparisons within the ANOVA, and analyses that considered the pharmacokinetic variability are suggestive for a possible anti-hyperalgesic or analgesic effect of GABAergic drugs.
Primary endpoint
Capsaicin injection induces secondary hyperalgesia, a phenomenon related to central sensitization [26]. The method has been used in several pharmacological studies. It could detect the effects of different drugs, such as opioids, NMDA-antagonists and anticonvulsants [27]–[29], whereas it did not reveal anti-hyperalgesic effects in some studies on cannabinoids and antidepressants [30], [31]. In another study on cannabinoids, the intensity of capsaicin-induced pain, but not the area of hyperalgesia, detected drug effects [32].
We found a statistically significant interaction of the “factors” drug and time. An increase in the area of static hyperalgesia was observed after administration of placebo, but not after clobazam and clonazepam. Our interpretation is that the benzodiazepines prevented the increase in the area of hyperalgesia, thereby displaying an anti-hyperalgesic effect. This is consistent with the large amount of animal data on GABA-agonists [4], [5].
Secondary endpoints
Results suggestive for an analgesic effect of the benzodiazepines were obtained with all three intramuscular pain models and with cuff algometry. No analgesic effect could be detected with the other pain models employed. The finding that muscle pain models were particularly sensitive to the effects of the benzodiazepines is consistent with a previous investigation on remifentanil: this opioid caused a significantly higher increase in the electrical muscular pain threshold than in the electrical cutaneous pain threshold [33]. The results of these two studies suggest that muscle pain models may be important tools for future pharmacological investigations.
The results with intramuscular stimulation were more robust for the model with repeated (five) stimulations, compared with single stimulation. The 5 repeated stimuli induce temporal summation, a phenomenon related with short-lasting central sensitization [21]. Interestingly, the NMDA-antagonist ketamine increased the pain threshold to repeated stimulation, but had no effect on the pain threshold after single stimulation [21]. This finding is consistent with the established role of NMDA mechanisms in central sensitization [34]. Accordingly, the effects of GABAergic drugs may be better detected with models that induce temporal summation, given the well-know role of GABA-receptors in modulating spinal cord hyperexcitability [3].
Psychomotor performance
Clonazepam impaired psychomotor performance more than clobazam and tolterodine. This effect was not observed anymore after flumazenil. The results confirm previous data that clobazam may be associated with less sedation than clonazepam, at least at the doses investigated [8], [9].
Effect of flumazenil
The effect of clobazam and clonazepam was not influenced by flumazenil in most cases. Occasionally, the difference with placebo persisted after administration of this drug. In our opinion, this is likely the result of a too low dose of flumazenil. This dose was chosen in order to minimize the risk of severe complications, such as seizures, in our population of healthy subjects [14].
Pharmacokinetics and pharmacogenetics
Plasma levels of clobazam exhibited an unexpected high degree of variability. Pharmacogenetic do not explain this variability, since all except one subject were CYP 2C19 extensive metabolizers. The reason for this variability is probably related to delayed absorption. All subjects were tested in the morning and had been instructed to take a very light breakfast at home. Considering the transfer time to the research unit and the additional 2-hours time before drug administration (interview, instruction, training of tests), it is unlikely that our healthy subjects had still some gastric content at the time of drug intake. Delayed absorption, whatever the cause, still remains the most likely explanation for this finding.
Strengths and limitations
This study implied an extensive exploration of pain mechanisms, including hyperalgesia, cutaneous and muscular pain sensitivity, temporal summation, response to different types of sensory stimuli and conditioned pain modulation. The same investigator performed all the experiments. Pharmacokinetic and pharmacogenetic investigations were included. We are not aware of similar comprehensive studies in this area [7].
The study has also limitations. The capsaicin model does not allow more than 2 h time for testing of hyperalgesia [17]. Based on the pharmacokinetic results, an experimental model with a longer time profile may be more sensitive in detecting the anti-hyperalgesic effects of clobazam or possible active metabolites. Our design implied a single dose administration, and the effect of higher, repeated doses or long-term applications remain unclear.
Do benzodiazepines display anti-hyperalgesic and analgesic action?
Research on benzodiazepines in clinical pain conditions is sparse. A systematic review on intrathecal midazolam for postoperative pain found some evidence of efficacy [35]. Another systematic review on the efficacy of clonazepam in neuropathic pain and fibromyalgia could not identify any studies that satisfied the inclusion criteria [36].
We detected an effect of the drugs with part of the tests employed. Thus, a conclusive answer to this question cannot be given. However, in the context of experimental work in rodents and humans, this partial result is not surprising. Failure to detect efficacy with part of the experimental pain modalities is a common finding also with drugs characterized by established clinical efficacy [7], [37]. Different experimental pain modalities likely reflect different dimensions of nociceptive processes, as confirmed by a recent large investigation on healthy subjects [38]. Drugs probably affect these dimensions differentially, leading to different sensitivities of pain models to the analgesic effect of drugs.
We cannot rule out that part of the results were positive by chance. However, some arguments are suggestive for a true anti-hyperalgesic and analgesic effect of the drugs investigated. For the primary endpoint, a significant result was obtained with the main analysis on the interaction of drug with time. The result on the primary endpoint is consistent with the findings of basic research. A further argument is the pattern of response that emerges by a collective analysis of the secondary endpoints, as summarized in table 4. None of the pressure pain modalities revealed an effect of clobazam and clonazepam; none of the tests that employed von Frey hair stimulation detected any effect; and none of the cutaneous electrical pain tests yielded any significant result. Conversely, all the three intramuscular pain tests were able to detect analgesic effects of the benzodiazepines. These effects were consistent with the findings obtained with cuff algometry, a model that probably acts, at least in part, by nociceptive stimulation of the muscles. Because of this cluster pattern, it is unlikely that the significant effects that were detected are the result of chance. Rather, they confirm the finding of the aforementioned study, which found responses to different stimulation modalities to represent distinct dimensions of pain perception [38]. Such dimensions seem to be affected differently by analgesic drugs.
Accordingly, our study provides information for future investigations that will analyze the effect of benzodiazepines. The data indicate that models of deep pain and hyperalgesia are probably most suited.
There was no difference between clobazam and clonazepam regarding the effect on the primary endpoint, i.e. the area of static hyperalgesia. However, clonazepam was associated with stronger impairment of psychomotor performance. This suggests that clobazam may offer potential advantages in the clinical management of pain; further clinical investigations on this compound would be desirable.
Conclusions and perspectives
The present study provides some evidence for an anti-hyperalgesic and analgesic action of agonists at the benzodiazepine-binding site of GABAA receptors. The results are not conclusive, since positive findings were obtained with part of the analyses. Furthermore, the clinical significance of the possible effects of the specific drugs investigated is questionable. Nevertheless, the data support further research in the field of GABA-modulation. GABA-agonists that selectively act at subtypes of GABAA-receptors produce potent antihyperalgesia in animal models, in the absence of sedation [6], [39], [40]. Some subtypes of GABAA receptor modulators were investigated in clinical studies for the treatment of anxiety [41], [42]. Analgesic or antihyperalgesic actions of these selective compounds have not been tested in humans. The perspective holds that, due to the minimally sedating properties of these selective drugs, doses higher than for unselective compounds might be used: the consequent higher receptor occupancy could lead to significant anti-hyperalgesic or analgesic effects in clinical pain conditions. Finally, the results of the present study provide information on which models have to be selected for future investigations: methods that induce hyperalgesia and deep pain may be most promising for detecting the effects of GABAA-agonists.
Acknowledgments
The authors thank Dr. H.U. Zeilhofer, University of Zurich, for his helpful suggestions. The authors thank the Clinical Trial Unit (CTU) of Bern and the Hospital Pharmacy for organizing randomization, blinding and drug preparation.
Author Contributions
Conceived and designed the experiments: MC PV MB JD LA. Performed the experiments: PV. Analyzed the data: PV MC. Contributed reagents/materials/analysis tools: MB JD. Wrote the paper: MC PV. Discussed the results: MC PV MB JD LA. Commented on the manuscript: MC PV MB JD LA.
References
- 1. Woolf CJ, Salter MW (2000) Neuronal plasticity: increasing the gain in pain. Science 288: 1765–1769.
- 2. Curatolo M, Arendt-Nielsen L, Petersen-Felix S (2006) Central hypersensitivity in chronic pain: mechanisms and clinical implications. Phys Med Rehabil Clin N Am 17: 287–302.
- 3. Zeilhofer HU (2005) The glycinergic control of spinal pain processing. Cell Mol Life Sci 62: 2027–2035.
- 4. Knabl J, Zeilhofer UB, Crestani F, Rudolph U, Zeilhofer HU (2009) Genuine antihyperalgesia by systemic diazepam revealed by experiments in GABAA receptor point-mutated mice. Pain 141: 233–238.
- 5. Reichl S, Augustin M, Zahn PK, Pogatzki-Zahn EM (2011) Peripheral and spinal GABAergic regulation of incisional pain in rats. Pain
- 6. Knabl J, Witschi R, Hosl K, Reinold H, Zeilhofer UB, et al. (2008) Reversal of pathological pain through specific spinal GABAA receptor subtypes. Nature 451: 330–334.
- 7. Staahl C, Olesen AE, Andresen T, Arendt-Nielsen L, Drewes AM (2009) Assessing efficacy of non-opioid analgesics in experimental pain models in healthy volunteers: an updated review. Br J Clin Pharmacol 68: 322–341.
- 8. Wildin JD, Pleuvry BJ, Mawer GE, Onon T, Millington L (1990) Respiratory and sedative effects of clobazam and clonazepam in volunteers. Br J Clin Pharmacol 29: 169–177.
- 9. van der Meyden CH, Bartel PR, Sommers DK, Blom M, Pretorius LC (1989) Effects of clobazam and clonazepam on saccadic eye movements and other parameters of psychomotor performance. Eur J Clin Pharmacol 37: 365–369.
- 10. Wiffen P, Collins S, McQuay H, Carroll D, Jadad A, et al. (2005) Anticonvulsant drugs for acute and chronic pain. Cochrane Database of Systematic Reviews 20: CD001133.
- 11. Hugel H, Ellershaw JE, Dickman A (2003) Clonazepam as an adjuvant analgesic in patients with cancer-related neuropathic pain. Journal of Pain and Symptom Management 26: 1073–1074.
- 12. Riss J, Cloyd J, Gates J, Collins S (2008) Benzodiazepines in epilepsy: pharmacology and pharmacokinetics. Acta Neurol Scand 118: 69–86.
- 13. Nilvebrant L, Hallen B, Larsson G (1997) Tolterodine–a new bladder selective muscarinic receptor antagonist: preclinical pharmacological and clinical data. Life Science 60: 1129–1136.
- 14. Whitwam JG, Amrein R (1995) Pharmacology of flumazenil. Acta Anaesthesiol Scand Suppl 108: 3–14.
- 15. Gazerani P, Andersen OK, Arendt-Nielsen L (2007) Site-specific, dose-dependent, and sex-related responses to the experimental pain model induced by intradermal injection of capsaicin to the foreheads and forearms of healthy humans. J Orofac Pain 21: 289–302.
- 16. Scanlon GC, Wallace MS, Ispirescu JS, Schulteis G (2006) Intradermal capsaicin causes dose-dependent pain, allodynia, and hyperalgesia in humans. J Investig Med 54: 238–244.
- 17. Torebjörk HE, Lundberg LE, LaMotte RH (1992) Central changes in processing of mechanoreceptive input in capsaicin-induced secondary hyperalgesia in humans. Journal of Physiology (London) 448: 765–780.
- 18. Witschi R, Punnakkal P, Paul J, Walczak JS, Cervero F, et al. (2011) Presynaptic alpha2-GABAA receptors in primary afferent depolarization and spinal pain control. Journal of Neuroscience 31: 8134–8142.
- 19. Chitour D, Dickenson AH, Le Bars D (1982) Pharmacological evidence for the involvement of serotonergic mechanisms in diffuse noxious inhibitory controls (DNIC). Brain Res 236: 329–337.
- 20. Ge HY, Madeleine P, Arendt-Nielsen L (2004) Sex differences in temporal characteristics of descending inhibitory control: an evaluation using repeated bilateral experimental induction of muscle pain. Pain 110: 72–78.
- 21. Arendt-Nielsen L, Sonnenborg FA, Andersen OK (2000) Facilitation of the withdrawal reflex by repeated transcutaneous electrical stimulation: an experimental study on central integration in humans. European Journal of Applied Physiology 81: 165–173.
- 22. Polianskis R, Graven-Nielsen T, Arendt-Nielsen L (2002) Spatial and temporal aspects of deep tissue pain assessed by cuff algometry. Pain 100: 19–26.
- 23. Hindmarch I (1980) Psychomotor function and psychoactive drugs. Br J Clin Pharmacol 10: 189–209.
- 24. Wang H, Bolognese J, Calder N, Baxendale J, Kehler A, et al. (2008) Effect of morphine and pregabalin compared with diphenhydramine hydrochloride and placebo on hyperalgesia and allodynia induced by intradermal capsaicin in healthy male subjects. J Pain 9: 1088–1095.
- 25. Pullar T, Haigh JR, Peaker S, Feely MP (1987) Pharmacokinetics of N-desmethylclobazam in healthy volunteers and patients with epilepsy. British Journal of Clinical Pharmacology 24: 793–797.
- 26. LaMotte RH, Shain CN, Simone DA, Tsai EF (1991) Neurogenic hyperalgesia: psychophysical studies of underlying mechanisms. Journal of Neurophysiology 66: 190–211.
- 27. Poyhia R, Vainio A (2006) Topically administered ketamine reduces capsaicin-evoked mechanical hyperalgesia. Clin J Pain 22: 32–36.
- 28. Wallace MS, Ridgeway B, Leung A, Schulteis G, Yaksh TL (2002) Concentration-effect relationships for intravenous alfentanil and ketamine infusions in human volunteers: effects on acute thresholds and capsaicin-evoked hyperpathia. Journal of Clinical Pharmacology 42: 70–80.
- 29. Gottrup H, Juhl G, Kristensen AD, Lai R, Chizh BA, et al. (2004) Chronic oral gabapentin reduces elements of central sensitization in human experimental hyperalgesia. Anesthesiology 101: 1400–1408.
- 30. Wallace MS, Barger D, Schulteis G (2002) The effect of chronic oral desipramine on capsaicin-induced allodynia and hyperalgesia: a double-blinded, placebo-controlled, crossover study. Anesthesia and Analgesia 95: 973–978.
- 31. Kraft B, Frickey NA, Kaufmann RM, Reif M, Frey R, et al. (2008) Lack of analgesia by oral standardized cannabis extract on acute inflammatory pain and hyperalgesia in volunteers. Anesthesiology 109: 101–110.
- 32. Wallace M, Schulteis G, Atkinson JH, Wolfson T, Lazzaretto D, et al. (2007) Dose-dependent effects of smoked cannabis on capsaicin-induced pain and hyperalgesia in healthy volunteers. Anesthesiology 107: 785–796.
- 33. Curatolo M, Petersen-Felix S, Gerber A, Arendt-Nielsen L (2000) Remifentanil inhibits muscular more than cutaneous pain in humans. British Journal of Anaesthesia 85: 529–532.
- 34. Dickenson AH, Sullivan AF (1987) Evidence for a role of the NMDA receptor in the frequency dependent potentiation of deep rat dorsal horn nociceptive neurones following C fibre stimulation. Neuropharmacology 26: 1235–1238.
- 35. Ho KM, Ismail H (2008) Use of intrathecal midazolam to improve perioperative analgesia: a meta-analysis. Anaesthesia and Intensive Care 36: 365–373.
- 36. Corrigan R, Derry S, Wiffen PJ, Moore RA (2012) Clonazepam for neuropathic pain and fibromyalgia in adults. Cochrane Database of Systematic Reviews 5: CD009486.
- 37. Staahl C, Olesen AE, Andresen T, Arendt-Nielsen L, Drewes AM (2009) Assessing analgesic actions of opioids by experimental pain models in healthy volunteers - an updated review. Br J Clin Pharmacol 68: 149–168.
- 38. Neziri AY, Curatolo M, Nuesch E, Scaramozzino P, Andersen OK, et al. (2011) Factor analysis of responses to thermal, electrical, and mechanical painful stimuli supports the importance of multi-modal pain assessment. Pain 152: 1146–1155.
- 39. Di Lio A, Benke D, Besson M, Desmeules J, Daali Y, et al. (2011) HZ166, a novel GABAA receptor subtype-selective benzodiazepine site ligand, is antihyperalgesic in mouse models of inflammatory and neuropathic pain. Neuropharmacology 60: 626–632.
- 40. Zeilhofer HU, Mohler H, Di Lio A (2009) GABAergic analgesia: new insights from mutant mice and subtype-selective agonists. Trends in Pharmacological Sciences 30: 397–402.
- 41. Rudolph U, Knoflach F (2011) Beyond classical benzodiazepines: novel therapeutic potential of GABA(A) receptor subtypes. Nature Reviews Drug Discovery 10: 685–697.
- 42. de Haas SL, de Visser SJ, van der Post JP, de Smet M, Schoemaker RC, et al. (2007) Pharmacodynamic and pharmacokinetic effects of TPA023, a GABA(A) alpha(2,3) subtype-selective agonist, compared to lorazepam and placebo in healthy volunteers. Journal of Psychopharmacology 21: 374–383.