Introduction

Top Abstract Introduction Methods Results Discussion References

Several of the pharmacotherapies that are currently used to treat opioid dependence and withdrawal have markedly different pharmacologies, although their primary effects are mediated by mu opioid receptors. Naltrexone is a mu-receptor selective antagonist that attenuates the behavioral effects of morphine-like opioid agonists (e.g., heroin). Naltrexone can decrease the use of opioids in patients who are highly motivated to remain drug abstinent; however, in many patients, the success of naltrexone is limited by lack of compliance (Renault, 1981). Other maintenance therapies, using mu receptor selective agonists (e.g., methadone), often decrease the use of other opioids by producing effects that are qualitatively similar to the drug of abuse (Kreek, 1992). Consequently, the primary pharmacological therapy in opioid dependent individuals consists of regular treatment with opioid agonists. The success of methadone for the treatment of opioid dependence is attributed to its oral bioavailability, long duration of action, and sufficient efficacy to attenuate and prevent withdrawal (for review see Kreek, 1992). More recently, LAAM has been approved as a maintenance therapy and its long duration of action prevents the emergence of withdrawal when administered every other or every third day (Ling et al., 1994). In addition, the opioid mixed agonist/antagonist buprenorphine (Negus et al., 1989; Young et al., 1984) is currently under evaluation as a maintenance therapy and purportedly has several advantages over other therapeutics including low efficacy (therefore a limited degree of toxic effects; Walsh et al., 1994) and a long duration of action. The long duration of action of buprenorphine can provide effective treatment when administered every other or every third day (Bickel and Amass, 1995). These pharmacologic therapies provide a variety of treatment options that can be catered to the specific needs of individuals.

Methadone is a mu agonist with a long duration of action and is the most commonly used pharmacotherapeutic for opioid dependence. When administered acutely, methadone and morphine have a similar duration of behavioral effects (e.g., analgesia); however, methadone accumulates with repeated dosing resulting in a prolonged duration of its behavioral effects (Kreek, 1992). In mouse brain, methadone has higher affinity for mu ([³H] dihydromorphine-labeled) binding sites than for either kappa ([³H] ethylketocyclazocine-labeled) or delta ([³H] D-ala²-leu⁵-enkephalin-labeled) binding sites (Neil, 1984). In nonhuman primates, methadone substitutes for the discriminative stimulus effects of other mu agonists such as codeine and morphine (Schaefer and Holtzman, 1977; Teal and Holtzman, 1980; Bertalmio et al., 1992) and, as with many other mu agonists, methadone maintains responding under i.v. self-administration procedures (Harrigan and Downs, 1978). Collectively, these studies demonstrate that the behavioral effects of methadone are mediated by mu opioid receptors.

LAAM is a congener of methadone and has a very long duration of mu agonist activity. Although both LAAM and methadone attenuate opioid withdrawal in human and nonhuman primates (Aceto et al., 1992; Kreek, 1992; Ling et al., 1994), LAAM has a longer duration of action than methadone. The formation of active metabolites and possibly other factors, such as drug-tissue binding and enterohepatic recycling, contribute to the long duration of action of LAAM (Finkle et al., 1982; Henderson et al., 1976). LAAM is N-demethylated into a secondary amine (nor-LAAM) and further N-demethylated into a primary amine (dinor-LAAM) with the two metabolites having longer plasma half-lives than the parent compound in both human and nonhuman primates (Henderson et al., 1976; Kaiko and Inturrisi, 1975; Mule and Misra, 1978). In decreasing electrically induced contractions of the guinea pig ileum (Nickander et al., 1974), as well as in behavioral studies in a number of species (Bertalmio et al., 1992; Holtzman, 1979; McGivney and McMillan, 1981), both of the primary metabolites of LAAM have potencies that are equal to or greater than the parent compound (Bertalmio et al., 1992; Holtzman, 1979; McGivney and McMillan, 1981). Moreover, nor-LAAM and dinor-LAAM have higher affinities than LAAM for mu ([³H] (D-Ala²-Me-Phe⁴, Glyol⁵) enkephalin [DAMGO]-labeled) binding sites in rhesus monkey cortex (Bertalmio et al., 1992). In rats and monkeys, LAAM, nor-LAAM and dinor-LAAM substitute for the discriminative stimulus effects of other mu agonists such as morphine and codeine (Bertalmio et al., 1992; Holtzman, 1979) and also maintain responding under i.v. self-administration procedures (Bertalimo et al., 1992; Harrigan and Downs, 1978; Young et al., 1978, 1979). Together, these data are consistent with the view that active metabolites contribute to the pharmacologic profile of LAAM.

Buprenorphine is a long-acting, low-efficacy mu agonist. In rhesus monkey cortex and the rat brain, buprenorphine binds to mu sites with equal to or greater affinity for kappa sites and with the least affinity for delta sites (Richards and Sadee, 1985; Sadee et al., 1982; Woods et al., 1992). Buprenorphine dissociates very slowly from mu receptors and, therefore, has a long duration of action (Hambrook and Rance, 1976); this slow dissociation produces conditions under which the behavioral effects of buprenorphine can be prevented, but not easily reversed, by competitive antagonists (France et al., 1984; Shannon et al., 1984). Buprenorphine has low efficacy at mu opioid receptors and has either agonist or antagonist effects, depending on the particular conditions under which it is evaluated. For example, buprenorphine has antinociceptive effects when a low (e.g., 50°C water) but not a high (e.g. 55°C water) intensity stimulus is used (Walker et al., 1995). Moreover, under conditions where buprenorphine does not have antinociceptive effects, it attenuates the antinociceptive effects of higher efficacy agonists (e.g., alfentanil; Walker et al., 1995; Woods et al., 1992). Buprenorphine also maintains responding under i.v. self-administration procedures (Mello et al., 1981; Winger et al., 1992), although under some conditions, rates of responding for buprenorphine are lower than rates maintained by higher efficacy agonists (Lukas et al., 1983; Young et al., 1984). Collectively, these studies support the notion that buprenorphine is a long-acting mu agonist with low efficacy.

Drug discrimination is a widely used procedure which is thought to be predictive of the subjective effects of drugs in humans (Holtzman, 1990; Preston and Bigelow, 1991). Drug discrimination procedures can be pharmacologically selective with subjects responding on the training drug-appropriate lever only after the administration of drugs that have effects that are similar to those of the training drug. In addition to simple two-choice discriminations between a drug and vehicle, discriminations can also be established between a drug and vehicle in subjects treated chronically with a second drug and can provide additional information relevant to the pharmacologic effects of the training drug. For example, the discriminative stimulus effects of naltrexone are not mediated by an opioid mechanism in morphine-naive subjects, whereas the discriminative stimulus effects of naltrexone are mediated by an opioid mechanism in morphine-treated subjects (France and Morse, 1986; Valentino et al., 1983; Weissman, 1978). Under the latter condition, it has been postulated that subjects discriminate between a withdrawal condition and a dependent condition because antagonist lever responding increases in either a dose-related manner after the administration of an antagonist (precipitated withdrawal) or a time-related manner after the discontinuation of agonist treatment (deprivation induced withdrawal; for reviews, see Emmett-Oglesby et al., 1990; France, 1994). In nonhuman subjects, drug discrimination procedures appear to provide suitable conditions for characterizing the behavioral effects of drugs, in general, as well as for identifying relative drug efficacies and evaluating drug interactions in a manner that cannot be used in humans.

The current study examined the discriminative stimulus effects of LAAM, buprenorphine and methadone in rhesus monkeys receiving 3.2 mg/kg/day of morphine and discriminating between s.c. injections of saline and 0.01 mg/kg of naltrexone; these conditions are sensitive to both mu agonists and mu antagonists (France and Woods, 1989; France et al., 1990). One purpose of the current study was to systematically compare and contrast the durations of action, potencies and relative efficacies of methadone, LAAM and buprenorphine using drug discrimination procedures. In addition, drug interactions were assessed to determine whether the discriminative stimulus effects of morphine or naltrexone were modified by methadone, LAAM or buprenorphine. Finally, the durations of the discriminative stimulus effects of nor-LAAM and dinor-LAAM were studied to determine the relative contribution of these metabolites to the discriminative stimulus effects of the parent compound, LAAM.

	Methods

Top Abstract Introduction Methods Results Discussion References

Subjects

One male (HI) and five female adult rhesus monkeys (Macaca mulatta) weighing between 5.4 and 10.2 kg were individually housed in stainless-steel cages with free access to food (Teklad monkey chow, Madison, WI) and water; monkeys received fresh fruit and peanuts several times each week and were maintained under a 14-hr light/10-hr dark schedule. Five monkeys (WI, GR, LO, PI and DA) had been treated with morphine for longer than 7 yr and the other monkey (HI) had been treated with morphine for longer than 4 mo; all monkeys had received other opioid and nonopioid drugs in previous studies.

Apparatus

Monkeys were seated in either metal or Lexan primate chairs within ventilated, sound-attenuating chambers that contained two or three response levers, a food cup and an array of red stimulus lights. Each chair was equipped with a pair of shoes containing brass electrodes to which brief (250 msec) electric shocks (3 mA) could be delivered from a remote a.c. shock generator. Control of experiments and data recording were accomplished with a microprocessor, interface and commercially available software (MedAssociates, Inc, St. Albans, VT).

Behavioral Procedure

Training. Monkeys received 3.2 mg/kg (s.c.) of morphine 3 hr before daily sessions and discriminated between s.c. injections of 0.01 mg/kg of naltrexone and saline. Daily sessions consisted of multiple (2-6), discrete, 15-min cycles with each cycle comprising a 10-min timeout, during which the chamber was dark and responses had no programed consequence, followed by a 5-min response period, during which monkeys responded under a FR-5 schedule of stimulus-shock termination. Under these conditions, shocks were scheduled to occur every 15 sec in the presence of red stimulus lights. Monkeys could extinguish the stimulus lights and avoid the scheduled shocks for 30 sec by completing five consecutive responses on the correct (i.e., injection-appropriate) lever, as determined by an injection (saline or naltrexone) administered during the first minute of the time out; the right lever was correct after naltrexone, and the left lever was correct after saline for three of the monkeys and the lever orientation was reversed for the other monkeys. Responses on the injection-inappropriate lever reset the FR requirement on the injection-appropriate lever. Two of the chambers contained a third lever that was not associated with stimulus lights; responses on this lever always reset the response requirement on the other levers. The response period ended after 5 min had elapsed or four shocks had been delivered, whichever occurred first; in the latter case, stimulus lights were extinguished for the remainder of the 5-min period. Training sessions consisted of a saline injection or a "sham" injection (i.e., the chamber door was opened and closed without an injection) during each of 0 to 6 cycles; naltrexone was administered on the last or second to the last cycle on selected days.

Testing. Test sessions were identical to training sessions except that monkeys could extinguish the stimulus lights and avoid scheduled shocks by completing five consecutive responses on either lever. Drugs were compared for their relative durations of action in reversing naltrexone lever responding in monkeys acutely deprived of morphine. Under these conditions, saline was substituted for the daily dose of morphine (i.e., 27-hr morphine deprivation) and a single dose of LAAM, nor-LAAM, dinor-LAAM, methadone or morphine was administered at the beginning of the first cycle followed by seven "sham" cycles.

Data analyses. Results of drug discrimination studies are presented as the average percentage of responses on the naltrexone lever (i.e., % DR) ± 1 SEM plotted as a function of either dose or time. The % drug responding was not included for an individual monkey either when the response rate was less than 20% of the average response rate of the five previous saline training cycles in which the testing criteria had been attained, or when more than one shock was delivered (this restriction only occurred when doses of naltrexone larger than 0.0032 mg/kg were studied in combination with methadone). In morphine-treated monkeys, test compounds were considered to have substituted for the naltrexone discriminative stimulus when responding was at least 90% on the naltrexone-appropriate lever. In morphine-deprived monkeys, test compounds were considered to have substituted for the morphine discriminative stimulus (i.e., reversed responding on the naltrexone lever) when no more than 10% of the responses were on the naltrexone-appropriate lever. To estimate relative potencies among drugs for reversing naltrexone lever responding in monkeys acutely deprived of morphine, maximal decreases in naltrexone lever responding for individual monkeys (over the two hour session; Fig 2) were averaged (± 1 SEM) and plotted as a function of dose (Fig 3). ED₅₀ values were calculated for individual subjects by linear regression when three or more data points were available and by interpolation when only two data points (one above and one below 50% drug responding) were available, and these values were averaged among subjects. Single-dose apparent affinity estimates for naltrexone in combination with either morphine or methadone were calculated with a modified equation of Tallerida et al. (1979) where pK_B = -log [B/(dose ratio-1)]; the variable "B" represents moles/kg of body weight. Over the 5 yr during which these studies were conducted, three monkeys (WI, LO and DA) were unable to complete the studies because of health problems that were unrelated to the current experiment; the number of monkeys for each study is indicated in the figure legends.

Drugs

The compounds studied were buprenorphine hydrochloride, LAAM, nor-LAAM, dinor-LAAM, methadone hydrochloride, morphine sulfate and naltrexone hydrochloride. All compounds were generously provided by the Research Technology Branch, National Institute on Drug Abuse, Rockville, MD. Compounds were dissolved in sterile water and injected s.c. in a volume of 0.1 to 4.0 ml. Doses are expressed in mg/kg of body weight in terms of the salt form. Doses of morphine of more than 100.0 mg/kg were not studied due to potential toxic effects.

	Results

Top Abstract Introduction Methods Results Discussion References

When 3.2 mg/kg of morphine was administered 3 hr before experimental sessions, monkeys responded on the saline lever after the administration of saline (fig. 1; square above SAL). Naltrexone dose dependently increased responding on the drug lever with more than 90% drug-appropriate responding occurring at the training dose of 0.01 mg/kg of naltrexone. When saline was administered 3 hr before experimental sessions (i.e., 27-hr morphine deprivation), monkeys responded on the naltrexone lever after the administration of saline (fig. 1; circle above SAL); under this condition, morphine dose dependently decreased naltrexone lever responding with less than 10% drug-appropriate responding occurring at a dose of 5.6 mg/kg of morphine. These doses of naltrexone and morphine did not modify rates of responding (data not shown).

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Discriminative stimulus effects of naltrexone and morphine in rhesus monkeys (n = 6) treated with 3.2 mg/kg/day of morphine and discriminating between 0.01 mg/kg of naltrexone and saline. Under one condition, a dose of 3.2 mg/kg of morphine was administered 3 hr before cumulative doses of naltrexone (squares). Under a second condition, saline was substituted for the daily doses of morphine 3 hr before cumulative doses of morphine (circles). Under both conditions, saline was administered on the first cycle (shaded symbols above SAL). Ordinate: average percentage of responses on the naltrexone key (% DR) ± 1 SEM. Abscissa: dose in mg/kg of body weight.

To determine the duration of the discriminative stimulus effects of morphine, methadone, LAAM, nor-LAAM and dinor-LAAM, single doses of these drugs were administered to monkeys acutely deprived of morphine. A small dose of morphine (0.1 mg/kg) did not decrease naltrexone lever responding in any of the monkeys (fig. 2; left top panel). Larger doses of morphine (0.32 and 1.0 mg/kg) dose dependently decreased naltrexone lever responding. The onset for morphine in decreasing naltrexone responding was slow with the peak effect occurring approximately 1 hr after administration; naltrexone lever responding increased later in the session. A dose of 3.2 mg/kg of morphine partially (26%) decreased naltrexone lever responding at 0.25 hr and maximally decreased naltrexone lever responding in all monkeys for the remainder of the 2-hr session; responding gradually shifted back to the naltrexone lever over a 24-hr period.

View larger version (45K): [in this window] [in a new window]   Fig. 2.   Duration of action of single doses of morphine (n = 4), methadone (n = 4), LAAM (n = 3-6), nor-LAAM (n = 3) and dinor-LAAM (n = 3) for reversing naltrexone lever responding in monkeys acutely deprived of morphine. Abscissa: time (hr) after the administration of drug. See figure 1 or additional details.

The potency of methadone (fig. 2; right top panel) for reversing naltrexone lever responding was similar to morphine. Doses of 0.32 and 1.0 mg/kg of methadone partially decreased naltrexone lever responding. A larger dose of methadone (3.2 mg/kg) decreased naltrexone lever responding to less than 10% for longer than 2 hr; responding gradually shifted to the naltrexone lever over an 8-hr period. Rates of responding were not altered up to a dose of 3.2 mg/kg of morphine or methadone (data not shown).

Injections of LAAM, nor-LAAM or dinor-LAAM dose dependently decreased naltrexone lever responding in monkeys acutely deprived of morphine. A dose of 0.32 mg/kg of LAAM did not reverse naltrexone lever responding in any of three monkeys over the 2-hr session (fig. 2; middle left panel). At times longer than 1.25 hr after the administration of a dose of 1.0 mg/kg of LAAM, naltrexone-lever responding was only partially decreased. Although 3.2 and 5.6 mg/kg of LAAM decreased naltrexone lever responding, these effects were not maximal until 45 to 60 min after drug administration. There were no discriminative stimulus effects evident 24 hr after the administration of 1.0 mg/kg of LAAM in any of six monkeys. A dose of 3.2 mg/kg of LAAM reversed naltrexone lever responding in only one of the six monkeys 24 hr after administration. In all five monkeys tested, a dose of 5.6 mg/kg of LAAM reversed naltrexone lever responding for more than 24 hr, and in one monkey this dose of LAAM fully reversed naltrexone lever responding for more than 48 hr, but not more than 72 hr (data not shown). Two hr after the administration of 5.6 mg/kg of LAAM, response rates were not different from response rates after saline (data not shown). Twenty-four hr after the administration of 5.6 mg/kg of LAAM, the average response rate was decreased to 66% of the average control response rate and monkeys displayed signs of intoxication (i.e., mydriasis, hypolocomotion and decreased frequency of ventilation) that were more obvious than signs displayed 2 hr after LAAM administration. A dose of 10.0 mg/kg of LAAM fully reversed naltrexone lever responding in two monkeys (data not shown); approximately 12 hr after the administration of this dose of LAAM, one monkey displayed signs of intoxication and the other monkey was unresponsive to physical stimulation. Naltrexone reversed the condition of the latter monkey. Due to the potential for toxic effects, doses of LAAM larger than 5.6 mg/kg were not studied in other monkeys.

Nor-LAAM dose-dependently decreased naltrexone lever responding (fig. 2; middle right panel). A dose of 0.032 mg/kg of nor-LAAM did not modify responding, whereas doses of 0.1 and 0.32 mg/kg partially reversed responding on the naltrexone lever. A dose of 0.56 mg/kg of nor-LAAM completely reversed responding on the naltrexone lever with the maximum effect occurring 45 to 60 min after administration; this dose of nor-LAAM maximally reversed naltrexone lever responding for up to 4 hr and partially 8 hr after administration.

Similarly, dinor-LAAM dose dependently reversed naltrexone lever responding (fig. 2; left bottom panel). A dose of 0.1 mg/kg of dinor-LAAM did not attenuate naltrexone lever responding, whereas doses of 0.032 and 1.0 mg/kg partially decreased naltrexone lever responding. A larger dose (1.78 mg/kg) maximally reversed naltrexone lever responding for up to 4 hr and partially 8 hr after administration. Rates of responding were not modified by nor-LAAM or dinor-LAAM (data not shown).

The maximal effect obtained for each dose of drug in individual monkeys was averaged and is plotted as a function of dose in figure 3. The ED₅₀ for morphine was 0.69 ± 0.37 mg/kg and the ED₅₀ for methadone was 1.34 ± 0.39 mg/kg. The ED₅₀ for LAAM was 1.69 ± 0.51 mg/kg; nor-LAAM and dinor-LAAM were 20- and 5-fold more potent than LAAM with ED₅₀s of 0.09 ± 0.04 and 0.39 ± 0.17 mg/kg, respectively.

View larger version (28K): [in this window] [in a new window]   Fig. 3.   Potency of morphine (n = 4), methadone (n = 3), LAAM (n = 5), nor-LAAM (n = 3) and dinor-LAAM (n = 3) for reversing naltrexone lever responding in monkeys acutely deprived of morphine. See figure 1 and "Methods" for additional details.

Naltrexone attenuated the discriminative stimulus effects of morphine and methadone. Monkeys acutely deprived of morphine responded on the naltrexone lever after the administration of saline (fig. 4; open circle above SAL) and a dose of 0.1 mg/kg of methadone. Three monkeys switched their responding predominantly to the saline lever after receiving a cumulative dose of 3.2 mg/kg of methadone; another subject responded 17% on the naltrexone lever with this dose of methadone and exclusively on the saline lever with a larger dose (10.0 mg/kg) of methadone. These doses of methadone alone did not modify response rates, whereas doses of methadone larger than 10.0 mg/kg substantially decreased responding, and in some monkeys caused respiratory depression that was reversed by naltrexone. Morphine-deprived monkeys responded on the naltrexone lever after receiving a dose of 0.0032 mg/kg of naltrexone (shaded circle above NTX); this dose of naltrexone shifted the methadone dose-effect curve 2-fold to the right. Single dose apparent-affinity estimates (pK_B) for naltrexone in combination with methadone were 8.39, 8.41 and 8.41 for individual monkeys (average pK_B = 8.40); this dose of naltrexone did not shift the methadone dose-effect curve in a fourth monkey. In the presence of a larger dose of naltrexone (e.g., 0.032 mg/kg) methadone failed to reverse responding on the naltrexone lever up to doses of methadone (32.0 mg/kg) that decreased rates of responding and produced tremors and muscle twitches. Because the rate-decreasing effects of methadone were not antagonized by naltrexone, larger doses of naltrexone were not studied in combination with methadone. Similar to results obtained with methadone, a dose of 0.0032 mg/kg of naltrexone shifted the morphine dose-effect curve 2-fold to the right. Single dose apparent affinity estimates for naltrexone in combination with morphine were 7.95, 8.03 and 8.13 for individual monkeys (average pK_B = 8.04).

View larger version (31K): [in this window] [in a new window]   Fig. 4.   Dose effect curves for the discriminative stimulus effects of cumulative doses of methadone (n = 4) and morphine (n = 3), each administered alone and in combination with 0.0032 mg/kg of naltrexone in monkeys acutely deprived of morphine. A dose of 0.0032 mg/kg of naltrexone (NTX) or saline (SAL) was administered on the first cycle. See figure 1 for additional details.

Naltrexone also antagonized the discriminative stimulus effects of LAAM. Doses of 0.01 and 1.0 mg/kg of naltrexone fully attenuated the discriminative stimulus effects of an acute injection of 5.6 mg/kg of LAAM for the duration of the 2-hr session (fig. 5). The longer duration of action of LAAM, as compared to naltrexone, was evidenced by a partial reversal of naltrexone lever responding 24 hr after the administration of combinations of LAAM and naltrexone.

View larger version (29K): [in this window] [in a new window]   Fig. 5.   Discriminative stimulus effects of 5.6 mg/kg of LAAM alone and in combination with either 0.01 or 1.0 mg/kg of naltrexone (n = 3) in monkeys acutely deprived of morphine. See figures 1 and 2 for additional details.

LAAM and methadone potentiated the discriminative stimulus effects of morphine in monkeys acutely deprived of morphine. Twenty-four hr after the administration of 3.2 mg/kg of either LAAM or methadone, monkeys responded predominantly on the naltrexone lever (fig. 2); the morphine dose-effect curve determined under either of these conditions was shifted 3-fold to the left of the respective control dose-effect curve (data not shown). Morphine dose-effect curves redetermined 24 hr after either 0.56 mg/kg of nor-LAAM or 1.78 mg/kg of dinor-LAAM were not different from the morphine dose-effect curve determined under control conditions (data not shown). Monkeys acutely deprived of morphine responded on the naltrexone lever 1 hr after the administration of 1.0 mg/kg of methadone (fig. 6; triangle above SAL). This dose of methadone shifted the morphine dose-effect curve 4-fold to the left as compared to the morphine dose-effect curve determined under control conditions. Similarly, monkeys acutely deprived of morphine responded on the naltrexone lever 2 hr after the administration of 1.0 mg/kg of LAAM (square above SAL). This dose of LAAM shifted the morphine dose-effect curve 18-fold to the left of the morphine dose-effect curve obtained under control conditions.

View larger version (27K): [in this window] [in a new window]   Fig. 6.   Discriminative stimulus effects of cumulative doses of morphine administered alone (n = 4) and either 2 hr after the administration of 1.0 mg/kg of LAAM (n = 4) or 1 hr after the administration of 1.0 mg/kg of methadone (n = 3) in monkeys acutely deprived of morphine. See figure 1 for additional details.

Buprenorphine had morphine-like discriminative stimulus effects in some monkeys and naltrexone-like discriminative stimulus effects in other monkeys. Buprenorphine fully (i.e.,  90% DR) substituted for the naltrexone discriminative stimulus in three morphine-treated monkeys (WI, GR and PI; fig. 7, left panel). Doses of 0.32 and 1.0 mg/kg of buprenorphine substituted for naltrexone in one of two tests in a fourth monkey (DA); buprenorphine never substituted in a fifth monkey (LO). When the same doses of buprenorphine were studied in the same monkeys acutely deprived of morphine, buprenorphine did not attenuate naltrexone lever responding in three monkeys (WI, GR and PI; fig. 7, right panel). In contrast, one monkey (LO) responded partially (35-60%) on the naltrexone lever after these same doses of buprenorphine; partial drug lever responding consisted of responding on both the saline and naltrexone levers during both tests.

View larger version (23K): [in this window] [in a new window]   Fig. 7.   Discriminative stimulus effects of cumulative doses of buprenorphine in individual morphine-treated monkeys (WI, GR, PI, LO and DA; left panel) or individual morphine-deprived monkeys (WI, GR, PI and LO; right panel). Each dose-effect curve is the average of two determinations in each monkey; SEM are not included in the left panel for clarity. See figure 1 for additional details.

The antagonist effects of buprenorphine had a long duration of action in monkeys (GR and PI) in which buprenorphine had naltrexone-like discriminative stimulus effects. Dose-effect curves for morphine that were determined 1, 6 or 10 days after the administration of 3.2 mg/kg of buprenorphine are shown in figure 8. Up to a dose of 100.0 mg/kg of morphine, naltrexone lever responding was decreased to only 89% one day after the administration of buprenorphine. The morphine dose-effect curve was shifted 3-fold to the right of the control (pre-buprenorphine) morphine dose-effect curve 6 days after the administration of buprenorphine and was similar to the control (pre-buprenorphine) morphine dose-effect curve 10 days after the administration of buprenorphine.

View larger version (30K): [in this window] [in a new window]   Fig. 8.   Time course of the antagonist effects of 3.2 mg/kg of buprenorphine in monkeys (GR and PI) that responded on the naltrexone lever after the administration of buprenorphine alone. On three different occasions, a morphine dose-effect curve was determined 1, 6 or 10 days after a test with buprenorphine. See figure 1 and "Methods" for additional details.

Similarly, the agonist effects of buprenorphine had a long duration of action in monkeys (DA and LO) in which buprenorphine had some morphine-like effects (i.e., buprenorphine did not substitute for naltrexone and partially reversed naltrexone lever responding). For 1 and 2 day(s) after the administration of a small dose of buprenorphine (0.1 mg/kg), monkeys acutely deprived of morphine responded predominantly on the saline lever (16% DR; fig. 9, diamond and square above SAL). Three days after the administration of buprenorphine, monkeys responded an average of 78% on the naltrexone lever after the administration of saline and the morphine dose-effect curve redetermined on that day was shifted to the left of the morphine dose-effect curve determined under control conditions. In combination with buprenorphine, a 32-fold smaller dose of morphine was required to fully reverse responding on the naltrexone lever compared to the dose required to fully reverse responding on the naltrexone lever before the administration of buprenorphine.

View larger version (24K): [in this window] [in a new window]   Fig. 9.   Time course of the agonist effects of 0.1 mg/kg of buprenorphine in monkeys (DA and LO) that responded on the saline lever after the administration of buprenorphine alone. Cumulative doses of morphine were administered after the first cycle on the third day after the administration of buprenorphine. See figure 1 and "Methods" for additional details.

In monkeys in which buprenorphine had some morphine-like effects (DA and LO), large doses of naltrexone were required to increase responding on the naltrexone lever after the administration of buprenorphine. Twenty-four hr after the administration of a cumulative dose of 3.2 mg/kg of buprenorphine (same test as fig. 7; left panel), both morphine-treated monkeys responded on the saline lever after the administration of saline (data not shown). This dose of buprenorphine shifted the naltrexone dose-effect curve an average of 52-fold to the right of the naltrexone dose-effect curve determined before the administration of buprenorphine. In these two monkeys, more than 89% responding on the naltrexone lever occurred at doses of either 0.56 or 1.0 mg/kg of naltrexone (data not shown).

	Discussion

Top Abstract Introduction Methods Results Discussion References

In morphine-treated rhesus monkeys, methadone, LAAM, nor-LAAM and dinor-LAAM had behavioral effects that were qualitatively similar to the behavioral effects produced by other mu agonists. Similar to the effects obtained with the mu agonists alfentanil and nalbuphine under identical conditions (France and Woods, 1989; France et al., 1990), naltrexone lever responding was reversed by methadone, LAAM, nor-LAAM and dinor-LAAM. The specificity of mu agonists in reversing naltrexone lever responding under these conditions has been demonstrated previously in studies showing that naltrexone lever responding was not attenuated by agonists that act at kappa (France and Woods, 1989) or delta (France, unpublished data) opioid receptor types or by nonopioids (France and Woods, 1989; Holtzman, 1985). Moreover, the relative potencies of drugs in reversing responding on the naltrexone lever in the current study were similar to their relative potencies under other conditions. For example, nor-LAAM and dinor-LAAM were 20- and 5-fold more potent, respectively, than LAAM in reversing naltrexone lever responding. Similarly, nor-LAAM was 30-fold more potent than LAAM in producing codeine-appropriate responding and dinor-LAAM had a potency intermediate to nor-LAAM and LAAM (Bertalmio et al., 1992). In the current study, morphine and methadone were equipotent in reversing naltrexone lever responding. Similar to results obtained under other behavioral conditions, morphine and methadone were equipotent in maintaining i.v. self-administration responding and in substituting for the codeine discriminative stimulus (Bertalmio et al., 1992; Harrigan and Downs, 1978). These results confirm mu agonist actions for the discriminative stimulus effects of methadone, LAAM, nor-LAAM and dinor-LAAM.

Receptor selective antagonists have been used to identify specific receptor types that mediate the discriminative stimulus effects of agonists. For example, apparent affinity estimates (i.e., pA₂ and pK_B) for receptor-selective opioid antagonists (e.g., naltrexone) are used to differentiate agonist effects that are mediated through mu, kappa or delta receptors (Bertalmio and Woods, 1987; Comer et al., 1993; Negus et al., 1993). In the current study, the average single-dose apparent affinity estimates for naltrexone in combination with methadone and morphine were 8.4 and 8.0, respectively; these values are similar to apparent affinity estimates for naltrexone in combination with these agonists in other discrimination studies (morphine 8.3, France et al., 1990; methadone 8.4, Gerak and France, 1996), for naltrexone in combination with other mu agonists, such as alfentanil (8.7; France et al., 1990) and nalbuphine (8.1; Gerak and France, 1996), and are different from apparent affinity estimates for naltrexone in combination with kappa agonists such as enadoline (6.1; Butelman et al., 1993). In a previous study, a dose of 0.032 mg/kg of naltrexone shifted the dose-effect curves for the discriminative stimulus effects of morphine or alfentanil approximately 32-fold to the right (France et al., 1990). In the current study, comparable shifts in the dose-effect curve for the discriminative stimulus effects of methadone could not be determined with this dose of naltrexone, perhaps because naltrexone failed to attenuate the rate-decreasing effects of methadone. An inability of naltrexone to shift the discriminative stimulus effects of methadone further to the right also has been observed in monkeys discriminating between saline and nalbuphine (Gerak and France, 1996), further suggesting that methadone has some behavioral effects that are not mediated by opioid receptors.

Although many of the behavioral effects of LAAM are similar to other mu agonists, LAAM has a unique pharmacology which is thought to make it especially well suited for the treatment of opioid abuse (i.e., longer duration of action than other mu agonists). The long duration of LAAM is thought to be due to the formation of active metabolites (Finkle et al., 1982; Henderson et al., 1977; Kaiko and Inturrisi, 1975). A comparison of equi-effective doses (i.e., doses that decreased naltrexone lever responding to less than 10%) of LAAM, nor-LAAM and dinor-LAAM demonstrated that the duration of action of LAAM was at least three times longer than either of its metabolites. Other studies in monkeys have shown that, after an oral dose of 2.0 mg/kg of LAAM, the plasma half-lives of LAAM, nor-LAAM and dinor-LAAM are 6.9, 24.3 and more than 96 hr, respectively (Mule and Misra, 1978). In comparison, after an oral dose of 0.5 mg/kg of nor-LAAM, the plasma half-live of nor-LAAM (9.9 hr) is approximately 2.5 times shorter than its half-life after the administration of LAAM (Misra et al., 1980) suggesting that the long duration of action of LAAM is due to a slow biotransformation to active metabolites with intermediate durations of action rather than to a rapid biotransformation of LAAM to active metabolites with very long durations of action. Although the potency difference between oral and s.c. LAAM has not been established, data from other studies (Bertalmio et al., 1992; Billings et al., 1974; Henderson et al., 1977; Holtzman, 1985; Kaiko and Inturrisi, 1975) together with the current study, support the notion that it is the rate of LAAM metabolism and subsequent formation of nor-LAAM and dinor-LAAM that determines the long duration of behavioral effects of LAAM.

Numerous studies have shown significant individual differences in the onset and peak plasma concentrations of nor-LAAM and dinor-LAAM in both human and nonhuman primates (Billings et al., 1974; Henderson et al., 1977; Kaiko and Inturrisi, 1975; Misra et al., 1976; Mule and Misra, 1978). Although differences in plasma levels of a drug might not correspond to differences in its behavioral effects, many studies have shown considerable differences among subjects in the toxic effects of LAAM, even among subjects receiving the same dose of LAAM over long periods of time (Aceto et al., 1992; Crowley et al., 1985; Downs, 1979; Downs et al., 1977; Henderson et al., 1977; Moerschbaecher, et al., 1983). Similarly, in the current study the behavioral effects of LAAM were variable among individuals. For example, a dose of 5.6 mg/kg of LAAM reversed naltrexone lever responding for more than 48 hr in one monkey, whereas this dose of LAAM reversed naltrexone lever responding for just 24 hr in the other monkeys. Differences in the sensitivities of monkeys to the behavioral effects of other mu agonists did not predict differences in the behavioral effects of LAAM. Not only was there greater variability in the behavioral effects of LAAM among subjects compared to other mu agonists, the margin of safety appeared smaller with LAAM compared to other mu agonists. For example, a dose of LAAM just 0.25 log unit larger than a dose required to fully reverse naltrexone lever responding (5.6 mg/kg) produced substantial sedation in monkeys. In comparison, doses of other drugs (e.g., morphine and alfentanil) more than 1 log unit larger than doses required to fully reverse naltrexone lever responding are required to produce any sedation in these monkey (current study; France and Woods, 1989). Together with previous studies, the current study suggests that differences in the behavioral effects of LAAM among subjects might be related to individual differences in the rate of LAAM metabolism. Moreover, these results suggest that in vivo, drugs that depend on the formation of active metabolites to produce behavioral effects have greater variability among subjects and smaller margins of safety than drugs that do not require active metabolites to produce behavioral effects. These results in monkeys are similar to studies in humans that have reported striking individual variability in the behavioral effects of LAAM and suggest that maintenance doses of LAAM in humans need to be carefully tailored to individuals because a dose of LAAM which is effective for treating opioid dependence in some might be ineffective or even toxic in others.

The behavioral effects of opioid agonists administered in combination can differ depending on drug history (Craft and Dykstra, 1992; Young et al., 1991). Under the dosing conditions of the current study, LAAM and methadone potentiated the discriminative stimulus effects of morphine, as evidenced by leftward shifts in the morphine dose-effect curve. Similarly, methadone also potentiates the antinociceptive effects of morphine under a shock titration procedure in squirrel monkeys (Craft and Dykstra, 1992). Thus, it is possible that other effects of opioids (e.g., heroin), such as respiratory depression, might also be potentiated by methadone or LAAM in subjects that are not tolerant to opioids. In contrast to the acute dosing conditions in the current study, LAAM and methadone might not be expected to potentiate the behavioral effects of morphine under chronic dosing conditions. For example, when methadone or LAAM is used as a maintenance therapy for long periods of time, cross-tolerance would be expected to develop to the behavioral effects of other mu agonists. Consistent with this notion, LAAM and methadone have been shown to attenuate the subjective effects of other opioids in humans, presumably because of the development of tolerance to LAAM or methadone and cross-tolerance to other opioids (e.g., heroin: Kreek, 1992; Levine et al., 1973). These results emphasize that the behavioral effects of drugs can be attenuated or potentiated depending on whether an individual is tolerant or not tolerant to other pharmacologically related drugs.

The behavioral effects of drugs can also be modified by the level of dependence. In the current study, buprenorphine had morphine-like discriminative stimulus effects in some monkeys and naltrexone-like discriminative stimulus effects in other monkeys. In previous studies, buprenorphine had either mu agonist or mu antagonist effects, depending on the treatment conditions. Under conditions requiring a low degree of receptor stimulation to produce behavioral effects (i.e., when there were many spare receptors), buprenorphine often has agonist effects; for example, buprenorphine substitutes for etorphine (a mu agonist) in monkeys discriminating between etorphine and saline (Young et al., 1984). In contrast, under conditions requiring a high degree of receptor stimulation to produce behavioral effects (i.e., when there were few or no spare receptors), buprenorphine has antagonist effects; for example, buprenorphine precipitates a naltrexone-like withdrawal in morphine-treated monkeys (3.0 mg/kg/6 hr of morphine) (Fukase et al., 1994; Gmerek, 1984).

In contrast to studies where buprenorphine had either agonist or antagonist effects, in the current study, buprenorphine had both agonist and antagonist effects, a finding that is consistent with the view that the behavioral effects of buprenorphine are related to the level of opioid dependence. Similarly, under conditions of low dependence in humans, buprenorphine has agonist effects, whereas, under conditions of high dependence, buprenorphine has antagonist effects (for review see Bickel and Amass, 1995). Although the agonist or antagonist effects of buprenorphine appear to be related to the level of opioid dependence, it is not clear why buprenorphine had qualitatively different effects in monkeys treated identically with morphine. There was no evidence, for example, that these monkeys had substantially different sensitivities to agonist or antagonist actions of opioids. The ED₅₀ values for naltrexone among these monkeys varied by only 3-fold (range was 0.002-0.006 mg/kg) and monkeys DA and LO were only 3-fold more sensitive than monkeys PI and GR to the discriminative stimulus effects of morphine (compare control morphine dose-effect curves from figs. 8 and 9). The ED₅₀ values for other agonists varied by less than 2-fold among these monkeys. These results suggest that the sensitivity of subjects to the behavioral effects of mu agonists, as indicated by the potency of agonists and antagonists, does not fully predict the behavioral effects of buprenorphine. The individual variability observed in the behavioral effects of buprenorphine in morphine-treated monkeys confirms the difficulty in predicting the behavioral effects of buprenorphine in opioid-dependent humans.

Although buprenorphine had either agonist or antagonist effects in individual monkeys, like LAAM, its behavioral effects were long lasting. Twenty-four hr after the administration of a dose of 3.2 mg/kg of buprenorphine, the morphine and naltrexone dose-effect curves were shifted more than 40-fold to the right of the respective control dose-effect curves. The durations of these behavioral effects were similar as evidenced by the rate at which sensitivity returned to the discriminative stimulus effects of morphine or naltrexone. These long durations of behavioral effects of buprenorphine are consistent with other studies in pigeons (France et al., 1984), monkeys (Liguori et al., 1996; Walker et al., 1995) and humans (Bickel and Amass, 1995; Jasinski et al., 1978; Rosen et al., 1994). One theory of opioid receptors is that the mu receptor has different affinities for ligands depending on the tertiary conformation of the receptor (for a review, see Taylor, 1990). Agonists preferentially bind to a low affinity state of the receptor whereas antagonists bind to both a low and a high affinity state of the receptor (Carroll et al., 1988). That the duration of both the agonist and antagonist effects of buprenorphine were similar in the current study suggests that qualitative differences in the behavioral effects of buprenorphine among subjects was not due to variations in receptors (numbers or state) among individuals.

The behavioral effects of buprenorphine in combination with morphine varied depending on the discriminative stimulus effects of buprenorphine in individual monkeys. Buprenorphine antagonized the effects of morphine in some monkeys (those that responded on the naltrexone lever after receiving buprenorphine) and potentiated the effects of morphine in other monkeys (those that responded on the naltrexone lever after receiving buprenorphine). Attenuation of the discriminative stimulus effects of morphine by buprenorphine in monkeys is consistent with studies in humans where the positive subjective effects of high efficacy mu agonists are attenuated by the acute (Walsh et al., 1995) or chronic (Bickel et al., 1988; Jasinski et al., 1978; Rosen et al., 1994) administration of buprenorphine. Collectively, these studies provide evidence for the low efficacy of buprenorphine which appears to confer a high margin of safety even after buprenorphine is administered in combination with other higher-efficacy opioids. That is, under conditions in which low efficacy agonists do not have maximal effects or produce antagonistic actions, they can be shown to attenuate the effects of higher efficacy agonist (Kenakin, 1993). Similar to other low efficacy agonists (i.e., nalbuphine) (Gerak et al., 1994), buprenorphine has limited respiratory depressant effects and doses larger than those that maximally decrease respiration have no further effect (Liguori et al., 1996; Walsh et al., 1994). In the current study, a dose of 3.2 mg/kg of buprenorphine antagonized the discriminative stimulus effects of morphine and, by virtue of the ability to safely administer what otherwise would be a toxic dose of morphine (i.e., 100.0 mg/kg), antagonized the respiratory depressant effects of morphine. Buprenorphine also antagonizes the respiratory depressant effects of the mu agonist levorphanol in monkeys (Liguori et al., 1996). These results extend previous studies demonstrating that buprenorphine is a particularly safe compound, even in combination with other opioids, and might be an especially useful addition to current opioid maintenance therapies in individuals with a low level of opioid dependence.