Introduction

Top Abstract Introduction Methods Results Discussion References

The frequent administration of some drugs often results in the development of dependence that can be quantified by the severity of withdrawal that occurs after either the discontinuation of drug treatment or the administration of a pharmacologic antagonist. Of particular clinical relevance are the adverse signs and symptoms (i.e., withdrawal) that often emerge on the discontinuation of drug use and that can contribute to the continued drug use. Moreover, individuals often report an increased motivation to obtain and use drugs during periods of withdrawal (Schuster et al., 1995) that could contribute significantly to the maintenance or re-initiation of drug use.

LAAM is a mu opioid agonist that is used as a maintenance therapy in opioid-dependent patients. In human (Kreek, 1992; Ling et al., 1994) as well as non-human primates (Aceto et al., 1992), LAAM attenuates opioid withdrawal and has a long duration of action that effectively prevents the emergence of withdrawal when it is administered as infrequently as thrice weekly. Two active metabolites of LAAM, l-alpha acetylnormethadol (i.e., nor-LAAM) and l-alpha acetyldinormethadol (i.e., dinor-LAAM), appear to contribute to its long duration of action (Finkle et al., 1982; Henderson et al., 1977) and, in many species, these metabolites have behavioral effects that are similar to the effects of other mu agonists and potencies that are equal to or greater than the parent compound, LAAM (Bertalmio et al., 1992; Holtzman, 1979; McGivney and McMillan, 1981). The combination of mu agonist activity and a long duration of action appears to account for the relative success of LAAM as a maintenance therapy for opioid dependence.

This study assessed the behavioral effects of chronic LAAM administration in rhesus monkeys, a species that has been used extensively to study opioid dependence and withdrawal (Aceto et al., 1992; Downs et al., 1977; Holtzman and Villarreal, 1973). In non-human primates, opioid dependence and withdrawal are typically assessed using three general approaches. The first approach consists of observational procedures where unconditioned behavioral signs of withdrawal are rated after either the discontinuation of chronic drug treatment (i.e., deprivation-induced withdrawal; Aceto et al., 1992) or the administration of a pharmacologic antagonist (i.e., precipitated withdrawal; Gmerek and Woods, 1985; Gmerek et al., 1987). Generally, withdrawal signs are pharmacologically specific for a class of drugs and are typically reversed only by drugs that have pharmacologic actions similar to the treatment drug. For example, the discontinuation of chronic morphine treatment produces signs of withdrawal that are reversed by other mu opioid agonists and not by kappa opioid agonists (Gmerek and Woods, 1985; Gmerek et al., 1987).

A second approach for studying the behavioral effects of withdrawal uses simple schedules of reinforcement. For example, in agonist-treated subjects, the termination of drug treatment or the administration of an antagonist can reliably disrupt conditioned behavior (e.g., responding maintained by a schedule of food presentation). Such disruptions in behavior have been observed after the termination of chronic treatment with various different drugs including phencyclidine (Slifer et al., 1984), tetrahydrocannabinol (Beardsley et al., 1986), cocaine (Woolverton and Kleven, 1988) and opioids (Holtzman and Villarreal, 1973; Thompson and Schuster, 1964). Moreover, reinitiation of responding occurs after several days or, more immediately, after the administration of the treatment drug or a pharmacological equivalent; these findings support the view that the disruptive effects of drug deprivation are related to withdrawal.

A third approach for studying withdrawal involves drug discrimination procedures in which agonist-treated subjects are trained to discriminate between an antagonist and vehicle. Because antagonist-lever responding increases in a dose-dependent manner after the administration of an antagonist, or in a time-dependent manner after the termination of agonist treatment, it is thought that responding on the antagonist-associated lever represents a withdrawal condition, whereas responding on the vehicle-associated lever represents a dependent condition (for reviews, see Emmett-Oglesby et al., 1990; France 1994b). In general, drugs with pharmacological actions similar to the treatment drug reverse antagonist-associated responding, whereas drugs with dissimilar pharmacologic actions do not reverse antagonist-associated responding (France and Woods, 1989; Holtzman, 1985).

This study examined the rate-altering effects, discriminative stimulus effects and other behavioral effects that occur on administration of an opioid antagonist in LAAM-treated monkeys and those effects that occur on discontinuation of twice daily injections of LAAM. Because of the long duration of LAAM, our study used multiple dependent variables to assess whether precipitated withdrawal was different from deprivation-induced withdrawal. These studies on opioid withdrawal in LAAM-treated subjects could be especially relevant to the long term use of LAAM as a maintenance therapy in humans. To date, no published study has systematically assessed the relationship between the discriminative stimulus and other behavioral manifestations of withdrawal in LAAM-dependent primates.

	Methods

Top Abstract Introduction Methods Results Discussion References

Subjects. Two male and two female adult rhesus monkeys (Macaca mulatta) were individually housed in stainless-steel cages with free access to water. Monkeys received fresh fruit and peanuts several times each week and were maintained under a 14-hr light/10-hr dark schedule. Access to food (Teklad monkey food) in the home cage was unrestricted during stimulus-shock termination procedures and restricted during food-maintained responding procedures; in the latter case, monkeys were maintained at no less than 90% of their free-feeding weights. All monkeys were experimentally naive at the initiation of these studies.

Apparatus. Monkeys were seated in either aluminum or Lexan chairs within ventilated, sound-attenuating chambers. Each chamber contained three response levers; the center lever could be extended (available) into the operant chamber or retracted out of the reach of the monkey (unavailable). Located above each lever were red and green stimulus lights. An externally mounted pellet dispenser could deliver 300 mg banana-flavored pellets (product F0179 Bio-serve, Frenchtown NJ) to a food cup located below the center lever. Each chair was equipped with a pair of shoes containing brass electrodes to which brief (250 msec) electric shock (3 mA) could be delivered from a remote a.c. shock generator. The feet of the monkeys were placed in shoes during all phases of these studies. Control of experiments and data recording were accomplished with a microprocessor as well as a commercially available interface and software.

Behavioral studies. Initially, three monkeys were trained to respond under a FR-10 schedule of food presentation and the antinociceptive, ventilatory and rate-decreasing effects of acutely administered opioids and nonopioids were assessed. After the completion of these acute drug studies (data from these studies will be presented in a subsequent report), sensitivity to the rate-decreasing effects of naltrexone was examined when the same three monkeys received LAAM daily.

Behavioral signs. Video tapes were evaluated by an experienced observer who was blind to experimental conditions. Behavioral signs [e.g., grimacing, vocalization, retching, vomiting, holding abdomen, lying on side, wet dog shakes, yawning, coughing and masturbation (male monkeys only)] that are commonly observed during mu opioid withdrawal (e.g., Katz, 1986) were scored for the 25-min session. Each period was divided into twenty-five 1-min bins and behavioral signs were recorded as either present or absent during each bin [i.e., a maximum frequency of 25 occurred when the sign was observed during every min (bin) of an observation period].

Data analyses. The rate-decreasing effects of naltrexone under the simple FR schedule in monkeys receiving LAAM daily are presented as a percentage of the average rate of responding during the 10 cycles immediately preceding LAAM treatment. Results of drug discrimination studies are presented as the average percentage of responses on the naltrexone lever (i.e., percent drug responding; %DR) ±1 S.E.M. plotted as a function of either dose or time. The %DR was not included for an individual monkey either when the response rate was less than 20% of the control rate (i.e., average response rate for that monkey for the five previous saline training cycles in which the testing criteria had been attained) or when more than two shocks were delivered. Response rates are presented as a mean percentage ±1 S.E.M. of control response rates. Naltrexone dose-effect curves were determined approximately monthly throughout these studies; although there was very little variation among naltrexone dose-effect curves (see "Results"), in figures the dose-effect curves for other drugs are compared to the naltrexone dose-effect curve that was determined at a similar time. In LAAM-treated monkeys, test compounds were considered to have substituted for the naltrexone discriminative stimulus when they produced at least 90% naltrexone-appropriate responding. Substitution studies typically ended in a particular subject when that subject responded at least 90% on the naltrexone lever or when more than two shocks were delivered; thus, the largest dose studied was not necessarily the same among all subjects. Nalbuphine and morphine were studied up to a maximum dose of 56.0 and 32.0 mg/kg, respectively. ED₅₀ values were calculated for individual subjects by linear regression when three or more data points were available and by interpolation when two data points (one above and one below 50% DR) were available; these ED₅₀ values were averaged among subjects. Differences between dose-effect curves (potency) were identified with two-way repeated measures analyses of variance using ED₅₀ values (P < .05). For behavioral signs, the frequency with which a particular sign was observed was averaged among subjects (±1 S.E.M.) and was plotted as a function of either dose or time. Data that are presented for unconditioned behavioral signs of withdrawal were obtained in all monkeys with the exception of masturbation which represents data for only the two male monkeys.

Drugs. The compounds studied were: LAAM, morphine sulfate, naloxone hydrochloride and naltrexone hydrochloride (the Research Technology Branch, National Institute on Drug Abuse, Rockville, MD), ketamine hydrochloride (Fort Dodge Laboratories, Inc., Fort Dodge, IA), nalbuphine hydrochloride (Mallinckrodt Inc., St. Louis, MO) and quadazocine methanesulfonate (Sterling-Winthrop Research Institute, Rensselaer, NY). With the exceptions of LAAM and ketamine, all compounds were dissolved in sterile water and injected s.c. in a volume of 0.1 to 4.0 ml. LAAM was dissolved in 85% H₂O, 10% emulphor and 5% EtOH; a small quantity of 5 molar NaOH was added to increase the pH to 6. Ketamine was diluted, as needed with saline, from a commercially available solution. Doses are expressed in mg/kg of body weight in terms of the salt forms.

	Results

Top Abstract Introduction Methods Results Discussion References

In the dose-finding study, the average rate of food-maintained responding for three monkeys before daily treatment was 1.22 ± 0.22 responses/sec. Rates of lever pressing were not altered during a week of treatment with 0.32 mg/kg/day of LAAM or during a subsequent week of treatment with 1.0 mg/kg/day (fig. 1, circles and squares, respectively, above days 1 and 6). After an increase in the daily dose of LAAM to 1.78 mg/kg, rates of responding were slightly decreased (triangles above days 1 and 6), primarily because one monkey responded at 75% or less of the control rate throughout treatment with 1.78 mg/kg/day of LAAM.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Rates of responding engendered under a FR schedule of food presentation in three monkeys that were treated with increasing doses of LAAM. LAAM was administered daily 2 hr before experimental sessions with each dose administered for 7 consecutive days. On the 7th day of treatment for each dose of LAAM, saline was administered during with first cycle followed by cumulative doses of naltrexone during subsequent cycles. Ordinate: average rate of responding calculated as a percentage (±1 S.E.M.) of the average rate of responding before LAAM treatment. Abscissa: left, effects of saline on days 1, 6 and 7 (Sal) each dose of LAAM treatment; right, dose of naltrexone in mg/kg body weight administered on day 7 of each week of treatment.

After 7 days of treatment with 0.32 mg/kg of LAAM, doses of naltrexone up to 0.1 mg/kg did not modify rates of responding (fig. 1). In contrast, a dose of 0.1 mg/kg of naltrexone decreased the average rate of responding to 51% of the control rate after 1 wk of treatment with 1.0 mg/kg of LAAM. Still smaller doses of naltrexone decreased rates of responding when the daily dose of LAAM was increased to 1.78 mg/kg (fig. 1, triangles).

Discrimination training commenced after the terminal dosing conditions were established (1.0 mg/kg/12 hr of LAAM) with monkeys satisfying the criteria for adequate stimulus control after an average of 46 training sessions (range = 20-76). The mean rate of responding (control) in four monkeys receiving LAAM was 2.3 ± 0.1 responses/sec. Naltrexone dose-dependently increased responding on the drug lever with a dose of 0.01 mg/kg producing more than 90% DR (fig. 2, circles, top); doses of naltrexone that occasioned drug-lever responding also decreased rates of responding (circles, bottom).

View larger version (27K): [in this window] [in a new window]   Fig. 2.   Discriminative stimulus effects of opioid antagonists in four LAAM-treated monkeys discriminating between naltrexone and saline and responding under a FR schedule of stimulus shock termination. Drugs were studied in an individual subject up to a dose that occasioned at least 90% drug-lever responding and numbers in parentheses represent the number of monkeys contributing to the respective data points (i.e., doses of quadazocine larger than 0.1 mg/kg were studied only in two monkeys). Ordinate: top, average percentage of responses on the naltrexone lever (%DR) ± 1 S.E.M.; bottom, average rate of responding calculated as a percentage (±1 S.E.M.) of the control rate of responding. Abscissa: dose in mg/kg body weight; saline (Sal) was administered on the first test cycle.

Other opioid antagonists substituted completely for the naltrexone discriminative stimulus in LAAM-treated monkeys. Naloxone had a similar potency to naltrexone, both as a discriminative stimulus and in decreasing rates of responding (fig. 2, squares, top and bottom), and quadazocine was 10-fold less potent than naltrexone and naloxone. A dose of 0.1 mg/kg of quadazocine produced at least 90% drug-lever responding in two monkeys, whereas doses of 0.32 or 1.0 mg/kg of quadazocine were required to produce at least 90% drug-lever responding in the other monkeys.

Morphine, nalbuphine and ketamine failed to substitute for naltrexone (data not shown), with each compound producing <10% DR at all doses. Moreover, up to doses of 32.0 and 56.0 mg/kg, respectively, morphine and nalbuphine did not modify rates of responding. In contrast, a dose of 3.2 mg/kg of ketamine eliminated responding in all monkeys.

Morphine, nalbuphine and ketamine also were studied in combination with naltrexone to determine whether they modified the discriminative stimulus effects of naltrexone. Morphine shifted the naltrexone discrimination dose-effect curve to the right (fig. 3, top) as evidenced by significant (P < .05 for 3.2 and 10.0 mg/kg) increases in the naltrexone ED₅₀: a 2-fold increase with a dose of 3.2 mg/kg of morphine and a 4-fold increase with a dose of 10.0 mg/kg. When administered alone, doses of 3.2 and 10.0 mg/kg of morphine increased rates of responding to more than 135% of the control rate (data above Sal); these doses of morphine also shifted the dose-effect curve for rate-decreasing effects of naltrexone to the right (fig. 3, bottom panel).

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Discriminative stimulus effects of naltrexone alone and in combination with 3.2 or 10.0 mg/kg of morphine (n = 4). Morphine was administered 1 hr before experimental sessions and saline (Sal) was administered on the first cycle. See figure 2 for additional details.

At similar doses, nalbuphine was more effective than morphine in modifying the discriminative stimulus effects of naltrexone. Doses of 3.2 and 10.0 mg/kg of nalbuphine also shifted the naltrexone dose-effect curve to the right with the naltrexone ED₅₀ increasing (P < .05) 5- and 14-fold, respectively, in the presence of nalbuphine (fig. 4, top). When administered alone, neither dose of nalbuphine modified rates of responding; however, each dose appeared to attenuate the rate-decreasing effects of naltrexone (bottom).

View larger version (26K): [in this window] [in a new window]   Fig. 4.   Discriminative stimulus effects of naltrexone alone and in combination with 3.2 or 10.0 mg/kg of nalbuphine (n = 4). Nalbuphine was administered at the beginning of the first cycle (i.e., 20-min pretreatment; data above Nal). See figure 2 for additional details.

In contrast to morphine and nalbuphine, each of which significantly increased the naltrexone ED₅₀ for discriminative stimulus effects, a dose of 0.32 mg/kg of ketamine failed to significantly alter the potency of naltrexone as a discriminative stimulus (P > .05; fig. 5, top). This dose of ketamine also failed to modify the rate-decreasing effects of naltrexone (bottom).

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Discriminative stimulus effects of naltrexone alone and in combination with 0.32 mg/kg of ketamine (n = 4). Ketamine was administered at the beginning of the first cycle (i.e., 20-min pretreatment; data above Ket). See figure 2 for additional details.

LAAM treatment was discontinued in order to compare the discriminative stimulus effects of naltrexone to the discriminative stimulus effects of LAAM deprivation. Monkeys varied markedly in their response to LAAM deprivation and results from individual subjects are presented in fig. 6. Ten hr after the last injection of 1.0 mg/kg of LAAM, all four monkeys responded exclusively on the saline lever (fig. 6, leftmost open symbols, all panels). For monkey O (upper left), responding was predominantly on the naltrexone lever for 5 days after discontinuation of LAAM treatment. This monkey responded predominantly on the saline lever 6 and 7 days after discontinuation of LAAM treatment. Monkey C responded on the saline lever 2 and 3 days after discontinuation of LAAM treatment; at times less than 2 days and longer than 3 days, this monkey responded predominantly on the naltrexone lever (upper right). Monkeys K and T each responded predominantly on the naltrexone lever either 1 or 1.5 days after discontinuation of LAAM treatment and, thereafter, on the saline lever (lower panels).

View larger version (30K): [in this window] [in a new window]   Fig. 6.   Discriminative stimulus effects of LAAM deprivation in four monkeys (different panel for each subject). Twice daily injections of saline were substituted for LAAM. In addition, acute injections of naltrexone (0.1 mg/kg) were tested 10 hr, 4 and 8 days after discontinuation of LAAM treatment. Abscissa: days after the termination of LAAM treatment. See figure 2 and "Methods" for additional details.

When administered 10 hr after an injection of 1.0 mg/kg of LAAM, a dose of 0.1 mg/kg of naltrexone occasioned more than 95% naltrexone-lever responding in all four monkeys (fig. 6, leftmost closed symbols, all panels) and decreased average rates of responding to 69% of control rates (data not shown). When administered 4 or 8 days after the discontinuation of LAAM treatment, the same dose of naltrexone occasioned predominantly naltrexone-lever responding in two monkeys (O and C) and saline-lever responding in the remaining two monkeys (K and T). In contrast to the marked rate-decreasing effects that were observed with this dose of naltrexone 10 hr after LAAM, response rates were not altered by naltrexone 4 or 8 days after discontinuation of LAAM treatment (data not shown).

On a second occasion, morphine was substituted for LAAM to determine whether another mu agonist would attenuate the naltrexone lever responding that occurred when LAAM treatment had been discontinued. All four monkeys responding exclusively on the saline lever when 3.2 mg/kg/6 hr of morphine was substituted for 1.0 mg/kg/12 hr of LAAM (data not shown). Rates of lever pressing were not changed when morphine was substituted for LAAM (not shown).

A third study further examined reversal by morphine of deprivation-induced naltrexone-lever responding. Figure 7 shows the individual data obtained with cumulative doses of morphine 10 days after discontinuation of twice daily LAAM treatment. A cumulative dose of 1.0 mg/kg fully reversed naltrexone-lever responding in monkey O and a dose of 32.0 mg/kg fully reversed naltrexone-lever responding in monkey C. Subjects K and T responded exclusively on the saline lever prior administration of any morphine (points above Sal, fig. 7) and continued to do so throughout the morphine dose-effect determination. Morphine decreased rates of responding in a dose-related fashion in 10-day LAAM-deprived monkeys (lower panel) with a cumulative dose of 10.0 mg/kg decreasing the average rate to 56% of control rates. In contrast, up to a dose of 32.0 mg/kg, morphine had no rate-decreasing effect when monkeys were receiving LAAM twice daily (data not shown).

View larger version (26K): [in this window] [in a new window]   Fig. 7.   Discriminative stimulus and rate-decreasing effects of saline (Sal) and cumulative doses of morphine 10 days after the discontinuation of LAAM treatment. See figure 2 for additional details.

The frequency with which other behavioral signs were observed when saline was administered 10 hr after LAAM are shown by the shaded circles in figure 8; with the exception of "holds abdomen," behavioral signs that often occur during opioid withdrawal were not present under this condition. In contrast, acute administration of 0.1 mg/kg of naltrexone 10 hr after LAAM (fig. 8, shaded triangles) elicited a variety of behaviors including grimacing, wet-dog shakes and vomiting.

View larger version (34K): [in this window] [in a new window]   Fig. 8.   Behavioral signs observed during 25-min periods at various times after the discontinuation of LAAM treatment. In addition, acute injections of 0.1 mg/kg of naltrexone were assessed 10 hr, 4 and 8 days after discontinuation of treatment. Ordinates: the average (±1 S.E.M.) number of signs. See figure 7 and "Methods" for additional details.

Similarly, withdrawal-associated behavioral signs were also observed when twice daily LAAM injections were discontinued (fig. 8, open circles) with a peak in signs occurring 1 to 3 days after the last injection of LAAM. For example, lying on side, grimacing and masturbation occurred during the first 3 days of LAAM deprivation and not thereafter. Vocalizations, holding abdomen and wet-dog shakes also occurred with greater frequency during the 3 days immediately after the discontinuation of LAAM treatment. In contrast, retching, vomiting and coughing were not observed during 9 days of LAAM deprivation. With the exception of an apparent increase in vocalization 4 days after discontinuation of LAAM, acute administration of 0.1 mg/kg of naltrexone (open triangles) did not further increase the frequency of withdrawal-associated behavioral signs (compare days 4 and 8 for naltrexone to 6 and 8 for saline). Finally, some behavioral signs (e.g., retching, vomiting and coughing) were observed when naltrexone was administered 10 hr but not 4 or 8 days after the discontinuation of LAAM treatment.

	Discussion

Top Abstract Introduction Methods Results Discussion References

The magnitude of dependence and withdrawal that develop after repeated drug treatment can be quantified by changes in behavior after either the discontinuation of drug treatment or by the administration of a pharmacological antagonist. In monkeys receiving morphine chronically, behavioral signs such as wet-dog shakes, agitation and yawning occur after the discontinuation of morphine treatment or the administration of an opioid antagonist (Gmerek et al., 1987; Katz, 1986). Similarly, in this study, the discontinuation of LAAM treatment or the administration of naltrexone resulted in the expression of some behavioral signs of mu opioid withdrawal, although in the former case the magnitude and frequency of these signs were less than typically observed after discontinuation of treatment with morphine. For example, coughing, retching and vomiting that are often observed after discontinuation of morphine (e.g., 3.2 mg/kg/6 hr) treatment (e.g., Gmerek et al., 1987) were not observed when LAAM treatment was discontinued. Nevertheless, several qualitatively similar effects emerge after discontinuation of treatment with either LAAM or morphine, thereby implicating a common mechanism (mu receptor) for dependence on each of these opioids.

It is likely that the development of dependence on LAAM was important for establishing naltrexone as a discriminative stimulus in this study. A discrimination has been established with an opioid antagonist in monkeys treated subchronically with a mu agonist (France, 1994a), although most often opioid antagonists are studied in subjects receiving agonist chronically (e.g., France and Woods, 1989; Holtzman, 1985). In one study, the dose of naloxone that was established as a discriminative stimulus in morphine-treated rats could not be established as a discriminative stimulus in untreated rats (Weissman, 1978). When a large dose of naltrexone was used as the training stimulus in untreated subjects, the discriminative stimulus effects of naltrexone appeared to be unrelated to opioid actions (Valentino et al., 1983). In contrast, the discriminative stimulus effects of naltrexone clearly were mediated by mu opioid receptors in monkeys receiving LAAM; the relative potencies of naltrexone, naloxone and quadazocine for producing drug-lever responding were similar to their potencies for producing (other) signs of withdrawal (Gmerek et al., 1987), as well as for producing naltrexone-like discriminative stimulus effects in monkeys receiving morphine chronically (France and Woods, 1989). Acute injections of the mu-selective agonists nalbuphine and morphine (France and Woods, 1989; Gerak and France, 1996) shifted the naltrexone dose-effect curve to the right, further supporting the view that the discriminative stimulus effects of naltrexone in LAAM-treated monkeys involve mu opioid receptors. The larger shifts to the right in the naltrexone dose-effect curve that were obtained with nalbuphine, as compared to the same doses of morphine, might be due to the higher affinity that nalbuphine has for mu binding sites, as compared to morphine (Emmerson et al., 1996). Collectively, these results indicate that the discriminative stimulus effects of naltrexone are mediated by mu opioid receptors in LAAM-treated monkeys.

Two-choice discrimination procedures in drug-treated subjects, like the one used in our study, cannot easily identify the qualitative nature of the training stimuli. For example, the stimulus condition that is associated with an antagonist could represent withdrawal or could simply indicate the absence of agonist. The notion that monkeys in this study were discriminating between withdrawal and nonwithdrawal stimuli, and not simply between the presence and absence of agonist, is supported by the observance of certain behavioral effects upon discontinuation of LAAM treatment. Upon discontinuation of LAAM treatment, naltrexone-lever responding was maximal when other behavioral signs of withdrawal were also maximal (i.e., 1-2 days after discontinuation of treatment). Thereafter, responding was distributed between the saline and the naltrexone levers, among the various monkeys. It is unlikely that variations in response patterns among subjects (2-10 days after the discontinuation of LAAM treatment) simply represent a loss of stimulus control because morphine completely reversed naltrexone-lever responding 10 days after the discontinuation of LAAM treatment. It is also unlikely that the lack of sustained naltrexone-lever responding in some monkeys, after discontinuation of LAAM treatment, was due to the long duration of LAAM or one of its metabolites, because naltrexone did not increase naltrexone-lever responding 4 or 8 days after the discontinuation of LAAM. Indeed, variations in response patterns were consistent within, but not between, individuals, perhaps reflecting individual differences in sensitivity to and expression of opioid withdrawal.

Compared to the effects obtained when LAAM treatment was discontinuing, a greater number and magnitude of withdrawal-associated behavioral effects were observed after the administration of naltrexone. In rats, a less severe withdrawal was observed after the termination of LAAM treatment as compared to the withdrawal signs that were evident after the administration of naltrexone (Young et al., 1979a). Similarly, in monkeys, the frequency of withdrawal signs, decreases in response rates and the proportion of responses made on the naltrexone lever were less after the discontinuation of LAAM treatment as compared to the administration of naltrexone. These results suggest that differences in the magnitude of withdrawal between deprivation-induced and precipitated withdrawal might be related to the duration of action of LAAM and its metabolites. This notion is further supported by other studies showing that the magnitude of withdrawal is less after the termination of LAAM treatment as compared to the magnitude of withdrawal observed after the termination of treatment with shorter acting opioids (e.g., morphine and methadone; Young et al., 1977, 1979a,b).

Early studies in humans reported numerous toxic effects with LAAM (e.g., respiratory depression; Fraser and Isbell, 1952; Keats and Beecher, 1952) which might have resulted from the use of small inter-injection intervals and the accumulation of LAAM and its metabolites. More recent studies have varied dosing conditions and demonstrated LAAM to be a safe and effective treatment for opioid dependence (Blaine et al., 1981; Tennant et al., 1986); however, the rate of metabolite formation and the peak levels of metabolites in plasma after either a single dose or multiple doses of LAAM can be highly variable among individuals (Billings et al., 1974; Henderson et al., 1977; Kaiko and Inturrisi, 1975). Similarly, previous studies in non-human primates have reported marked individual differences in the behavioral effects of LAAM, even among subjects receiving the same dose of LAAM over long periods of time. Chronic dosing with 2.0 mg/kg of LAAM, administered orally three times per week for periods of up to 1 yr, was reported to produce mild to severe signs of opioid intoxication and, occasionally, lethal effects, but only in some monkeys (Downs et al., 1977; Misra and Mule, 1977, 1978). In another study, variations in the response to LAAM among subjects necessitated an adjustment in dose to avoid toxic effects (Crowley et al., 1985). In our study, 1.0 mg/kg of LAAM, administered s.c. every 12 hr, was not associated with any behavioral signs of toxicity over a 2-yr study period. Notably, the total dose of LAAM administered in the current study (14.0 mg/kg per week, s.c.) was more than twice the total dose of LAAM administered by others (p.o. 6.0 mg/kg per week; Downs et al., 1977; Misra and Mule, 1977, 1978). Differences between our study and other studies with regard to toxicity, might be related to the dosing intervals; in our study, LAAM was administered every 12 hr, whereas LAAM was administered every 24 or 48 hr in other studies. A shorter dosing interval with smaller doses might more closely approximate steady-state conditions or, at least, a condition whereby plasma concentrations of LAAM and its metabolites change relatively little across days. Because the formation of active metabolites appears to be important for the behavioral effects of LAAM, the effectiveness of LAAM in treating opioid dependence could vary markedly among individuals, thereby complicating empirical assessment of its relative utility for this indication.

The rate at which LAAM is metabolized to nor-LAAM and dinor-LAAM might also account for differences among studies. When drugs are administered orally they are more susceptible to rapid hepatic metabolism (i.e., first-pass effect) than when they are administered by other routes (e.g., s.c.). Because hepatic metabolism is primarily responsible for the conversion of LAAM to its active metabolites, nor-LAAM and dinor-LAAM (e.g., Billings et al., 1974), differences among studies could be related to variations in the rate of metabolism of LAAM by different routes. Early studies in humans reported that orally administered LAAM had a faster onset on action, with effects observed within 1.5 hr, whereas s.c. administered LAAM had effects only after 14 hr (Fraser and Isbell, 1952). The homogeneity in response to twice daily injections of LAAM among the four monkeys in our study, as well as the general lack of toxic effects obtained with dependence-producing doses of LAAM, suggest that the variability in effects reported by others might be related to dosing interval and route of administration. It is clear that oral dosing with LAAM should be carefully tailored to individuals to maximize therapeutic effects without generating toxicity.