Pharmacodynamic interactions of a solid formulation of sodium oxybate and ethanol in healthy volunteers

Abstract

Aim

The pharmacologic effects of sodium oxybate (SO) have a number of similarities with those of alcohol. This study evaluated the pharmacodynamic interaction of SMO.IR (a solid immediate release formulation of SO) and alcohol (0.7 (males) or 0.57 (females) g kg^–1 alcohol using 40% vodka).

Methods

In a randomized, double-blind, double-dummy, crossover trial, 24 healthy volunteers received randomly a) 2.25 g SMO.IR and placebo alcohol preparation, b) 2.25 g f SMO.IR and alcohol, c) 2.25 g SMO.IR matching placebo and alcohol and d) 2.25 g of SMO.IR matching placebo and placebo alcohol preparation. Objective and subjective cognitive parameters, adverse events and vital signs were assessed before, 15 and 165 min after treatment administration.

Results

Alcohol produced the expected cognitive impairment and the expected subjective sedation rapidly after intake (from 15 min). The objective effects of SMO.IR were much less pronounced than those of alcohol. The reverse was observed for subjective complaints, which were related to lesser stimulation and greater sedation. Nevertheless, 165 min after administration this sedation feeling was less with SMO.IR than with alcohol. There was a significant interaction between SMO.IR and alcohol at 15 min (i.e. increase in alertness and stimulation and decrease in sedation). In addition, an isolated mild decrease in digit vigilance accuracy occurred at 165 min post-dose after the combination. The co-administration of SMO.IR and alcohol was safe and well-tolerated.

Conclusion

SMO.IR and alcohol have distinct adverse effect profiles. The objective effects of SMO.IR are much less marked than those of alcohol. No deleterious interaction was observed.

WHAT IS ALREADY KNOWN ABOUT THIS SUBJECT

• Sodium oxybate (SO) has been approved for the treatment of alcohol dependence in Europe. Post-market surveillance has found no significant safety issues in alcohol relapsing patients taking SO, even though SO and alcohol share several pharmacological targets and effects.

• Illicit SO (also called ‘street GHB’ or ‘liquid ecstasy’) can lead to abuse, misuse, addiction, criminal use and/or overdosing, and may have long term neurotoxic effects.

• SMO.IR, a solid immediate-release formulation of sodium oxybate, is being developed to reduce the risk of criminal or accidental misuse of the drug and to obtain worldwide approval for the treatment of alcohol dependence.

WHAT THIS STUDY ADDS

• SMO.IR and alcohol have distinct adverse effect profiles. The objective sedative effects of SMO.IR are much less marked than those of alcohol and no interaction was observed.

• The dose of SMO.IR used in this study (2.25 g) is in the upper range of the therapeutic interval for alcohol dependence and is 30%–80% higher than commonly prescribed doses.

• The low level of interaction between alcohol and SMO.IR may account for the good safety profile of the drug even in the context of alcohol relapses.

Introduction

Sodium oxybate (SO) is the name of the sodium salt of gamma-hydroxybutyric acid when used as a therapeutic agent. Gamma-hydroxybutyric/gammabutyrate (GHB) is an endogenous short chain fatty acid, which can be found in several brain areas. GHB is closely related to gamma-aminobutyric acid (GABA) and, like GABA, is a cerebral inhibitory neurotransmitter [1,2]. GHB has a GABA-ergic effect through its metabolic transformation into GABA and as an agonist of GABA-B receptors. More recently, several GABA-A receptor subunits were also identified as possible target candidates for GHB binding (see [3] for a review). More precisely, the α4βδ GABA-A receptor appears to be a relevant candidate for the elusive GHB-receptor [4]. GHB has a high affinity for GHB receptors which do not recognize GABA. GHB also has an indirect effect on dopaminergic, serotonergic, glutamatergic and opioidergic neural pathways [5,6].

SO is approved in a variety of European and American countries as an intravenous anaesthetic [7,8] and as a liquid oral medication for patients with narcolepsy-cataplexy [9–11] or alcohol dependence [12,13].

It has been well-demonstrated in animal and human studies that SO reduces alcohol consumption and suppresses alcohol withdrawal syndrome [13,14]. The pharmacological effects of SO present several similarities with those of alcohol [15–17]. Thus, these alcohol-mimicking effects on the central nervous system are of particular interest in alcohol use disorders (AUDs) [18].

AUDs include alcohol withdrawal syndrome, alcohol abuse and alcohol dependence [19]. AUDs are very common in all developed countries and in numerous emergent countries [20]. AUDs have been associated with severe somatic and psychiatric problems, including liver cirrhosis, hepatocarcinoma, myocardial toxicity and neuropsychiatric dysfunctions. Alcoholism results in an increase in morbidity and mortality, a reduced quality of life at the individual level and a variety of societal, public health and illegal consequences. Therapeutic approaches toward AUDs combine various individual and group psychotherapies, pharmacotherapy aiming at facilitating alcohol withdrawal, reducing alcohol craving and contributing to the maintenance of abstinence, psychosocial interventions and patient associations [21,22].

In several randomized, placebo- or active comparator-controlled clinical studies, SO has consistently shown efficacy in the treatment of acute alcohol withdrawal syndrome [6,23,24], in the maintenance of long term alcohol abstinence [25–27], and in the management of treatment-resistant chronic alcoholics [28]. No serious drug-related effects have been reported for SO in clinical studies [12,29]. Some side effects, e.g. dizziness, sleepiness and tiredness, usually of mild intensity, were significantly more frequent in subjects on SO than in those on placebo. These adverse effects usually occurred in the 3–4 first weeks of treatment and resolved thereafter without a need for study drug withdrawal [24,30].

In the past 10 years, a growing body of literature has focused on the abuse, misuse and criminal uses of GHB [31–33]. Illicit GHB can lead to abuse, misuse, addiction, criminal use and/or overdosing, and may have long term neurotoxic effects [5,34]. For this reason, there has been some concern about the use of this medication in an addict population. However, in the countries where SO has been approved for the treatment of alcohol dependence for a number of years, SO addiction has been infrequent and readily manageable among alcoholics [13]. SO withdrawal syndrome is rare as well as rapidly and completely responsive to an oral benzodiazepine treatment [35,36].

The present study evaluated a solid immediate release formulation of sodium oxybate (SMO.IR). SMO.IR is being developed to provide physicians and patients with a solid oral formulation, which might reduce the risk of abuse/misuse as well as the criminal use of the drug, and to obtain worldwide approval for the treatment of alcohol dependence and narcolepsy-cataplexy.

This new formulation was previously tested in an open, randomized, crossover study aimed at defining the PK profile of a single oral dose (2.25 g) of SMO.IR [37]. Results suggested that SMO.IR was rapidly absorbed (t_max = 30 min) and eliminated (t_1/2 = 34 min). In a subsequent single dose (2.25 g) PK study carried out in 46 male and female healthy volunteers, SMO.IR was found to be bioequivalent to the same dose of the liquid formulation of SO which is approved for alcohol dependence (Alcover®, CT Farmaceutico, Sanremo, Italy) [38].

Because both SO and alcohol act centrally and interact with the GABA-ergic system, a pharmacodynamic interaction might occur in the CNS as a result of additive, synergistic or antagonistic effects when these two drugs are taken together. Such interaction might result in potentially serious adverse reactions in relapsing alcohol-dependent patients, although such adverse reactions have not been reported in countries where SO has been approved for the treatment of alcohol dependence. Therefore, this exploratory study was conducted, in healthy volunteers without a history of drug and/or alcohol abuse, to assess the possibility for potential CNS interactions between alcohol and SMO.IR. The main objective was to compare the pharmacodynamic effects of a single dose of SMO.IR, of alcohol and their potential pharmacodynamic interaction. To evaluate the safety of a single dose administration of SMO.IR, of alcohol and of the combination of SMO.IR and alcohol was the secondary objective.

Methods

This was a randomized, double-blind, double-dummy, four-way crossover, placebo-controlled study conducted at a single clinical pharmacology centre (BIOTRIAL, Rennes, France) according to the ethical principles stated in the revised version of the Declaration of Helsinki (WMA, 2008), after approval by an Independent Ethics Committee (CPP Ouest VI, Brest, France) and authorization from the French Medicine Agency (AFSSAPS).

Participants

The study population consisted of 24 healthy adults (91.7% Caucasian, 4.2% Asian and 4.2% others), 50% males. Their demographic characteristics were the following: mean age 28.7 (range 18–43) years, mean height 174.3 (range 155–191) cm and mean weight 74 (range 54–90) kg. All had given their written informed consent. The main inclusion criteria were to be a healthy male or female aged between 18 to 45 years of age, non-smoker or smoking less than 5 cigarettes per day and, for women, to use an effective method of contraception. The main exclusion criteria were a history of alcohol abuse and/or alcohol addiction, or history of addiction to cannabis, heroin or other opiates, cocaine, crack, amphetamines or GHB, excessive consumption of beverages with xanthine bases (>six cups or glasses/day). Given the potential target population for the therapeutic use of SMO.IR, a mixed gender sample was used. It has been reported that women and men do not differ in their extent of cognitive impairment with low alcohol doses [39]. However, a gender difference in alcohol level has been reported and is due mainly to a smaller gastric metabolism in females and to between-gender differences in body water content. The pharmacokinetic gender difference may increase the vulnerability of women to the effects of alcohol [40]. Therefore the doses of alcohol were adjusted for female participants and the factor 'gender' was included in the statistical analysis of the pharmacodynamics.

Study design

A randomized, double-blind, placebo-controlled design was chosen for unbiased assessment of the pharmacodynamic interaction of SMO.IR and alcohol in healthy volunteers. A four way crossover design was adopted to minimize the variability. Randomization was performed using four sequences of treatments according to a (4 × 4) Williams Latin square design, with stratification on gender. Six subjects were included in each of the four following sequences: sequence 1 = A/B/C/D, sequence 2 = B/D/A/C, sequence 3 = C/A/D/B and sequence 4 = D/C/B/A, with treatment A = 2.25g SMO.IR granules and placebo alcohol preparation, treatment B = 2.25g SMO.IR granules and 0.7 g kg^–1 alcohol for males and 0.57 g kg^–1 alcohol for females, treatment C = SMO.IR matching placebo granules and 0.7 g kg^–1 alcohol for males and 0.57 g kg^–1 alcohol for females and treatment D = SMO.IR matching placebo granules and placebo alcohol preparation.

Experimental procedures

Prior to the screening examination, the nature of the study was explained to all volunteers. Each volunteer was asked to give written informed consent before any screening procedures start. After providing informed consent, subjects were examined for eligibility during the screening visit between day –21 and day –2, prior to entry into the study.

Screening assessments included physical examination, body weight and height, alcohol breath test, urine test for drugs of abuse, medical and medication history, serum pregnancy (for female participants), clinical laboratory tests (haematology, biochemistry and urinalysis), serology (HIV, hepatitis B and C), 12-lead ECG, vital signs (supine and standing systolic and diastolic blood pressure, heart rate) and percentage of oxygen saturation (SpO₂). Screening assessments were performed after an overnight fast (from midnight). Subjects who fulfilled all the inclusion criteria and none of the exclusion criteria were included in the study.

Any medication (except contraception) was forbidden within 1 week before dosing. Caffeine and xanthine consumption as well as smoking were not permitted while in the study centre. The subjects were asked to refrain from intensive physical exercise throughout the conduct of the study (from screening visit until the end of study visit).

As detailed in Table 1, the inpatient study period lasted 24 h (from day –1 evening to day 1 evening) for each of the four study periods except for period 1 (from day –1 afternoon to day 1 evening to enable training for the CNS tests). There was a wash-out period of at least 2 days between each dosing.

Table 1.

Schedule of assessments

	Screening	1st dosing period**		2nd dosing period**		3rd dosing period**		4th dosing period**
Assessment	Day –21 to Day –2	Day –1	Day 1	Day –1	Day 1	Day –1	Day 1	Day –1	Day 1	Follow-up††
Admission to the centre		✓		✓		✓		✓
Discharge from the centre			✓		✓		✓		✓
Written informed consent	✓
Serology§	✓
Inclusion/exclusion criteria	✓	✓		✓		✓		✓
Medical and medication history	✓
Physical examination*	✓		✓		✓		✓		✓	✓
Alcohol breath test and urine drugs of abuse test†	✓	✓	✓	✓	✓	✓	✓	✓	✓
Pregnancy test, if applicable§§	✓	✓		✓		✓		✓		✓
Vital signs and SpO2‡	✓		✓		✓		✓		✓	✓
12-lead ECG	✓		✓		✓		✓		✓	✓
Clinical laboratory safety tests¶	✓		✓		✓		✓		✓	✓
Study drug administration			✓		✓		✓		✓
Body sway, saccadic eye movements, CRT, CTT, DV, NWM, SWM, Bond & Lader VAS, ARCI49, BAES		✓‡‡	✓		✓		✓		✓
Alcohol breath concentration measure			✓		✓		✓		✓
Concomitant medication	←-----------------------------------------------------------------------------------------------------------------------------------------→
Adverse events	←-----------------------------------------------------------------------------------------------------------------------------------------→

^*Including body weight, height and BMI calculation. Height only performed at screening.

^†The drugs of abuse test was also randomly carried out during the in-house stay. The 1 h 10 min alcohol breath test was only performed when the 3.5 h alcohol breath concentration measure was positive.

^‡Includes supine and standing systolic and diastolic blood pressures, heart rates and SpO₂. Vitals signs were performed as follows, for any given scheduled time point Hx: In supine position (after a 10 min rest): Hx, followed by two repeat measurements in supine position: Hx + 1min, Hx + 2 min. Then in standing position (after 3 min in standing position): Hx + 5 min.

^§Includes HIV, hepatitis B and hepatitis C.

^¶Includes haematology, biochemistry and urinalysis.

^**Each dosing period was separated by at least 2 days of washout period.

^††Follow-up assessments were performed 2–5 days after the last dosing day.

^‡‡Habituation test for SEM and training sessions for the others CNS tests on day –1 of period 1 only.

^§§Serum pregnancy test at screening and urine pregnancy test on day –1 of each period and follow-up visit.

Medication

Investigational medical product: SMO.IR was supplied by D&A PHARMA as individual sticks of 2.25 g SO granules for oral administration. The dose of SMO.IR (2.25 g) was selected to sensitize the detection of SO–ethanol interactions. This dose is close to the maximum unit dose administrated to alcohol-dependent patients (50–100 mg kg^–1 three times daily).

Placebo

SMO.IR placebo was supplied by D&A PHARMA as individual sticks of granules for oral administration, indistinguishable from SMO.IR. The composition was the same as SMO.IR (excipients), except for the active component.

Alcohol

Alcohol (ethanol) was supplied by Biotrial in the form of 40% volume of vodka from commercially available batches mixed with 300 ml of apple juice. The dose of alcohol (0.7 g kg^–1 alcohol for males and 0.57 g kg^–1 alcohol for females) was proven to induce mild effects on objective and subjective cognitive measures [41–43]. This dose is also able to be potentiated by psychotropic drugs [44–47] and is in the range of the legal breath alcohol concentration (BrAC) limits tolerated in a wide range of European countries.

Placebo for alcohol

Blinding was achieved by giving 300 ml apple juice. Apple juice was provided by Biotrial from commercially available batches.

Treatments were administered after at least 8 h of fasting on the morning of day 1 of each dosing period. Throughout the inpatient periods, breath alcohol concentrations, safety and pharmacodynamic assessments were performed. The schedule of the assessments is detailed in Appendix S1.

Breath alcohol concentrations

BrAC was measured using a Dräger 7410 ethylometer. BrACs were analyzed to determine the alcohol concentrations at the times of evaluation, i.e. around the expected peak alcohol concentration (0.25–1 h), and during recovery (2.75–3.5 h).

Pharmacodynamics

Saccadic eye movement test

The analysis of eye movements in response to changing visual stimuli provides a simple and reproducible method of detecting changes in attention and sedation in healthy subjects. This method was also demonstrated to be sensitive to both sedative-like and stimulant-like effects of various psychotropic drugs [44], including alcohol effects [45–48]. Different types of eye movements are sensitive to alcohol effects (for a review see [41]). In this study, saccadic eye movements, smooth pursuit and antisaccades were measured.

Saccadic eye movements

Saccadic eye movements are rapid changes of gaze designed to centre a target of interest in the fovea. Eye movements were measured using a non-invasive infrared device (eye-brain tracker™). After removing the potential artefacts, the peak velocity and latency were identified and recorded.

Antisaccades

In this test, the subject had to resist looking at the stimulus when it moved. Instead he had to look an equal distance in the opposite direction. The main parameters were the errors (expressed as percentages) corresponding to a fixation in the same direction as the target, and the latency to correct response (i.e. response to the opposite site of the target)

Smooth pursuit

The subject had to follow a target moving in a sinusoidal manner at a constant speed of 30°s^–1. The average root mean square i.e. the distance between the target and the position of the eye during the smooth pursuit was the main parameter.

Body sway

This is an objective, sensitive, reliable and non-invasive method designed to assess the effects of drugs and alcohol on postural stability [44,45,49–52].

Body sway was recorded using a force-platform (Win Posture®, Medicapteurs; Balma, France). Measurements of body sway (1 min with eyes open and 1 min with eyes closed) were recorded as recommended by the International Society of Posturography [53]. The length (x axis and y axis) and area of the postural oscillations were then calculated.

Choice reaction time (CRT)

Alcohol impairs the speed of information processing in complex attentional tasks [41,46,54] and leads to reaction time increases in choice reaction time tasks [46,54,55].

This test evaluated the speed at which a subject was able to respond to a complex visual stimulus. Performance was assessed as the speed of processing (in ms), and the answer accuracy expressed as the percentage of correct answers.

Critical tracking test (CTT)

The CTT is sensitive to alcohol effects [45,46,54,56,57]. It measures eye-hand coordination and delays in visual motor response.

The performance was expressed in pixels and consisted of the average distance between the target and the cursor.

Digit vigilance (DV)

This test assessed selective attention i.e. the ability to detect a target stimulus during a concentrated attention task. Selective attention has been found to be impaired by alcohol [47,58].

The performance was the speed of processing measured as the mean reaction time and the answer accuracy measured as the percentage of correct answers.

Numeric working memory (NWM) and spatial working memory (SWM)

Inconsistent results were found in the studies that have examined the acute effects of alcohol on working memory [41]. These inconsistent findings can be explained by the fact that the different components of working memory (phonological loop, visuo-spatial sketchpad and central executive, [59]) might be differentially impacted on by alcohol [42,60]. Therefore, a numeric working memory task and a visuo-spatial working memory task were used in this study [61].

For both tasks, the mean reaction time and the percentage of correct answers were measured.

Bond & Lader VAS (B&L VAS)

The Bond & Lader Visual Analogue Scales consist of 16 bipolar self-rating 100 mm long lines between two opposite adjectives [62]. The measurements were expressed in mm and the scores consisted of three derived factors, alertness, contentedness (well-being) and calmness.

Addiction research centre inventory 49 (ARCI)

This questionnaire consisted of 49 ‘true–false’ items for assessments of mood according to five scales [63–65], each corresponding to one class of abused drug: morphine-benzedrine group (MBG; measure of euphoria), amphetamine group (AG: stimulant-like effects), benzedrine group (BG: intellectual efficiency and energy), pentobarbital-chlorpromazine alcohol group (PCAG: sedation and alteration of motivation) and lysergic acid (LSD: dysphoria and somatic effects).

Biphasic alcohol effects scale (BAES)

This questionnaire consisted of 15 items, describing feelings commonly experienced after drinking alcohol [66]. The parameters were analyzed to provide two BAES sub-scores, a stimulation sub-score and a sedation sub-score.

These three questionnaires were self-administered and computer-assisted.

Safety criteria

Safety assessments included physical examination, alcohol breath test, SpO₂ measured with a fingertip pulse oximeter, vital signs, 12-lead ECG, clinical safety laboratory tests (haematology, biochemistry and urinalysis) and adverse events collection.

Statistical methods

All data were summarized as means and standard errors of the means. Analyses of variance (anova) were conducted on SAS® software 9.2 release (SAS Institute Inc. Cary NC, USA) for pharmacodynamic data. A potential pharmacodynamic drug–alcohol interaction was evaluated by an analysis of variance applied to a 2 × 2 factorial interaction model [67,68] performed on change from baseline at each measurement time with treatment, period, carry-over and gender as fixed effect effects and subject as a random effect (carry-over was removed from the model in the case of a non-significant effect). Gender data are not shown here as no gender effect was seen.

The treatment effect was partitioned into four levels (placebo, alcohol, SMO.IR, SMO.IR + alcohol) and orthogonal contrasts were computed to test SMO.IR*alcohol interaction [i.e. potentiation = (SMO.IR + alcohol) + (placebo + placebo alcohol) vs. (placebo + alcohol) + (SMO.IR + placebo alcohol )], the main drug effect [i.e. SMO.IR effect = (SMO.IR + alcohol) + (SMO.IR + placebo alcohol) vs. (placebo + alcohol) + (placebo + placebo alcohol)] and alcohol effect [i.e. alcohol effect = (SMO.IR + alcohol) + (placebo + alcohol) vs. (SMO.IR + placebo alcohol) + (placebo + placebo alcohol)], with a fixed α level of 5%.

Results are presented as estimates, 95% confidence intervals and P values (statistical level of significance for P < 0.05, two-tailed test). As indicated previously, no formal statistical hypothesis was tested.Thus all P values are interpreted in an exploratory sense and therefore no adjustment for multiplicity was done.

Results

Study population

Forty-four subjects were screened and 24 subjects were included. These enrolled subjects were determined to be healthy as established by medical, clinical and laboratory examinations. One randomized subject withdrew her consent (for personal reasons) after completing the third dosing period and thus was not dosed in the fourth period. Consequently, 24 subjects were included in the safety analyses and 23 in the pharmacodynamic analyses.

Pharmacodynamic results

The effects of alcohol

When compared with baseline performance, alcohol had significant deleterious effects on several subjective, physiological and performance parameters within 1 h and 2 h 45 min post-dose (see Table 2).

Table 2.

Main results obtained for alcohol effect

		Change from baseline
		t +15min			t+2h45
		estimate	95%CI	F ratios,df,P value	estimate	95%CI	F ratios,df,P value
Saccadic eye movement	Saccadic eye movement – peak velocity	−35.8	[–45.49, –26.12]	F(1,60) = 54.67, P < 0.0001	–27.75	[–40.94, –14.56]	F(1,60) = 17.72, P < 0.0001
	Saccadic eye movement – latency	19.69	[9.86, 29.51]	F(1,60) = 16.07, P = 0.0002	15.68	[6.99, 24.37]	F(1,60) = 13.02, P = 0.0006
	Anti-saccades - latency	22.35	[12.04, 32.66]	F(1,57) = 18.86, P < 0.0001	9.3	[–0.01, 18.61]	F(1,60) = 0.08, P = NS
Body sway	Area eyes closed	0.18	[0.02, 0.35]	F(1,62) = 4.92, P = 0.0302	0.13	[–0.07, 0.33]	F(1,62) = 1.74, P = NS
	Length eyes closed	0.11	[0.02, 0.19]	F(1,62) = 6.45, P = 0.0136	0.01	[–0.07, 0.1]	F(1,62) = 0.13, P = NS
CTT	Mean distance	10.62	[-5.03, 26.27]	F(1,63) = 1.84, P = NS	3.58	[1.26, 5.9]	F(1,63) = 9.5, P = 0.003
CRT	% good answers	–1.19	[–2.37, –0.02]	F(1,63) = 4.14, P = 0.0461	–0.39	[–1.21, 0.43]	F(1,63) = 0.91, P = NS
Digit vigilance	Mean reaction time	26.25	[13.93, 38.56]	F(1,62) = 18.14, P < 0.0001	4.95	[–6.69, 16.58]	F(1,63) = 0.72, P = NS
NWM	Mean reaction time	–14.04	[–73.5, 45.43]	F(1,62) = 0.22, P = NS	–45.68	[–88.47, –2.89]	F(1,60) = 4.56, P = 0.0368
	% good answers	–4.46	[–7.19, –1.73]	F(1,59) = 10.67, P = 0.0018	–0.92	[–3.73, 1.89]	F(1,63) = 0.43, P = NS
SWM	Mean reaction time	103.83	[19.01, 188.64]	F(1,62) = 5.99, P = 0.0172	75.03	[–26.38, 176.44]	F(1,63) = 2.19, P = NS
	% good answers	–5.65	[–9.84, –1.46]	F(1,62) = 7.26, P = 0.0091	2.04	[–3.34, 7.42]	F(1,60) = 0.58, P = NS
B & L VAS	Alertness	–3.34	[–10.29, 3.61]	F(1,62) = 0.92, P = NS	–8.48	[–14.56, –2.4]	F(1,63)=7.77, P = 0.007
ARCI	Amphetamine	0.92	[0.18, 1.66]	F(1,62) = 6.20, P = 0.0155	0.06	[–0.44, 0.56]	F(1,63) = 0.06, P = NS
	Benzedrine (BG)	0.14	[–0.75, 1.04]	F(1,62) = 0.10, P = NS	–0.98	[–1.79, –0.17]	F(1,63) = 5.86, P = 0.0184
	MBG	2.34	[1.1, 3.59]	F(1,62) = 14.17, P = 0.0004	–0.03	[–0.83, 0.77]	F(1,63) = 0.01, P = NS
	PCAG	1.14	[–0.36, 2.64]	F(1,62) = 2.31, P = NS	3.08	[1.78, 4.38]	F(1,63) = 22.40, P < 0.0001
BAES	Sedation	2.81	[–1.18; 6.81]	F(1,62) = 1.98, P = NS	9.07	[4.41, 13.74]	F(1,63) = 15.12, P = 0.0002
	Stimulation	6.04	[1.64, 10.44]	F(1,62) = 7.52, P = 0.008	–3.16	[–6.78, 0.45]	F(1,63) = 3.05, P = NS

CTT, critical tracking test; CRT, choice reaction time; NWM, numeric working memory; SWM, spatial working memory; B & L VAS, Bond & Lader visual analogue scales; ARCI, Addiction Research Centre Inventory; MBG, morphine-benzedrine group; PCAG, pentobarbital-chlorpromazine group; BAES, biphasic alcohol effects scale; df, degree of freedom

Within 1 h post-dose, alcohol significantly decreased peak velocity and increased latencies in saccadic eye movements. Latencies were also significantly increased in the antisaccades test. Significant increased length and area of postural oscillations were recorded for body sway in eyes closed condition. Fifteen minutes after alcohol ingestion, performance accuracy was significantly impaired in NWM and SWM tests and in CRT. Reaction times were substantially increased in DV and SWM tests. Regarding the subjective effects, alcohol had few immediate effects. Significant increases in disinhibition (ARCI -Amphetamine scale and Stimulation in BAES) and in euphoria were observed.

Considering the longer time effects (i.e. 2 h 45 min post-dose), alcohol significantly decreased peak velocity and increased latencies in saccadic eye movements, increased mean tracking distance in CTT and had a deleterious effect on mean reaction time on NWM. Regarding the subjective effects, alcohol significantly impaired the feeling of alertness (B&L VAS), increased the feeling of sedation (PCAG scale – ARCI and BAES) and decreased stimulation (BG scale – ARCI).

SMO.IR effects

When compared with baseline performance, SMO.IR also induced significant short (within 1 h post-dose) and long (2 h 45 min post-dose) time effects on subjective, physiological and performance parameters (Table 3).

Table 3.

Main results obtained for SMO.IR effect

		Change from baseline
		t +15min			t +2h45
		estimate	95%CI	F ratios,df,P value	estimate	95%CI	F ratios,df,P value
Saccadic eye movement	Saccadic eye movement – peak velocity	–9.58	[–19.28, 0.13]	F(1,60) = 3.89, P = NS	–18.44	[–31.66, –5.23]	F(1,60) = 7.79, P = 0.007
Body sway	Length eyes open	–0.08	[–0.15, –0.02]	F(1,59) = 6.17, P = 0.0158	–0.03	[–0.11, 0.06]	F(1,59) = 0.36, P = NS
CTT	Mean distance	10.73	[–4.92, 26.37]	F(1,63) = 1.88, P = NS	2.67	[0.34, 4.99]	F(1,63) = 5.27, P = 0.0251
CRT	Mean reaction time	–0.74	[–31.97, 30.50]	F(1,63) = 0.00, P = NS	30.77	[11.27, 50.28]	F(1, 63) = 9.94, P = 0.0025
Digit vigilance	Mean reaction time	15.95	[3.63, 28.27]	F(1,62) = 6.70, P = 0.012	17.7	[6.06, 29.34]	F(1,63) = 9.23, P = 0.0035
NWM	Mean reaction time	26.23	[-33.24, 85.69]	F(1,62) = 0.78, P = NS	56.6	[14.2, 99]	F(1,60) = 7.13, P = 0.0097
B & L VAS	Alertness	–7.41	[–14.36, –0.46]	F(1,62) = 4.54, P = 0.037	–6.84	[–12.91, –0.76]	F(1,63) = 5.05, P = 0.0281
	Contentedness	–2.49	[–6.89, 1.92]	F(1,62) = 1.27, P = NS	–6.01	[–9.62, –2.41]	F(1,63) = 11.11, P = 0.0014
ARCI	Benzedrine (BG)	–1.18	[–2.08, –0.29]	F(1,62) = 7.02, P = 0.0102	–0.97	[–1.77, –0.16]	F(1,63) = 5.70, P = 0.0199
	LSD	1.69	[0.86, 2.52]	F(1,62) = 16.48, P = 0.0001	1.02	[0.09, 1.95]	F(1,63) = 4.75, P = 0.033
	MBG	–0.2	[–1.44, 1.04]	F(1,62) = 0.10, P = NS	–0.9	[–1.7, –0.1]	F(1,63) = 5.07, P = 0.0279
	PCAG	3.57	[2.07, 5.07]	F(1,62) = 22.55, P < 0.0001	2.26	[0.96, 3.56]	F(1,63) = 12.12, P = 0.0009
BAES	Sedation	6.73	[2.74, 10.73]	F(1,62) = 11.83, P = 0.0013	4.6	[–0.06, 9.26]	F(1,63) = 3.89, P = NS
	Stimulation	–4.17	[–8.58, 0.23]	F(1,62) = 3.59, P = NS	–5.15	[–8.77, –1.53]	F(1,63) = 8.09, P = 0.006

CTT, critical tracking test; CRT, choice reaction time; NWM, numeric working memory; B & L VAS, Bond & Lader visual analogue scales; ARCI, Addiction Research Centre Inventory; LSD, lysergic acid; MBG, morphine-benzedrine group; PCAG, pentobarbital-chlorpromazine group; BAES, biphasic alcohol effects scale; df, degree of freedom

Within 1 h post-dose, SMO.IR had few effects on performance. A significant decrease in the length of postural oscillations in eyes opened condition (body sway) and a significant increase in response times in DV performance were observed. SMO.IR also had more effects on subjective ratings, as a significant decrease in alertness (B&L VAS) and stimulation (BG scale – ARCI) was observed. Significant increases in sedation (PCAG scale – ARCI and BAES) and in dysphoria as well as psychotomimetic changes (LSD scale – ARCI) were also observed.

SMO.IR had a significant deleterious effect on response speed performance (CRT, DV, NWM) as well as on psychomotor performance (CTT) and on saccadic eye movement peak velocity 2 h 45 min post-dose. SMO.IR also induced numerous significant effects on subjective feelings. Alertness and contentedness decreased (B&L VAS), as well as euphoria (MBG scale – ARCI) and stimulation (BG scale – ARCI and BAES). Sedation (PCAG - ARCI), dysphoria and psychotomimetic changes (LSD scale – ARCI) increased.

Alcohol and SMO.IR interaction effects

When administered concomitantly, alcohol and SMO.IR had no effect on performance or on physiological parameters within 1 h post-dose, but several significant interactions between SMO.IR and alcohol were observed in subjective parameters. Alertness (B&L VAS) and stimulation (BG scale – ARCI and BAES) increased significantly after co-administration of alcohol and SMO.IR, whereas sedation (PCAG scale – ARCI and BAES) decreased significantly.

No significant interaction between SMO.IR and alcohol was observed 2 h 45 min post-dose in subjective and physiological parameters. In performance parameters, an isolated interaction between SMO.IR and alcohol was observed: accuracy of DV was significantly decreased. These data are detailed in Table 4.

Table 4.

Main results obtained for SMO.IR and alcohol interaction effect

		Change from baseline
		t +15min			t +2 h 45 min
		estimate	95%CI	F ratios,df,P value	estimate	95%CI	F ratios,df,P value
Digit vigilance	% good answers	–0.67	[–5.58, 4.24]	F(1,62) = 0.07, P = NS	–6.66	[–11.72, –1.6]	F(1,63) = 6.93, P = 0.0106
B & L VAS	Alertness	10	[3.05, 16.95]	F(1,62) = 8.27, P = 0.0055	5.47	[–0.61, 11.54]	F(1,63) = 3.23, P = NS
ARCI	Benzedrine (BG)	1.53	[0.64, 2.43]	F(1,62) = 11.77, P = 0.0011	0.53	[–0.28, 1.34]	F(1,63) = 1.73, P = NS
	PCAG	–2.87	[–4.37, –1.37]	F(1,62) = 14.58, P < 0.0003	–0.75	[–2.05, 0.55]	F(1,63) = 1.32, P = NS
BAES	Sedation	–8.71	[–12.71, –4.71]	F(1,62) = 18.97, P < 0.0001	–0.79	[–5.45, 3.87]	F(1,63) = 0.12, P = NS
	Stimulation	4.54	[0.14, 8.94]	F(1,62) = 4.25, P = 0.0435	2.15	[–1.47, 5.77]	F(1,63) = 1.41, P = NS

B & L VAS, Bond & Lader visual analogue scales; ARCI, Addiction Research Centre Inventory; BG, benzedrine group; PCAG, pentobarbital-chlorpromazine group; BAES, biphasic alcohol effects scale; df, degree of freedom

A summary of the pharmacodynamic results and effects are displayed in Appendix S2 and S3, respectively.

Breath alcohol concentrations

Figure1 illustrates the mean BrAC for 2.25g SMO.IR and alcohol and for SMO.IR matching placebo and alcohol. There was no difference in BrAC between these treatment conditions as revealed by the statistical analyses performed to examine the differences between both drug administrations (F(1,157) = 2.98, NS). C_max (mg ml⁻¹) and t_max (h) of alcohol concentration in the treatment groups were very similar, whether alcohol was taken with (0.66 ± 0.16 and 0.79 ± 0.34, for C_max and t_max respectively) or without (0.69 ± 0.14 and 0.86 ± 0.53, for C_max and t_max respectively) co-administration of SMO.IR 2.25g. The AUC (mg h ml⁻¹) for alcohol alone (1.649 ± 0.304) was slightly greater than that for alcohol + SMO.IR (1.553 ± 0.377). In summary, whatever the time point, no pharmacokinetic interaction with alcohol was observed (F(3, 155) = 0.07, NS).

Figure 1.

Alcohol breath concentrations over time per treatment group. treatment B SMO.IR + alcohol; treatment C alcohol alone

Tolerability

No serious adverse event (AE) and no withdrawal due to an AE occurred throughout the study. A total of 113 AEs were reported in 24 subjects over the four periods. One hundred and ten were treatment emergent adverse events (TEAEs). The details of the reported AEs and TEAEs according to treatment are displayed in Table 5. Most TEAEs were mild or moderate and were expected reactions to either the ingestion of alcohol or to the sedative effect of SMO.IR that resolved spontaneously within a few hours after treatment administration.

Table 5.

Detail of the reported adverse events (AEs) according to study treatments

		SMO.IR		SMO.IR+Alcohol		Alcohol		Placebo
System organ class	Preferred term	(n =24)		(n =23)		(n =24)		(n =24)
		n (%)	AE	n (%)	AE	n (%)	AE	n (%)	AE
All	All	21 (87.5)	30	22 (95.7)	46	21 (87.5)	34	2 (8.3)	3
Gastrointestinal disorders	Any	2 (8.3)	2	4 (17.4)	5	1 (4.2)	1	0 (0.0)	0
	Abdominal pain - upper	0 (0.0)	0	0 (0.0)	0	1 (4.2)	1	0 (0.0)	0
	Nausea	1 (4.2)	1	3 (13.0)	3	0 (0.0)	0	0 (0.0)	0
	Vomiting	1 (4.2)	1	2 (8.7)	2	0 (0.0)	0	0 (0.0)	0
General disorders	Any	7 (29.2)	7	15 (65.2)	16	18 (75.0)	18	0 (0.0)	0
	Asthenia	4 (16.7)	4	4 (17.4)	4	2 (8.3)	2	0 (0.0)	0
	Feeling drunk	3 (12.5)	3	12 (52.2)	12	16 (66.7)	16	0 (0.0)	0
Nervous system disorders	Any	14 (58.3)	18	12 (52.2)	13	9 (37.5)	11	2 (8.3)	2
	Disturbance in attention	3 (12.5)	3	2 (8.7)	2	2 (8.3)	2	1 (4.2)	1
	Dizziness	9 (37.5)	9	4 (17.4)	4	1 (4.2)	1	0 (0.0)	0
	Dizziness postural	0 (0.0)	0	1 (4.3)	1	0 (0.0)	0	0 (0.0)	0
	Headache	2 (8.3)	2	2 (8.7)	2	5 (20.8)	5	1 (4.2)	1
	Psychomotor skills impaired	1 (4.2)	1	0 (0.0)	0	1 (4.2)	1	0 (0.0)	0
	Somnolence	3 (12.5)	3	4 (17.4)	4	2 (8.3)	2	0 (0.0)	0
Psychiatric disorders	Any	3 (12.5)	3	8 (34.8)	9	4 (16.7)	4	1 (4.2)	1
	Bradyphrenia	1 (4.2)	1	2 (8.7)	2	0 (0.0)	0	0 (0.0)	0
	Depressed mood	0 (0.0)	0	1 (4.3)	1	2 (8.3)	2	0 (0.0)	0
	Euphoric mood	2 (8.3)	2	6 (26.1)	6	2 (8.3)	2	1 (4.2)	1
Vascular disorders	Any	0 (0.0)	0	3 (13.0)	3	0 (0.0)	0	0 (0.0)	0
	Hot flush	0 (0.0)	0	2 (8.7)	2	0 (0.0)	0	0 (0.0)	0
	Hypotension	0 (0.0)	0	1 (4.3)	1	0 (0.0)	0	0 (0.0)	0

When SMO.IR was administered together with alcohol, the total number of the TEAEs (46) was slightly higher than when either SMO.IR alone (30) or alcohol alone (34) were administered.

The most frequently observed TEAEs were nervous system disorders (38.9%). Feeling drunk was a TEAE reported by 27.4% of the patients.

Clinical laboratory, vital signs and physical examinations for all subjects were normal and no abnormalities were observed while measuring SpO2.

Discussion

The results of the present study show that alcohol ingestion led to the expected effects on pharmacodynamic assessments. Rapidly after alcohol intake, performance was negatively impacted and some stimulant-like or disinhibitory effects on mood were observed. Three hours later, most of the performance and physiological impairments had vanished, but subjective effects increased and changed. From stimulant-like they became sedative-like effects. This biphasic alcohol-like effect on subjective ratings is well-described in the literature (see [69] for a review). The behavioural effects observed in this study are also in line with other published results [41,46,47,50,54], even if the first saccadic eye movement assessments were less sensitive than expected. The pharmacodynamic effects were assessed at two time points (15 min post-dose and 2 h 45 min post-dose). This time selection was based on the duration of the pharmacodynamic battery which took nearly 1 h to be completed. The peak alcohol effects are generally observed between 20–30 min and 1 h after alcohol ingestion [70]. Thus the full battery testing covered well the time course of the peak alcohol effects but the first assessments occurred shortly before the peak blood alcohol concentration. The saccadic eye movement tests were one of the first assessments of the battery (just after body sway), and this may explain that they were less sensitive than they would have been at the peak alcohol effects.

The pharmacodynamic effects of SMO.IR clearly suggest some sedative-like impairment characterized by a slight slowing down of processing speed 3 h after treatment intake and by subjective complaints. The subjective effects were present at both assessment time points, whereas cognitive impairments were quasi inexistent 15 min post-dose. These behavioural effects are more or less consistent with those described in the literature [71–74]. Little is known about the acute effects of SO on cognitive performance in healthy volunteers without a history of drug and/or alcohol abuse [74]. On the one hand, no psychomotor impairment was evidenced after administrations of low doses of SO (12.5 and 25 mg kg^–1 doses) [75]. On the other hand, more recent studies showed that cognition was impaired with higher doses of SO (see [74] for a review). For instance, significant cognitive impairments were shown with a 4.5g 70 kg^–1 dose [76] and significant deleterious effects on digit symbol substitution test and balance task were obtained with 60 and 72 mg kg^–1 doses (mean population weight: 71.9 kg) [73]. In contrast, in the present study the dosage was 2.25 g (mean population weight: 74 kg). Thus, the slight objective cognitive impairments observed are in line with results from previous studies conducted with the same doses of SO and with the same type of population. Regarding the subjective effects, a biphasic effect in the time course of SO which could be compared with that of alcohol was described [72]. There was first a psychostimulant effect during the first hour and then a sedative effect during the second hour after drug administration. Our data do not show such a psychostimulant effect at the first assessment time point (t +15min) but a sedative effect was present from the first time point. Interestingly several pharmacodynamic effects of SO are still observed 3 h post-dosing whereas SO is very rapidly cleared by hepatic metabolism (t_½ = 20–60 min) and has no active metabolite [5].

The cognitive impairment induced by SMO.IR in healthy volunteers is clearly less deleterious than a benzodiazepine-induced cognitive impairment in healthy volunteers. Benzodiazepine use has been associated with a wide range of cognitive side effects [46,77,78]. Few studies have compared SO and benzodiazepines on cognitive tasks [76–79], and due to differences in study populations and in cognitive tasks, the data of this study cannot be compared with Carter's data [76,79]. Nevertheless, the present results showed that sedation induced by SMO.IR remained mainly subjective and was less frequently observed in performance tasks. Thus the sedative-like effects of SMO.IR seem to be less pronounced than those generally observed with therapeutic doses of benzodiazepines that induce a slowing-down response and performance impairment (see [80] for an example of lorazepam-induced impairment on objective and subjective assessments or [81] for an overview of acute lorazepam effects on neurocognitive performance).

When SMO.IR was co-administered with alcohol, a significant interaction was observed 15 min post-dose, which antagonized the subjective sedative effects of alcohol. A mild decrease in selective attention was observed at 2 h 45 min post-dose after co-ingestion of alcohol and SMO.IR. No other effects were observed when both products were co-administered. Safety results are consistent with a previous SO and alcohol interaction study [82], indicating a slight increase in adverse events when both products are co-administered. Nevertheless, this co-administration was safe and well-tolerated. Taken together these results suggest that SMO.IR and alcohol have a distinct adverse effect profile. Objective sedative effects of SMO.IR are much less marked than those of alcohol, whereas subjective (mainly sedative) complaints of SMO.IR are more pronounced than those of alcohol. No potentiation was observed. These results are quite encouraging as they show that the administration of SO in alcohol dependent patients is safe in case patients would relapse during the treatment period. Thus SMO.IR seems to be a treatment allowing reduction of alcohol consumption (not alcohol abstinence). This is of particular interest for alcohol dependence treatment as a major bottleneck in alcohol therapy is that patients are often not ready or not willing to stop drinking.

Even if the study protocol was well-controlled and led to clear conclusions, some limitations have to be acknowledged. First, the subjects had no history of alcohol use/abuse and therefore did not reflect the targeted population. However, in subjects with histories of abuse of sedative/hypnotic drugs, a great variability in sedation ratings was demonstrated [79] and most of them remained awake and alert after an administration of 2 or 4 g 70 kg^–1 GHB. Moreover, only one dose of SMO.IR and one dose of alcohol were tested. Therefore the conclusions that can be drawn are limited. Consequently, it seems difficult to extend the results of the co-administration of SMO.IR and alcohol obtained in healthy volunteers to an alcohol-dependent or a drug-dependent population and/or to other doses of SMO.IR and alcohol. The pharmacodynamics effects were assessed at two time points (t +15 min and t +2 h 45 min) and this schedule may not have been optimal for assessing a peak alcohol effect as the latter is generally observed between 20–30 min and 1 h after ingestion [70]. As the full pharmacodynamic battery took near to 1 h to be completed, the first assessments occurred a little time before the peak alcohol concentration, but the full battery of tests covered well the time course of the peak alcohol effects. Another point regarding the schedule of assessments is the length of the assessment period that contributes to the question of the interaction between fatigue and performance. The completion of objective and subjective assessments took about one hour, but the participant did not have to concentrate and to perform during 1 full h. Indeed, this hour of testing comprised moving from one display (body sway, saccadic eyes movements, cognitive test computer) to the other, the adjustment of these displays (especially for saccadic eyes movements) and the launch/instructions for each test. Consequently, participants could switch between periods of concentration and periods during which there was no need to perform and during which their attention could wander. These off-periods may have helped to limit the fatigue induced by the cognitive effort required by each assessment, as well as the crossover design may have controlled this potential fatigue bias. Finally, as this study was an exploratory study no formal statistical hypothesis was tested and no adjustment for multiple comparisons was applied. Thus the results have to be interpreted carefully as this absence of corrections for multiple comparisons may have led to some false positives.

In conclusion, SMO.IR was well-tolerated in this population and the results are quite encouraging. The low level of interaction between alcohol and SMO.IR may account for the good safety profile of SMO.IR even in the case of alcohol relapses in alcohol-dependent patients. No deleterious clinically significant interaction between alcohol and SMO.IR on cognitive function was observed. Thus, this new formulation of SO, which might also reduce the risk of criminal or illegal use of the drug, presents a potentially new pharmacological approach for the treatment of alcohol withdrawal syndrome, reducing alcohol consumption or maintaining alcohol abstinence.

Competing Interests

All authors have completed the Unified Competing Interest form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare that MB, NF, AP and NP had support from D&A Pharma for the submitted work. There were no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years (except PV who is currently employed by D&A Pharma) and no other relationships or activities that could appear to have influenced the submitted work.

Supporting Information

Additional Supporting Information may be found in the online version of this article at the publisher's web-site:

Appendix S1 Schedule of Assessments on Dosing Days

Appendix S2 Summary of the main results – Pharmacodynamic testing

Appendix S3 Mean scores (and standard errors) expressed in raw data and changes from baseline for the different assessments performed in each study condition and at each time point