Evaluation of Anti-Hyperalgesic and Analgesic Effects of Two Benzodiazepines in Human Experimental Pain: A Randomized Placebo-Controlled Study

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. Trial Registration ClinicalTrials.gov NCT01011036


INTRODUCTION BACKGROUND
Neuropathic and nociceptive pain conditions are associated with plastic changes of the central nervous system [25]. These changes alter the processing of the nociceptive input, thereby leading to reduced pain thresholds to sensory stimuli, enhanced pain after suprathreshold stimulation and enlargement of the areas of local and referred pain [5]. These phenomena have clinical relevance, in that they contribute to pain and disability in patients suffering from chronic pain.
An important aspect of neuroplastic changes in inflammatory and neuropathic conditions is the reduction in inhibitory glycinergic and GABAergic control at dorsal horn neurons [26].
Such dis-inhibitory processes involve reduced responsiveness of certain glycine receptors and changes in the transmembrane chloride gradient of dorsal horn neurons [26].
Potentiation of GABA A receptor-mediated synaptic inhibition by benzodiazepines reverses pathologically increased pain sensitivity in animal studies [16]. Subtype-selective compounds targeting alpha2 and/or alpha3 subunit containing GABA A receptors [15] produce antinociception and antihyperalgesia in mice and rats without sedation and without tolerance induction [15]. These findings open new perspectives for a more selective targeting of pain pathways with GABAergic drugs.
Methods that explore specific mechanisms of pain in humans are today available [1]. Such models can be used to assess the effect of drugs on the different mechanisms related to central pain modulation. To date, little is known on the effects of GABA A receptor targeting 18.09.2009 3 drugs on nociceptive processes in humans. This information is important to evaluate the potential clinical usefulness of these compounds in pain management.
The benzodiazepine that is most commonly prescribed in pain treatment is clonazepam.
However, sedation strongly limits its clinical usefulness. Clobazam is another GABAergic compound that may cause less sedation than clonazepam [22; 24]. To our knowledge, its effects on pain processing have not been evaluated.

OBJECTIVE
To evaluate the effect of the GABA A -agonist clobazam on different mechanisms of pain processing by means of a multimodal experimental testing procedure in healthy volunteers.
The ultimate aims are twofold. First, we expect information on the potential clinical usefulness of clobazam in the clinical management of pain conditions. Second, compounds acting on subunits of the GABA-receptor complex that may lead to antinociception with no or minimal sedation are becoming available for human research [6]. This may change profoundly the role of GABA-agonists in pain management. The results of the present investigation will provide information on the pain mechanisms that are affected by GABAergic modulation for future investigations on more selective compounds. 18

DESIGN
This is an exploratory, phase II study with a randomized double-blind crossover design, using pain assessment methods that explore different nociceptive mechanisms. In this 3 arms study, clobazam 20 mg will be compared with clonazepam 1 mg (positive control) and tolterodine 1 mg (active placebo). At the end of each session the benzodiazepine antagonist flumazenil 0.2 mg will be administered to evaluate whether the observed effects are GABAmediated.

SETTING
Department of Anaesthesiology and Pain Therapy, University Hospital, Inselspital Bern,

Switzerland. Pharmacokinetic and genetic investigations at the Division of Clinical
Pharmacology and Toxicology, University Hospital of Geneva.

SUBJECTS
Sixteen healthy volunteers will be tested after obtaining written informed consent. • Current or past history of drug or alcohol abuse.
• Intake of any psychotropic drug currently or in the last month or chronic alcohol intake • Any concomitant illness • Current and regular intake of any drugs that might affect nociception • If an intercurrent illness necessitating a drug treatment occurs during the assay, a wash-out period of at least 4 half-lives will be left before the next session • If a volunteer withdraws from the assay ("drop-out") the values obtained hitherto will not be used and another volunteer will be enrolled.
WITHDRAWAL CRITERIA Volunteers will be informed that they can withdraw at any time without having to give an explanation.
In case of a short-time new diagnosed illness necessitating a drug treatment during the study (upper respiratory tract infection for example), a wash-out period of at least 4 half-lives will be left before the next session.

18.09.2009
If needed occasionally during the study, short half-life rescue analgesics (paracetamol, ibuprofen) will be allowed but have to be stopped at least 48 hour before the next experimental session.
In case of withdrawal, another volunteer will be enrolled and the data obtained will not be used in the analysis.

SELECTION AND RECRUITMENT PROCEDURES
Volunteers will be recruited by advertisement at the Inselspital and at the University of Bern. Only Caucasians will be included to minimize the genetic variations in the CYP2C19 gene.

Pharmacodynamic
Clobazam is a benzodiazepine prescribed in adults and children for all forms of anxiety and as an adjuvant in the treatment of epilepsy. Unlike 2 mg clonazepam, clobazam at the dose of 20 mg did not affect significantly cognitive and psychomotor functions, probably because of its unusual 1-5 chemical structure [22].

Choice of the dosage
The mean recommended dose in "The Swiss Compendium of Drugs" is 0.5/mg/kg daily, clobazam has been shown to be less sedative than 1 mg of clonazepam [24]. Therefore these doses have been chosen for the present study.

Safety profile
Like any benzodiazepine, the most common side effects of clobazam are drowsiness, sedation and hangover effect, which are dose-dependent. Sedation occurred in 38% of clobazam-treated patients and 44% of diazepam-treated patients in a placebo-controlled, double-blind trial on 114 outpatients with anxiety 1 . In this study, patients received mean doses of 59 mg clobazam daily and 25 mg diazepam daily. Dizziness occurred in 11% and 2% of diazepam-and clobazam-treated patients. Considering these data, a single dose of 20 mg of clobazam should be well tolerated.
Others common side effects of benzodiazepines such as tolerance and withdrawal syndrome are associated with chronic use and should not be present after a single dose. Finally the risk of respiratory depression is negligible at this dose and in absence of concomitant respiratory pathology [24].

Pharmacokinetics
Clobazam is well absorbed with a 90% biovailability. Maximum plasma concentration is reached within about 2 h (1.3 ± 0.2h after 10 mg and 1.6h ±0.3h after 20 mg in healthy volunteers). Clobazam is mainly metabolized in the liver through the cytochrome P450 enzyme system, notably into N-desmethylclobazam which is an active metabolite [10]. Considering its high bioavailability, the drug will be administered by mouth. Due to the long t 1/2 of the parent compound and of the metabolite, each session will take place 2 weeks apart.

Pharmacogenetics
CYP2C19 is polymorphic and people carrying the so-called *2 variant (a single base pair substitution in exon 5 of CYP2C19) have a truncated non functional protein. The activity of CYP2C19 is abolished in homozygous persons and decreased in heterozygous persons [17].
It has been show that the metabolic ratio N-desmethylclobazam/clobazam is higher in people carrying the CYP2C19*2 variant and correlated with more intense adverse effects such as sedation and dizziness [10; 17]. The N-desmethylclobazam/clobazam ratio is also increased after administration of a CYP3A4 inducer such as rifampicine, but to a lesser extend [17].

Pharmacodynamic
Clonazepam is a benzodiazepine prescribed in adults and in children in all forms of epilepsy. Moreover, clonazepam is widely used in practice to treat neuropathic pain and has been demonstrated efficient in myofascial pain [7], temporomadibular joint dysfunction [12], cancer related neuropathic pain [14] and in stomatodynia when used topically [11].

18.09.2009
Because of its clinical application in pain medicine, clonazepam is used here as a positive control.

Choice of the dosage.
Doses prescribed in chronic neuropathic pain vary between 0.5 and 2 mg, 1 mg being the average dosage. Based on the starting doses recommended in epilepsy (Swiss Compendium, Documed 2009) clonazepam 1 mg should be as efficient as clobazam 20 mg in terms of anti-hyperalgesic properties.

Safety profile
Clonazepam has the same side effect profile than clobazam with sedation and drowsiness being the most expected side effect. In patients treated for panic disorder, approximately 37% experienced somnolence and 8% experienced dizziness 1 .
As for clobazam, others common side effects of benzodiazepines such as tolerance and withdrawal syndrome are associated with chronic use and should not be present after a single dose. Finally the risk of respiratory depression is negligible at this dose and in absence of concomitant respiratory pathology [24].

Pharmacokinetic
Clonazeapm is well absorbed with a 90% biovailability. It is mainly metabolized in the liver trough the cytochrome P450 enzymes system, notably by the CYP3A4. Clonazepam is then excreted by the kidney and the t 1/2 is 20 to 40 hours 2 .

Pharmacodynamic
Tolterodine is an anticholinergic compound prescribed in hyperactive bladder.
Anticholinergic compounds usually cause some sedation and dry mouth, but are devoid of analgesic effects to the state of our knowledge. Therefore, tolterodine has been chosen as an active placebo.

Choice of the dosage
The starting dose usually recommended in the Swiss Compendium is 2 mg twice a day, which can be decreased to 1 mg twice a day. For tolerance reasons a dose of 1 mg will be given during this study.

Safety profile
The main side effects expected are tachycardia, constipation (6 to 7 %), xerostomia (15 to 30%), dizziness (2%) and somnolence (3%) 1 . These effects are however dose-dependent and should be self-limited after the low single dose chosen.

Pharmacokinetic
Tolterodine is well absorbed with a 77% oral bioavailability. It is metabolised in the liver through CYP2D6, which is polymorphic. In the so-called poor metabolisers, who have a reduced or an abolished activity of CYP2D6, CYP 3A4 is mainly involved in the metabolism of tolterodine. However t 1/2 of tolterodine is doubled in these people after a single dose. Elimination t 1/2 is 1.9 to 3.7 hours 2 .

Choice of the dosage
To avoid a risk of seizure, a dose of 0.2 mg iv will be given.

Safety profile
The main adverse effect of flumazenil is a risk of seizure. However it is essentially seen in patients who are relying on benzodiazepines to control seizures, are physically dependent on benzodiazepines, or who have ingested large doses of other drugs. It should not appear with a single dose of 0.2 mg in benzodiazepines naïve healthy volunteers 2 .

Pharmacokinetic
Flumazenil iv is extensively metabolised in the liver with a t 1/2 of about 50 minutes.

RANDOMIZATION, DRUG PREPARATION AND BLINDING
The subjects will receive the study medications in a randomized order. A computergenerated random list will be prepared by the hospital pharmacy and will be known only by the hospital pharmacy and by an investigator who will not participate in the experiments. 1  The drugs will be enclosed by the hospital pharmacy in the same type of capsule to assure blinding. The pharmacy will deliver the blinded capsules to the investigators, numbered according to the defined randomization order.

ENDPOINTS
The mechanisms investigated by the different tests are listed in table 1. The primary endpoint is the area of static hyperalgesia induced by intradermal capsaicin injection. The additional measures will be secondary endpoints. Before starting the experiment, training sessions of the pain tests will be performed until the subjects will be familiar with the testing procedures.
The body side to which tests are applied and the order of application of the pain tests will be defined before starting the recruitment using a computer-generated randomization. Each subject will be tested on the same side in all three experimental sessions.
Volunteers are not allowed to see the area tested or see any read-outs from any instruments.

INTRADERMAL CAPSAICIN
Several studies have used intradermal capsaicin, a hot tasting ingredient of certain pepper plants, for human experimental pain studies [8; 20]. 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 (healthy) skin. Secondary hyperalgesia after intradermal capsaicin is the result of central sensitization [21] and is sensitive to the action of GABAergic compounds in animal studies [16].

Preparation of capsaicin solution
The capsaicin solution will be prepared by the Hospital pharmacy and will undergo sterile filtration into sterile septum-sealed vials. The sterile solution (1 mg/mL) will be placed into sterile syringes and used for intradermal injection at ambient temperature.

Preparation of subjects
Subjects will remain semi-recumbent throughout the experiments and their arm will be held horizontally, supported by a pillow throughout the assessment. Skin temperature will be standardised to 32°C +/-1°C prior to capsaicin injection and recorded.
Within 2 h before capsaicin injection, the skin of the forearm will be cleaned using an antiseptic wipe and allowed to air dry. The injection site will be marked midway between the elbow and the wrist. Using a surgical skin marker, 4 lines will be drawn through the injection site, so that they intersect the corners of a regular octagon. These lines are marked every 0.5 cm from the center to the edges using a predefined grid and the individual points will be given a number from 1 to 10.
Using a tuberculin syringe (1 ml), fitted with 27-gauge disposable needle, 100 μL capsaicin will be injected epidermally into the skin, at the intersection of the 4 vectors previously drawn. A white wheal should appear during injection ensuring correct injection.
After injection the surrounding skin area becomes hyperalgesic to punctate mechanical stimuli (static hyperalgesia) and to stroking tactile stimulation (dynamic hyperalgesia).
Static mechanical hyperalgesia is judged to be present when a subject indicates that applying a punctuated probe within 5 cm of the injection site elicits a painful perception as compared to the non-painful, pricking perception elicited by the same probe at the contralateral control site. Only these subjects will be included in the study.

Assessment of injection pain
In order to document pain induction, the subject will record pain intensity from injection to 5 minutes post-injection or until the pain intensity returns to zero using a 10 cm electronic visual analogue pain scale (VAS), whereby 0 = no pain and 10 = worst pain imaginable.

Assessment of hyperalgesia
The area of secondary hyperalgesia will be assessed using a calibrated (60 g) von Frey hair.
The punctuated probe will be 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 is placed gently on the skin and the load is applied for 2 seconds.
The volunteers are asked to report when the sensation of the pricking changes to a "different", "unpleasant" or "burning pain" sensation. More than 4 seconds will elapse between consecutive stimuli in steps of 0.5 cm. This procedure will be repeated for each vector. The number corresponding to the marker at which sensation changes as described above will be noted and the sum of the numbers will be used as a proxy for the area.
Additionally, once complete boundaries had been defined, the areas are transcribed onto clear film and weighted.
The area of dynamic mechanical hyperalgesia is determined by gently stroking a hand-held cotton wool tip on the skin at a rate of approximately 1 cm/s. The borders of dynamic hyperalgesia are delineated similarly to the determination of static hyperalgesia. Subjects are asked to report when the sensation changes from a non-painful sensation to a painful/unpleasant sensation by stroking the skin with the cotton wool tip.

Assessment of stimulus-response function
Mechanical pain sensitivity will be assessed using a set of seven weighted pinprick stimuli to obtain a stimulus-response function. The test will be applied 1 cm inside the outer boarder of the pinprick hyperalgesic area, using a set of 7 custom-made graded von Frey hair mechanical stimulators with fixed stimulus intensities. The flat contact area of the stimulators is 0.2 mm diameter, and the 7 stimulators exert forces of 8, 16, 32, 64, 128, 256 and 512 mN. The order of stimulation will be defined randomly by computer. The stimulators will be applied at a rate of 2 s on, 2 s off. Subjects will be asked to give a pain rating for each stimulus on a 10 cm VAS. Two assessments for each stimulation will be made and the mean of these 2 measurements will be used for the data analysis.

PRESSURE PAIN
Pain detection and tolerance thresholds will be measured with an electronic pressure algometer [3] ( Algometer, Somedic) applied at the center of the pulp of the 2 nd toe. The probe has a surface area of 1 cm 2 . The pressure is increased from 0 at a rate of 30 kPa/s to a maximum pressure of 1200 kPa. Pain detection threshold is defined as the point at which the pressure sensation turns to pain. Pain tolerance threshold is defined as the point at which the subject feels pain as intolerable. The subjects are instructed to press a button when these points are reached. The algometer displays the pressure intensity at which the button is pressed. If the subjects do not press the button at a pressure of 1200 kPa, this value is considered as threshold. Three assessments will be made and the mean of these 3 measurements will be used for the data analysis.
For the assessment of the stimulus-response function, stimuli of 1. each stimulation, the volunteer rates the pain intensity on a 10 cm VAS. Two assessments for each stimulation will be made and the mean of these 2 measurements will be used for the data analysis.

DIFFUSE NOXIOUS INHIBITORY CONTROL (DNIC)
This method explores the endogenous modulation of nociceptive input. Under normal conditions, pain after application of a phasic nociceptive stimulus is attenuated by the application of an additional tonic noxious stimulus to a remote body region, reflecting diffuse endogenous inhibition [4; 9]. In the present study, pressure pain detection threshold and cold pressure test are used as phasic and tonic stimuli, respectively. An increase in pressure pain detection threshold immediately after cold pressure test is an indication of DNIC.

Cold pressor test
The device consists of a container separated in an outer and an inner part by a mesh screen.
The mesh screen prevents direct contact between the ice (placed in the outer part) and the hand of the subject (placed in the inner part). The outer part of the container is filled with ice that maintains the water on the inner side close to 0°C. The water is regularly mixed to maintain the temperature near to 0°C. The temperature of the water near the hand is monitored by a thermometer with a digital display (±0.1°C).
The subject places his hand, wide open and to the wrist, in the inner part of the container.
He is asked to keep it in the water until he feels an intolerable sensation of pain and is forced to remove his hand from the container, with a cut-off time of 2 min.

18.09.2009
Pain intensity will be continuously rated by the subject with an electronic VAS. The area under the pain intensity/time curve will be determined. If the hand is withdrawn before the end of the 2 min because of intolerable pain, the pain intensity will be considered to be 10 until the end of the 2 min period.

Pressure pain
Pressure pain detection threshold is measured again at the same time as the subject is withdrawing the hand from the water (one single measurement). DNIC is measured as the difference in pressure pain detection threshold between measurements after and before the cold pressure test.

CUTANEOUS SINGLE ELECTRICAL STIMULATION
Electrical stimulation will be performed through electrodes placed distal to the lateral malleolus. A 25 ms, train-of-five, 1 ms, square-wave impulse (perceived as one stimulus), will be delivered by a computer-controlled constant current stimulator (Digimeter, Neurospec). The current intensity will be increased from 1 mA in steps of 0.5 mA until a pain sensation will be evoked.
Three assessments will be made and the mean of these 3 measurements will be used for the data analysis. Three assessments will be made and the mean of these 3 measurements will be used for the data analysis.

CUTANEOUS REPEATED (20 STIMULI) ELECTRICAL STIMULATION
The stimulus burst used for single stimulus is repeated for a train of 20 pulses at 2 Hz, at an intensity corresponding to the temporal summation threshold (Digimeter, Neurospec).
Previous investigations using reflex assessments indicate that pain intensity may decrease during the stimulation [2], suggesting that this model may investigate endogenous pain modulation. During this 10 s stimulation, pain intensity will be continuously rated by the subject with an electronic VAS.

INTRAMUSCULAR ELECTRICAL STIMULATION (SINGLE AND REPEATED)
A needle is 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 described above are used. Three assessments for each stimulation modality will be made and the mean of these 3 measurements will be used for the data analysis.

CUFF ALGOMETRY
To explore response to an ongoing tonic deep-tissue painful stimulus, a tourniquet cuff is applied to the middle of the leg, at the level of the heads of the gastrocnemius and soleus muscle [19], at the side contralateral to the side of electrical stimulation. During the period of stimulation, the volunteer rates the pain intensity on an electronic VAS scale. The cuff will be inflated with compressed air until VAS 6 is reached [19]. The maximum allowed 18.09.2009 20 inflating pressure will be 200 kPa. The pressure will be maintained for 10 min or until the subjects will rate the pain as intolerable. Using this paradigm, part of the subjects experience a reduction in VAS during the stimulation [18], suggesting that the model may explore endogenous pain modulation.
The area under the curve VAS-time is computed. For those subjects who feel intolerable pain before 10 min, the time when the cuff is deflated is recorded and VAS 10 is extrapolated until 10 min for the calculation of the area under the curve.

SIDE EFFECTS
The psychomotor performance will be 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 information processing performance [13]. It is a two-minute paper-and-pencil test. The subject is required 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. Different versions of the test (i.e., different symbol-digit codes) will be used at each assessment.
The subjective degree of sedation will evaluated with a VAS, whereby 0 = not sleepy at all and 10 = extremely sleepy.
Any other side effect will be recorded.

PHENOTYPING
The investigation will take place once, 1-4 weeks before the first testing session.
In vivo activities of 2C19 and 3A4/5 will be assessed using micro-doses of omeprazole 2 mg (2C19) and midazolam 0.1 mg (CYP3A4/5), both administered orally at the same time. Significant effects (positive or negative) are not expected at such low doses. Two hours after the drugs intake, 2 venous blood samples (6 ml) are collected into heparinized tubes and centrifuged. According to a standardized protocol, the phenotype is determined by calculating the metabolic ratio between deconjugated midazolam and its metabolite 1'-OH-midazolam for CYP3A4 and the ratio of omeprazole and its metabolie OH-omeprazole for CYP2C19 2h after the drug intake. • DSST.
• Discharge after control of blood pressure, heart rate and clinical evaluation of cognitive functions.

SAMPLE SIZE CONSIDERATIONS -STATISTICAL ANALYSIS
The primary variable is the area of static hyperalgesia induced by capsaicin injection. The additional measures will be exploratory secondary variables. No data on the drugs under investigations are available for a sample size calculation. In a previous study that employed the capsaicin model, pregabalin caused a reduction in the area of hyperalgesia of 10.91 cm 2 , with a standard deviation of 11.54 [23]. This results in a sample size of 11, adopting a 5% level for statistical significance and an 80% power. In a conservative prediction of the effects of the drug that we investigate, we will study 16 subjects.
The data will be analyzed by two-way repeated measures ANOVA. 18.09.2009

RELEVANCE
In clinical practice benzodiazepines, and clonazepam in particular, are used as co-analgesics in chronic pain conditions. Their use is based on few randomized trials and mostly on an empirical basis. Sedation, memory impairment and tolerance development strongly limit the clinical usefulness of these compounds in pain therapy. The particular properties of clobazam, which seems to be relatively free of sedative properties, should offer an opportunity to modulate GABAergic pain transmission in humans with a smaller potential for sedation.
There is a need for developing new treatment strategies that modulate the plasticity changes associated with clinical pain conditions. Subtype-selective GABAergic drugs have potential important clinical applications in that they may attenuate such plasticity changes with minimal or no sedation. Such compounds may become available for human pain research in the near future.
No mechanistic study has profiled the action of drugs acting on GABA-A receptors in humans, which significantly impairs our understanding of the full potential of such drugs in pain management. The planned study would provide insights into the efficacy of these drugs on different mechanisms related to pain processing and therefore contribute to clarify their The principles established in the GCP guidelines apply to our study.

RISKS
Potential risks of the capsaicin model include temporary burning or stinging sensation at the site of capsaicin injection, irritation, redness, blistering, or severe and persistent burning at the application site. Rarely severe allergic reactions up to anaphylactic shock may occur. An anesthesiologist familiar with reanimation will be present during the experiment.
Possible risk of intramuscular electrical stimulation is local infection.
All other tests are not expected to cause any damage.
General remarks concerning safety: In case of adverse drug reaction the healthy volunteers will stay and be treated in the University Hospital of Bern as long as needed. Any unexpected adverse drug reaction will be notified to the Swissmedic.

INDEPENDENT DATA MONITORING COMMITTEE
The data quality and control will be reviewed by Dr. Andreas Siegenthaler, MD, Department of Anesthesiology and Pain Therapy, University Hospital of Bern. Controls will be made in intervals of 5 healthy volunteers.

DATA STORAGE
The data will be stored in a locked-up drawer for 5 years in the University Hospital of Bern.

FINANCING AND INSURANCE
This study is sponsored by a grant from the Swiss National Science Foundation (SPUM no.

33CM30_124117)
The public liability of the investigators will be covered by the insurance of the Inselspital Bern.

TIME SCHEDULE
This is a 24 months study, starting in October 2009. Volunteers will enter a 6 weeks double blind, placebo controlled, three arms, randomised period. The last subject will be included in October 2011.

TASKS OF AUTHORS
• Study design: all the authors.
• Draft of the study protocol: Michele Curatolo.
• Instruction of Pascal Vuilleumier in the test procedures: Alban Neziri and Michele Curatolo.
• Interpretation of findings: all authors.
• Final version of article: all authors.