The primary findings of the present study were 1) naltrexone produced modest, but dose-dependent, decreases in alcohol drinking, under both acute and chronic dosing conditions in baboons with long-term self-administration experience under a CSR, 2) naltrexone’s effects on alcohol intake were primarily by decreasing the magnitude of the first drinking bout, an effect also observed for intake of the non-alcoholic beverage, 3) in the context of alcohol access and ongoing drinking, the motivation to obtain alcohol in the CSR was largely unaffected by naltrexone, and 4) alcohol seeking responses, which were highly resistant to extinction, were dose-dependently and selectively decreased by naltrexone during within-session extinction tests. Each of these findings is discussed in detail below.
Throughout the current experiments, the non-alcoholic beverage maintained a higher rate of self-administration than alcohol. However, BPs determined prior to naltrexone testing for the Control Group matched BPs of the Alcohol Group. Similarly, BPs obtained during Experiment 3 (which did not include naltrexone pretreatment) were also similar for the two reinforcers, demonstrating the two fluids functioned as comparable reinforcers.
The FDA recommended dose of naltrexone for treatment of alcohol dependence is 50 mg, although higher doses (100–150 mg/kg) are used successfully in clinical trials (Anton et al 2003; Yoon et al 2011). Using an interspecies dose conversion formula (Dews 1976
; Mordenti and Chappell 1989
), the FDA recommended dose of 50 mg naltrexone would be comparable to a dose of 0.92 mg/kg in baboons. Thus, doses of 1 mg/kg or higher would be expected to reduce alcohol drinking in baboons.
To evaluate effects on self-administration responses, in Experiment 1, 0.32–3.2 mg/kg naltrexone was administered 50–60 min before alcohol access. Acute naltrexone dose-dependently reduced alcohol self-administration responses and total amount of alcohol consumed. Naltrexone administered for five consecutive days (Experiment 4) reduced alcohol self-administration responses and consumption in ways similar to that observed under acute administration conditions (Experiment 1). Individual differences were seen in the Control Group, with effects on the non-alcoholic beverage that were similar to those on alcohol in two of the four baboons (Experiment 1).
The concentration of alcohol (4% w/v) used in the present experiment has previously been shown to maintain stable intake. While baboons will drink high concentrations of alcohol (e.g., 16% w/v) during induction of alcohol drinking (Henningfield et al. 1981
), lower concentrations of alcohol are generally preferred and maintain higher rates of operant self-administration (Ator and Griffiths 1992
). Baboons will titrate volume of higher concentrations to achieve g/kg intake within a similar intoxicating range. Although the effects of naltrexone were not evaluated across a range of concentrations, previous research in nonhuman primates (Williams et al. 1997) demonstrated that naltrexone (0.1 mg/kg, administered 30 min prior to the session) decreased alcohol-reinforced responding similarly across a range of concentrations (0.25% w/v – 4% w/v), that maintained both low intake (0.25% w/v) and high intake (4% w/v), shifting the ethanol-concentration curve downward in an unsurmontable manner.
Studies in laboratory animals have consistently reported that naltrexone administration reduces alcohol self-administration and intake, and many studies have reported this effect was not selective for alcohol (Egli 2005
). For example, in nonhuman primates, doses of 0.1–3.0 mg/kg naltrexone, administered 30 min before self-administration, reduced self-administration of alcohol and Tang (Shelton and Grant 2001
), saccharin (Rodefer et al. 1999
), water (Williams and Woods 1999
), and sucrose (Williams et al. 1998
). Although it has been argued that the reduction of intake of alcohol and other orally consumed substances by naltrexone may, in part, be related to changes in palatability (Ferraro et al 2002
; Hill et al. 1997
), studies demonstrating decreased intravenous self-administration of alcohol in monkeys (Altshuler et al. 1980
; Williams et al. 1998
) provide evidence for opioid mediation of reinforcement, which is unrelated to any buccal or gustatory effects (Olszewski et al. 2011
All operant responses were recorded in real time allowing the examination of the within-session patterning of alcohol self-administration. Similar to previous reports with rats (Samson et al. 2000
), baboons (Weerts et al. 2006
) and other primates (Boyle et al. 1998
; Grant et al. 2009; Macenski and Meisch 1992
; Rodefer et al. 1999
; Vivian et al. 2001
; Williams et al. 1998
), baboons in the current study engaged in “loading;” the greatest magnitude of drinking occurred early in the alcohol self-administration period followed by lower rates of drinking for the remainder of the period. Naltrexone administered acutely (Experiment 1) or chronically (Experiment 4) did not delay onset of drinking or alter the general within-session patterning of drinking. Instead, the magnitude of the initial drinking bout was significantly reduced for both alcohol and the nonalcoholic reinforcer, further suggesting that naltrexone reduced reinforcing effects that were directly related to consumption. These data are consistent with human laboratory alcohol administration studies showing that naltrexone significantly decreased self-reported ratings of alcohol liking and best effects (King et al. 1997
; McCaul et al. 2000
; O’Malley et al. 2010).
Naltrexone did not decrease most alcohol seeking measures in C2 of the CSR. In Experiment 2, acute doses of naltrexone tended to decrease the number of responses directed towards obtaining the daily supply of alcohol, but did not prolong latency to obtain alcohol (FI response latency), and did not alter any seeking measure in the Control Group. In Experiment 4, chronically administered naltrexone also did not reduce BPs to obtain alcohol. This finding is consistent with a recent controlled laboratory study in social drinkers, which also found naltrexone did not alter alcohol seeking BPs (Setiawan et al. 2011
). The present data are consistent with previous research showing that alcohol seeking responses, which provide a measure of the motivation to drink in the CSR, are largely under the control of stimuli/cues which have acquired conditioned reinforcing and/or eliciting properties, and are highly resistant to change (Kaminski et al. 2008
). Previous research has shown that environmental or contextual stimuli associated with drug use are an important mitigating factor in relapse to drug use (Childress et al. 1992
; Collins and Brandon 2002
; Crombag and Shaham 2002
It is noteworthy then, that in Experiment 3, acute doses of naltrexone selectively facilitated extinction of seeking responses in the Alcohol Group. Naltrexone did not decrease extinction responding in a clear dose-related manner in any baboon in the Control Group (e.g., a decrease at 0.32 mg/kg was seen in one baboon, no change in two, and an increase in one). This finding is consistent with rodent studies showing naltrexone (doses ranging from 0.1–3.0 mg/kg) selectively facilitated extinction of responses previously maintained by alcohol (Bienkowski et al. 1999
; Czachowski and DeLory 2010
) and attenuated alcohol- and cue-induced reinstatement of alcohol self-administration (Bienkowski et al. 1999
; Lê et al. 1999; Liu and Weiss 2002
). The current study further determined that it selectively decreased resistance to extinction by suppressing seeking responses in the early phase of extinction and decreasing the length of time responding persisted. This finding is noteworthy given that under vehicle conditions alcohol showed greater resistance to extinction when compared to the non-alcoholic reinforcer.
The baboons in the current study had long-term self-administration experience under the CSR with either alcohol or the non-alcoholic beverage. In the Alcohol Group, baboons drank 1 g/kg/day of alcohol, which was maintained at stable levels 7 days/week. Mean blood alcohol level (BAL) of 88.2 mg/dl (>0.08%) were previously determined in these same baboons after comparable alcohol intake (mean 0.93 g/kg) (Kaminski et al. 2008
). This level of drinking is important, as relatively few models using outbred animals generate patterns of volitional alcohol intake that exceed a threshold associated with risk of harm (Leeman et al. 2010
). Problematic or “at risk” drinking in man includes patterns of drinking to intoxication (e.g., 0.8 to 1 g/kg, BAL > 0.08%) within a single drinking period (binge) as well as regular drinking at this level across days (heavy drinking). Thus, the level and pattern of alcohol intake by baboons under the CSR is pharmacologically relevant based on NIAAA definitions, particularly in the context of their long-term drinking history and exposure to alcohol-related cues in the drinking environment.
The observed decreases in seeking responses under extinction indicate that naltrexone facilitates extinction of persistent behaviors associated with alcohol-related cues. When alcohol was available, naltrexone produced modest decreases in alcohol drinking, but did not eliminate it. The results, then, support those studies showing that naltrexone may work, in part, by preventing drinking episodes from becoming a full-fledge relapse into heavy drinking (Anton et al. 1999
, 2004; O’Malley and Froehlich 2003
, Pettinati et al. 2006
). The consistency of the present results with available clinical data and observations suggests that the CSR procedure, in which seeking responding and consumption can be evaluated in the same session, is a valid model for evaluation of novel treatment medications.