Preclinical Evaluation of Pharmacotherapies for Treatment of Cocaine and Opioid Abuse Using Drug Self-Administration Procedures Nancy K

Home , Acetylmethadol, DPDPE, List of opioids, Naloxonazine, Naltrindole

ELSEVIER

REVIEW

Preclinical Evaluation of Pharmacotherapies for Treatment of Cocaine and Opioid Abuse Using Drug Self-Administration Procedures Nancy K. Mello, Ph.D., and S. Stevens Negus, Ph.D.

Drug abuse is a major public health problem, and the used clinically for other indications on drug self relationship between intravenous drug abuse and AIDS administration are critically examined. Where possible, the underscores the need for more effective treatment degree of concordance between clinical and preclinical medications. Animal models of drug self-administration are studies of drug abuse treatment medications is discussed. useful to systematically evaluate new treatment We conclude that drug self-administration models are medications and predict clinical efficacy. This review valuable for preclinical assessment of medication efficacy, summarizes the status of preclinical evaluations of and we recommend some strategies to further improve medications for treatment of cocaine and opiate abuse. The evaluation procedures. The discovery of more effective basic drug self-administration methodology and the medications for substance abuse treatment should be rationale for experimental designs and outcome criteria are facilitated by recent advances in behavioral science, described. Studies of the effects of dopamine or opioid pharmacology, neurobiology and medicinal chemistry. receptor agonists and antagonists as well as medications [Neuropsychophannacology 14:375--424, 1996]

KEY WORDS: Cocaine treatment; Heroin treatment; Opioids; (Adams and Durell 1984; Kozel and Adams 1986) was Substance abuse; Drug self-administration; Substance paralleled by an increased use of cocaine by opioid-de abuse treatment pendent and methadone-maintained patients (Kosten et Opioid abuse and cocaine abuse are major public health al. 1986, 1987; Cushman 1988; Condelli et al. 1991). problems (NIDA 1995), and the development of more Methadone, the traditional treatment for opioid abuse effective medications for drug abuse treatment is a con (Dole 1988), has not been as effective in reducing con tinuing challenge. Recently, a number of factors have current cocaine abuse (Kosten et al. 1986, 1987, 1989a). converged to make drug abuse treatment a highly visi Some patients begin or increase cocaine abuse during ble national priority. The cocaine epidemic of the 1980s methadone treatment (Cushman 1988; Chaisson et al. 1989; Condelli et al. 1991; Dunteman et al. 1992). There is no uniformly effective pharmacological treatment for This paper is based on a review presented at the 1993 Plenary Ses cocaine abuse per se, and concurrent abuse of both co sion of the American College of Neuropsychopharmacology. The caine and opioids presents an even greater challenge for Plenary Session was entitled "The Development of New Drugs for the development of treatment medications (Rawson et the Treatment of Alcohol and Cocaine Dependence: The Scientific Challenges." Preparation of this review was supported in part by al. 1991; Halikas et al. 1993; Tutton and Crayton 1993; grants no. DA00lOl, DA04059, and DA02519 from the National Mendelson and Mello 1996). Institute on Drug Abuse, NIH. The urgent need to develop new pharmacotherapies From the Alcohol and Drug Abuse Research Center, McLean Hos pital, Belmont, MA. is prompted in part by the link between intravenous use Address correspondence to Dr. Nancy K. Mello, Alcohol and of cocaine and heroin and the spread of AIDS. Intrave Drug Abuse Research Center, McLean Hospital, 115 Mill Street, Bel nous drug abuse and needle sharing have been esti mont, MA 02178. Received Julv 23, 1995; revised November 28, 1995; accepted mated to account for over 30 percent of all AIDS cases December 4, 1995. and over 50 percent of AIDS cases in women (National

NECROPSYCHOl'HARMACOLOCY IY%--VOI .. 14, NO. 6 © 1996 American College of Neuropsychopharmacology Published by Elsevier Science Inc. 0893-133X/96/$15.00 655 Avenue oi the Americas. New York, NY I 001(1 SSDI 0893-133X(95)00274-X 376 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6

Commission 1991; CDC 1993). Intravenous cocaine and drug self-administration models offer some method heroin abuse also may accentuate the risk for AIDS, ological advantages for the systematic evaluation of because each has direct immunosuppressive effects new medications. Unlike clinical trials, compliance with (Donahoe and Falek 1988; Pillai et al. 1991). One ap the treatment regimen is ensured, and there is no con proach to AIDS prevention is more effective drug abuse founding influence of unreported polydrug abuse. The treatment. Pharmacotherapies can provide treatment effects of a new treatment medication on an ongoing for larger numbers of patients than would be possible pattern of drug self-administration can be evaluated with nonpharmacological approaches and can also quantitatively under controlled experimental condi enhance more traditional treatment methods such as tions. Accurate baseline measures of the daily dose and counseling and rehabilitation. These arguments were patterns of drug self-administration can be determined summarized in a congressional staff report entitled before, during, and after administration of the treat "Pharmacotherapy: A Strategy for the 90s" developed ment medication, whereas in clinical studies the drug by a committee chaired by Senator J. R. Eiden (1989). abuse history and the usual pattern of drug use are of This report focused attention on the importance of de ten unknown. It is also possible to monitor medication veloping improved pharmacotherapies for drug abuse safety during exposure to the abused drugs and to treatment and led to the establishment of the Medica detect any adverse side effects that compromise the tions Development Division at the National Institute on health of the animal. Finally, uncontrolled social factors Drug Abuse, National Institutes of Health. such as expectancy or placebo effects and peer pressure With the impetus of congressional sanction and the cannot complicate the interpretation of data obtained in specter of an accelerating epidemic of AIDS associated animal models. Moreover, targeted preclinical trials are with intravenous drug abuse, scientific interest in phar more cost-effective than extensive clinical trials. macotherapies for drug abuse treatment has increased. This review describes the use of animal models of Investigators from a broad spectrum of disciplines are drug self-administration to evaluate potential medica now concerned with the development and evaluation tions for treatment of cocaine, opioid, and polydrug of new drug abuse treatment medications, and this abuse involving concurrent cocaine and opioid abuse. gathering momentum was reflected by the Plenary We have selected these three forms of substance abuse Session of the American College of Neuropsychophar because of their prevalence and medical and social viru macology, of which this review was a part. Of course, lence (Kreek 1987; DAWN 1993; Mendelson and Mello recognition of the value of pharmacotherapies for drug 1994; NIDA 1995). We have focused on the drug self-ad abuse treatment is not new. The basic arguments were ministration model because it has been studied exten clearly articulated by Dole and Nyswander (1965) in the sively across a number of laboratories in several species mid-1960s and federally endorsed in the 1973 National and appears to have clinical relevance. Where possible, Strategy for Drug Abuse Prevention published by the we have compared preclinical data with clinical evalua White House Special Action Office for Drug Abuse tions of established pharmacotherapies for drug abuse Prevention. Unfortunately, medical approaches to drug treatment, but many medications examined in preclini abuse treatment were often met by a series of quasi cal studies have not been approved for clinical testing. moralistic objections as well as arguments for enhanced The validation of preclinical models for evaluation of law enforcement and interdiction of drug supplies. Ad drug abuse treatment medications requires an ongoing vocacy for drug supply reduction strategies has per assessment of the concordance between preclinical sisted undiminished since the early 1900s and continues studies and outpatient studies. These interrelated pro to command the greatest allocation of federal resources cesses of validation and prediction are illustrated sche devoted to reducing drug abuse (Musto 1973). This po matically in Figure 1. litical emphasis has often obscured the real benefits of Drug self-administration models are especially valu methadone treatment for opioid dependence while ex able for evaluation of medications that substitute for or aggerating its limitations. antagonize the effects of cocaine and heroin. These medi The emerging consensus that more effective pharma cation strategies do not depend on any particular con cotherapies may reduce drug abuse and risk for AIDS ceptualization of the pathogenesis of drug abuse, and has stimulated interest in using preclinical models to the effects of antagonists and agonists on drug self-ad evaluate potential medications for drug abuse treat ministration can be interpreted in terms of pharmaco ment (Mello 1991, 1992; Balster et al. 1992; Spealman logical interactions with the target drug of abuse. How 1992; Woolverton and Kleven 1992). Animals will reli ever, in addition to pharmacological antagonists and ably self-administer most drugs that are abused by agonists, a wide range of medications approved for man, and it is reasonable to anticipate that medications other indications are often used clinically in substance that decrease drug self-administration should be more abuse treatment. For example, antidepressants such as clinically effective than medications that do not change desipramine and fluoxetine have been used clinically or increase drug self-administration. Moreover, animal for the treatment of abuse of cocaine alone and com- NEUROPSYCHOPHARMACOLOGY 1996-VOL. 1-l, NO. 6 Evaluation of Treatments for Drug Abuse 377

Figure 1. Cross validation of preclinical and clinical models. Validation of the effec tiveness of animal models for preclinical evaluation of drug abuse treatment medica tions requires assessing the degree of con cordance between preclinical studies and outpatient clinical studies. Eventually, the Pre-Clinical preclinical model should enable users to Clinical predict the potential effectiveness of new Pharmacotherapy Treatment pharmacotherapies. This interactive pro Evaluation Efficacy cess of cross-validation and prediction is es sential for refinement of the preclinical model. From Mello NK (1992): Behavioral Strategies for the Evaluation of New Phar macotherapies for Drug Abuse Treatment. NIDA Research Monograph 119. Washing ton, DC: Government Printing Office, pp 150-154.

bined cocaine and opioid dependence (Gawin and Kle cocaine self-administration are summarized: dopamine ber 1984; Tutton and Crayton 1993; Mendelson and antagonists (Table 1); dopamine agonists and partial ag Mello 1996). Insofar as a coexisting psychiatric disorder onists (Table 2); opioid antagonists, agonists, and mixed may contribute to the initiation and perpetuation of agonist-antagonists (Table 3); and medications ap drug abuse, this strategy has potential clinical useful proved for other indications (Table 4). Two categories of ness. Antidepressant medications also have been exam medication effects on opioid self-administration are ined in preclinical models of cocaine self-administra summarized in tabular form: opioid antagonists (Table tion, and these studies will be reviewed. A variety of 5), and opioid agonists and mixed agonist-antagonists other medications ranging from anticonvulsants to (Table 6). The effects of a diverse group of nonopioid ibogaine also have been examined in animal models of drugs on opioid self-administration are described in the cocaine and opioid self-administration, and we discuss text. The implications of findings from this emerging the implications of these studies for medication devel database for the design of optimal preclinical studies opment. Finally, a number of the studies described in for pharmacotherapy evaluation are discussed in the fi this review were designed to analyze the pharmacologi nal section of this review. cal mechanisms underlying the reinforcing effects of abused drugs rather than to evaluate the effects of po tential treatment medications. Despite a different em PRECLINICAL MODELS OF INTRAVENOUS phasis, results from these studies of drug interactions DRUG SELF-ADMINISTRATION often are very relevant to issues of treatment medica Drug Self-Administration: Basic Procedures tion development. This review is divided into four sections. The first During the 1960s, several independent groups reported section describes the basic methods used to conduct the first studies showing that monkeys and rats surgi drug self-administration studies and the types of data cally implanted with an intravenous catheter would obtained. ln this context, we examine the underlying as self-administer a wide range of opioids, stimulants, and sumptions and the advantages and limitations of vari alcohol (Weeks 1962; Thompson and Schuster 1964; ous study designs for evaluating treatment medications. Yanagita et al. 1965; Deneau et al. 1969). The procedures The rationale for several criteria used to determine currently used for drug self-administration studies are treatment medication efficacy are critically examined. similar across laboratories. A chronic intravenous cathe The next two sections review illustrative preclinical ter is surgically implanted into an accessible vein (usu studies of the effects of drugs on the self-administration ally a jugular or femoral vein) to permit infusion of a of cocaine, opioids, and cocaine/ opioid combinations. drug solution. During drug self-administration ses We have attempted to provide a comprehensive, al sions, the subject acquires drug injections by emitting though not exhaustive, summary of the basic preclinical some discrete response, such as pressing a response key literature in tabular form. These tables summarize stud or a lever on an operant panel. The number and pattern ies that examined drug effects on cocaine self-adminis of responses required for each injection are specified by tration (Tables 1-4) and opioid self-administration (Ta the schedule of reinforcement, and the availability of bles 5 and 6). Four categories of medication effects on drug injections under this schedule of reinforcement is 378 N.K. Mello and S.S. Negus NEUR0PSYCH0PHARMAC0L0GY 1996-V0L. 14, NO. 6

usually signaled by a stimulus, such as the illumination the abuse liability of novel compounds (Schuster and of a stimulus light on or above the response key. Param Johanson 1974; Thompson and Unna 1977; Brady et al. eters such as the schedule of reinforcement, the volume 1987), and there has been consistently good concor and duration of each injection, the duration and fre dance between clinical and preclinical assessments of quency of drug self-administration sessions, and the the reinforcing effects of drugs (Griffiths and Balster stimulus conditions in effect during drug self-adminis 1979; Brady and Lukas 1984; Balster 1991). Similarly, in tration sessions are controlled by computers. In addi sofar as the goal of pharmacotherapy is to reduce drug tion, alternative reinforcers such as food may be made self-administration, this drug self-administration model available. The dependent variables in drug self-admin also should be useful for the evaluation of new treat istration studies are usually the number of injections ment medications (Mello 1992). delivered or the rate of responding during each drug Drug self-administration studies have consistently self-administration session. obtained an inverted U-shaped dose-effect curve relat It is now well-established that animals will self-ad ing the dose of drug delivered in each injection (the unit minister most drugs that are abused by man, and there dose) to some metric of drug self-administration behav are many similarities in patterns of animal and human ior, such as response rate or the number of injections de drug self-administration (Mello 1979; Griffiths et al. livered. There is an extensive literature indicating that 1980; Brady and Lukas 1984). For example, drug self cocaine and opioid self-administration dose-effect curves administration procedures have been used to predict display the typical inverted U-shaped dose-effect curves

Figure 2. Number of injections/day or food pellets/day delivered to rhesus monkeys when different unit doses of cocaine or heroin were available during drug self-administration ses 80 100 sions. Abscissae: Unit dose cocaine in mg/kg/inj • (log scale) available during drug self-administra tion sessions. Left Ordinates: Number injections 60 delivered per day (open squares). Right Ordinates: 75 Number 1 gm food pellets delivered per day (filled circles). Points above "Sal" show data from 40 days when saline was available during drug ses 50 sions. Monkeys responded for food and drugs under a second-order FR 4 (VR16:S) schedule. /' Four food sessions were run daily from 6:00 to 20 25 7:00 A.M., 11:00 A.M. to noon, 3:00 to 4:00 P.M., D and 7:00 to 8:00 P.M. During each food session, D monkeys could earn up to 25 pellets, for a maxi 0 -,-- >, 0 mum of 100 pellets per day. Four drug sessions ca Sal 0.001 0.0032 0.01 0.032 0.1 were run daily from 7:00 to 8:00 A.M., noon to Unit Dose Cocaine (mg/kg/inj) 1:00 P.M., 4:00 to 5:00 P.M., and 8:00 to 9:00 P.M. C 0 During each drug session, monkeys could earn ;: up to 20 injections, for a maximum of 80 injec CJ G) tions per day. The monkeys were initially trained "i: 80 100 to self-administer a maintenance dose of 0.01 or • 0.032 mg/kg/injection of cocaine and saline, and each unit dose of cocaine or heroin was substi 60 75 tuted for the maintenance dose for 5 to 10 con secutive days and until responding for injections stabilized. Stability was defined as 3 consecutive 40 50 days during which the number of injections/ day was within 10% of the mean number of injections per day. Each point shows mean data from the last 3 days of each condition in 11 monkeys (top 20 25 panel) or three monkeys (bottom panel). Brackets D show the SE. Adapted from Negus SS; Mello NK; Lukas SE; Mendelson JH (1995a): Diurnal patterns -,-- 0 0 of cocaine and heroin self-administered in rhesus Sal 0.0001 0.00032 0.001 0.0032 0.01 monkeys responding under a schedule of multi Unit Dose Heroin (mg/kg/inj) ple daily sessions. Behav Pharmacol. 6:763-775. NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, '.\JO. 6 Evaluation of Treatments for Drug Abuse 379

under most conditions (Brady and Lukas 1984; Woods as sedation and stereotypies), which limit rates of oper et al. 1987). Similar dose-effect curves also were mea ant responding. Thus, along the ascending limb, rein sured for simultaneous self-administration of cocaine forcing effects predominate and response rates increase, and heroin combinations in a primate model of poly whereas along the descending limb, rate-decreasing ef drug abuse (Mello et al. 1995). Controlled clinical stud fects predominate and response rates decrease. In con ies have usually found that increasing drug doses are trast, food-maintained responding remains quite stable associated with increasing levels of drug self-adminis during the ascending limb of the cocaine and the heroin tration corresponding to the ascending limb of the dose-effect curve and decreases during the descending dose-effect curve in preclinical studies (see Griffiths et limb of the dose-effect curve. al. 1980 for review; Lamb et al. 1991). Detailed explora The slopes of the ascending and descending limbs, tion of the effects of high doses has often been limited the magnitude and width of the peak, and the position by considerations of subject safety in clinical research. of the curve along the abscissa can vary as a function of In practice, the drug abuser seldom knows the purity or several parameters, including the schedule of reinforce the dose of the drug purchased, and the empirical dose ment, the drug available for self-administration, and effect curve may span a range from a mild subjective the behavioral history of the subject. An example of high to a potentially lethal overdose. how the cocaine dose-effect curve in rats can vary as a An example of a dose-effect curve for cocaine and function of schedule parameters such as the fixed ratio heroin self-administration by rhesus monkeys is shown requirement and the postinjection timeouts is shown in in Figure 2. The cocaine dose-effect curve is character Figure 3. Notice that lengthening the postinjection time ized by an ascending limb (0.001-0.01 mg/kg/inj.}, a out from 20 seconds to 2 minutes changed the position peak (0.01-0.032 mg/kg/inj.) and a descending limb of the ascending limb, peak, and descending limb of the [0.032-0.1 mg/kg/inj. (see Figure 2, top panel)]. Along cocaine dose-effect curve. Thus, when postreinforce the ascending limb, increasing unit doses of cocaine re ment timeouts were brief, rats could rapidly self-ad sult in an increase in the number of injections delivered minister many cocaine injections and a low unit dose until some peak number of injections per day is ob was at the peak of the dose-effect curve. When succes tained. Along the descending limb, further increases in sive cocaine injections were separated by at least 2 min the unit doses of cocaine result in decreases in the num utes, higher unit doses were required to maintain re ber of drug injections. A similar dose-effect curve is sponding. These data illustrate that the schedule of shown for heroin self-administration (Figure 2, lower reinforcement is an important determinant of baseline panel). The peak of the heroin dose-effect curve is at a drug dose-effect curves. unit dose of 0.0032 mg/kg/inj., and the number of in Evaluation of Medication Efficacy jections decreases at higher unit doses. The shape of these dose-effect curves reflects an interaction between Study Designs. The basic study design used for pre the self-administered drug's reinforcing effects, which clinical evaluation of a potential pharmacotherapy is tend to increase response rates, and other effects (such similar to a single-blind, crossover trial in clinical re-

300 Figure 3. The effect of schedule require ments and postinjection timeout duration on the cocaine self-administration dose-effect w~ 250 curve. Saline and different unit doses of zC_o ...J ·- cocaine are shown on the abscissa. The per w cent of baseline control responding (total (/) Q) 200 <( cocaine injections per 3-hour session) is co LL Ql shown on the left ordinate. Cocaine self 0 2 150 administration maintained on a FR 5 to 20-sec f- .2 z .!: timeout (closed circles) and a FR 15 to 20-sec w~ 0- timeout (open circles) represents the average of a: .!!! 100 eight and four rats, respectively. Cocaine self W 0 a..~ administration maintained on a FR 15 to 2-minute timeout (open squares) was deter 50 mined in six or eight rats. Adapted and reprinted with permission from Caine SB and Koob GF (1994): Effects of dopamine D and 0 1 D antagonists on cocaine self-administration 0 0.063 0.125 0.250 0.500 2 under different schedules of reinforcement in COCAINE DOSE (mg/kg/inj) the rat. J Pharmacol Exp Ther 270(1):209-218. 380 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6

Possible Effects of Pharmacotherapy(•+A) on a Drug Dose-Response Curve (0)

Antagonism• Depression+ Enhancement A 80 , C , ._ 60 , 8. , 11) p I 5 40 I u I -~ 20 .!:

0 s 0.03 0.10 0.30 Drug Dose Per Injection (mg/kg) Figure 4. Medication effects on drug dose-response curve. Drug injections per day are shown on the ordinate and the drug dose per injection is shown on the abscissa of each panel. Injections per day during saline-maintained responding are shown above Son each abscissa. Open circles, drug dose-effect curve during saline control treatment; Solid squares, left panel, antago nism of the drug's reinforcing effects and rightward shift in the drug dose-response curve. Solid diamonds, center panel, depression of drug-maintained responding with no shift in the dose-response curve.Solid triangles, right panel, enhancement of the drug's reinforcing effects and leftward shift in the drug dose-response curve. search. First, baseline drug self-administration patterns but there is an overall reduction in responding (Figure are examined during placebo treatment with saline or a 4, center panel). vehicle control solution. Once drug self-administration Enhancement. Alternatively, the cocaine dose-re patterns are stable, the treatment medication is substi sponse curve may be shifted to the left during treat tuted for saline placebo treatment on a single occasion ment. If lower doses of cocaine now maintain the same or for several days. Placebo-control treatment is then re level of responding as during saline control treatment, sumed, and the rate at which drug self-administration these data suggest that the treatment medication is en returns to baseline control levels is measured. This se hancing cocaine's reinforcing effects (Figure 4, right quence may be repeated several times using different panel). doses of the treatment drug or different treatment It is important to note that both leftward and right drugs. It is important to study several doses of the treat ward shifts in the dose-effect curve decrease self-ad ment medication to determine the range of doses over ministration of some unit doses, but simultaneously in which it is maximally effective. crease self-administration of other unit doses. Leftward Medication Effects on Drug Dose-Effect Curves. Usu or rightward shifts, then, represent a change in the po ally, the effects of the treatment medications are first tency of the self-administered drug. Only a downward evaluated on a dose of the abused drug that maintains shift in the dose-effect curve involves a decrease in the high levels of drug self-administration, that is, a reinforc self-administration of all unit doses. From a treatment ing dose on the ascending limb or the peak of the drug perspective, medications that shift the cocaine or heroin dose-effect curve. If a treatment drug reduces self-ad dose-effect curve downward may have the most gen ministration of a reinforcing dose of the abused drug, eral therapeutic utility, because drug-taking behavior this is one indication of potential treatment efficacy. decreases across a broad range of unit doses. Medica However, a treatment drug may have different effects at tions that simply alter the potency of the self-adminis higher and lower doses of the abused drug, so a com tered drug and shift the dose-effect curve to the right plete dose-effect curve should also be examined. Three may be less satisfactory. It is apparent in Figure 4 (left possible effects of a treatment medication on a theoreti panel) that despite a rightward shift in the dose-effect cal cocaine dose-response curve are shown in Figure 4. curve, the drug available for self-administration retains its ability to maintain high rates of drug-taking behav Antagonism. If the dose-response curve for cocaine is ior and drug users may compensate for the change in shifted to the right during treatment, this suggests that potency by changing the amount of drug used per occa the medication is antagonizing the reinforcing effects of sion or the rate of drug self-administration. This phe cocaine. Higher doses of cocaine are required to main nomenon is illustrated in many of the studies we review, tain the same level of responding previously main in which treatment medication effects were evaluated tained by a lower dose of cocaine (Figure 4, left panel). against a single unit dose of the self-administered drug Depression. A second possibility is that the position located on the descending limb of the dose-effect curve. of the cocaine dose-effect curve may remain the same, In these studies, an increase in drug self-administration NEUROPSYCHOPHAR\1ACOLOCY 1996-VOL. 14, NO. 6 Evaluation of Treatments for Drug Abuse 381

often is taken as evidence that the underlying dose-ef passed results in reinforcement. One fundamental dif fect curve has shifted to the right and that the treatment ference between fixed-ratio and fixed-interval sched medication has antagonized the reinforcing effects of ules is the degree of control that the animal has over the the self-administered drug. Although an increase in frequency of drug injections. For example, in a fixed-ra drug self-administration may be interpreted as evi tio (FR) schedule of reinforcement, the interval between dence for antagonism, it is not a desirable goal for med successive drug injections is controlled by the animal's ications that are to be used in the clinical treatment of rate of response, and response rate and number of injec drug abuse. tions are likely to be highly correlated. In contrast, on a fixed-interval (Fl) schedule, the frequency of injections Duration of Treatment. Treatment medications can be is primarily determined by the reinforcement schedule, given in single acute doses or chronically over days or not by response rates. The duration of the postreinforce weeks. Pragmatic considerations, such as very limited ment timeout also influences the frequency of drug in quantities of a new medication, may dictate the use of jections (see Figure 3). an acute rather than a chronic treatment paradigm. Most investigators have relied exclusively on only However, most medications are used clinically on a one dependent variable and have not studied the corre chronic basis, and some pharmacotherapies require a lation between rate of operant responding and number long exposure before treatment effects are evident of drug injections. The issue is complicated by the fact (Gawin and Ellinwood 1988; Gawin 1991). In preclinical that drugs usually have a direct effect on rates of re studies, repeated daily exposures to the medication sponding that is cumulative during the drug self-ad may be necessary to detect a suppressant effect on drug ministration session. Moreover, drugs from different self-administration. Chronic treatment offers several im pharmacological classes may have different effects on portant methodological advantages over acute treatment response rates. For example, stimulants such as cocaine for preclinical evaluations. During chronic treatment, it may increase rates of responding initially, whereas opi is possible to measure the persistence of treatment med oids may decrease rates of responding. Although both ication effects and to determine whether an initial de measures have limitations, the number of injections has crease in drug self-administration remains constant a consistent definition from laboratory to laboratory over time. lf a decrease in drug self-administration is whereas the definition of response rates may vary. only transient, this could indicate that tolerance devel There is disagreement as to whether overall response oped to the treatment drug effect. For example, chronic rate (including postreinforcement pauses), running rate administration of a treatment medication for as little as (time from the first response to acquisition of each rein 8 to 10 days sometimes is sufficient to reveal tolerance forcer), or derivative measures (such as inter-response development in rhesus monkeys (Woolverton and Virus times distributions) provide the most sensitive index of 1989; Kleven and Woolverton 1990b; Gold and Balster operant response rates (Skjoldager et al. 1991). Al 1992; Negus et al. 1995b). In contrast, testing single though response rate data during medication treatment acute doses may lead to an erroneous conclusion of no are often presented as percent of control rates, these effect or may yield false positives because an initial de definitional and procedural differences complicate com crease in drug self-administration may not be sustained. parisons between laboratories. As research on preclini Treatment Outcome Measures. The dependent mea cal evaluation of treatment medications evolves, it is sures used to assess treatment medication efficacy usu likely that more attention will be directed toward inte ally are the number of drug injections taken during a grating information from these two dependent variables. period of drug availability and the rate of operant re The number of drug injections is a sensitive measure sponding for drug injections. As we have seen in Fig of medication efficacy (Mello et al. 1989, 1990b, 1992) ures 2 and 3, both behavioral measures are critically in and shows good concordance with clinical trials (Mello fluenced by the schedule of reinforcement, the drug et al. 1990b, 1993c; Mello 1991). However, most labora dose per injection, and the cumulative dose acquired tories limit the total number of drug injections available over a session. The importance of the schedule of rein per session to minimize the possibility of drug over forcement in determining drug self-administration has dose. This safety precaution could impose an artificial been extensively reviewed elsewhere, and discussion of ceiling effect that would prevent detection of a medica the many complex schedule permutations and the ef tion-induced enhancement of drug self-administration fects on operant behavior is beyond the scope of this re (Mello et al. 1990b). Because the effectiveness of chronic view (see Ferster and Skinner 1957; Dews 1958, 1981; treatment is defined in terms of a selective decrease in Kelleher and Morse 1968; Byrd 1981 for review). The ba drug self-administration, limiting the number of drug sic building blocks of complex schedules are ratio injections is seldom a serious problem. Unlimited ac schedules where the animal is reinforced after emitting cess to some drugs has led to severe toxicity in rhesus some number of responses, and interval schedules monkeys (Aigner and Balster 1978) so there is no practi where the first response after a specified interval has cal alternative to restricted drug access. 382 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6

Figure 5. Butorphanol effects on cocaine and food self-administration. 80 100 Nonselective effects of treatment with an opioid mixed agonist-antagonist analgesic, butorphanol, on cocaine and food self-administration by rhesus mon- 80 keys. The effects of daily treatment with 60 0 0 saline or butorphanol (0.01 to 0.30 mg/ i:i.. kg/ + day) on cocaine (solid circles) and t !E. food self-administration (solid squares) .... ;:;- c,:s 60 .. are shown. Saline treatment and each 40 dose of butorphanol were studied for

-~= '< 10 days. Each data point is the average ....u (± SE) of five subjects except for butor- Qj 40 phanol (0.3 mg/kg), which is an aver- ·=-.... age of four subjects. The statistical Qj f ·;= 20 significance of each change from the u 20 saline treatment baseline is indicated by 0 u * asterisks (**, p < 0.01). From Mello NK; Kamien JB; Lukas SE; Drieze J; Mendel- 0 son JH (1993d): The effects of nalbu- 0 phine and butorphanol treatment on Saline 0.01 0.03 0.1 0.3 cocaine and food self-administration by Butorphanol (mg/kg/day) rhesus monkeys. Neuropsychopharma- cology 8:45-55. Reprinted with permission.

Selectivity of Treatment Effects on Drug Self-Administration ure 5. At the lowest dose of butorphanol (0.01 mg/kg/ day), cocaine self.-administration decreased significantly, During preclinical evaluation of medication effects on but food self-administration remained slightly above drug-taking behavior, the principal focus is on the abil saline treatment levels. However, as treatment drug ity of the medication to change drug self-administra doses were increased, there was a parallel decline in tion. However, preclinical studies also provide an both food and cocaine self-administration. This means opportunity to examine a medication's adverse conse that the effects of the 0.03- to 0.3-mg/kg dose of butor quences that may limit its clinical usefulness. In drug phanol were not selective for cocaine, but rather re self-administration studies, the adverse effects of a flected a general disruption of operant behavior. medication usually are inferred from its effects on be Another strategy used to evaluate selectivity is to haviors other than self-administration of the target drug compare the effects of the treatment drug on self-ad of abuse. For example, the effects of a medication on un ministration of the target drug of abuse and self-admin conditioned behaviors such as posture or locomotor ac istration of a drug from another pharmacological class. tivity as well as adverse effects such as convulsions or If self-administration of the target drug is reduced and extreme sedation can be observed and recorded. self-administration of the comparison drug is not, then The extent to which a treatment medication selec it can be argued that the treatment drug is having a selec tively reduces drug self-administration is another im tive effect. An example of this approach is the evaluation portant consideration. If the treatment medication re of potential treatment medications on cocaine and alfenta duces drug self-administration to a greater extent or for nil self-administration (Winger et al. 1992; Winger 1994). a longer time than it reduces behavior maintained by Comparisons between two or more drugs from differ another reinforcer, this would suggest that the treat ent pharmacological classes also may clarify the mecha ment drug has selective effects on drug self-administra nisms by which the treatment medication reduces drug tion. Alternatively, if the treatment drug is sedating, this self-administration (Winger et al. 1992). One problem in could result in a parallel dose-dependent decrease in interpreting data from studies that compare medication the self-administration of both a drug and a nondrug effects on the target drug with an alternative drug or reinforcer such as food. An example of both a selective with nondrug reinforcers is whether or not each has and a nonselective decrease in cocaine self-administra equivalent reinforcing properties. However, determin tion during daily treatment with an opioid mixed ago ing the equivalence of reinforcers is a complex issue of nist-antagonist analgesic, butorphanol, is shown in Fig- continuing debate that is beyond the scope of this review. NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6 Evaluation of Treatments for Drug Abuse 383

Although medication-induced suppression of respond the clinical and the preclinical treatment evaluation lit ing maintained by other reinforcers is undesirable, the erature is based in part on the assumption that an effec interpretation of medication effects on the self-adminis tive treatment medication should make the abused tration of an alternative reinforcer may be complicated drug more like saline or like a lower dose of the abused by the same considerations that influence interpretation drug by antagonizing its reinforcing properties. Be of medication effects on self-administration of the target cause substituting saline for cocaine or heroin may re drug (i.e., the inverted U-shaped curve relating magni sult in an initial increase in drug-maintained respond tude of the reinforcer to rates of self-administration be ing, a treatment that increases drug self-administration havior; see Figures 2 and 3). Furthermore, evaluating may be classified as a potentially effective antagonist. medication effects on the self-administration of alterna This criterion for medication effectiveness usually is ap tive reinforcers addresses only one dimension of the plied in studies where one or more doses of the treat many adverse effects that could be produced by a med ment drug were administered on a single occasion ication. For example, a medication that produced a non (acute treatment). However, an increase in drug self-ad selective suppression of operant behavior at doses that ministration is not a desirable outcome in studies in produced no other toxic effects may have clinical utility which the medication is administered repeatedly over under some circumstances. Thus, an evaluation of the days or weeks (chronic treatment). In chronic treatment selectivity of a medication in decreasing drug self-ad studies, medication effectiveness is reflected in a sus ministration is an important step in a more general pro tained decrease in drug self-administration. The saline cess of characterizing that medication's adverse effects. analogy remains relevant insofar as self-administration One final medication effect that can be directly as of saline also decreases over several days of exposure. sessed in drug self-administration studies is the rein Some investigators have measured the number of days forcing effect of the medication itself. To evaluate a required for saline-maintained responding to decrease medication's reinforcing effects, it is substituted for co to a low level and used this as the basis for determining caine or heroin to determine the dose-range over which how long to evaluate each dose of a treatment medica it maintains operant behavior leading to its self-admin tion (Woolverton and Virus 1989; Kleven and Woolver istration. Because reinforcing effects in preclinical stud ton 1993). ies are indicative of abuse liability in humans, the abil If an initial increase in drug self-administration in ity of a drug to reinforce behavior is usually considered acute treatment studies was usually followed by a de to be an adverse effect. However, in the clinical treat crease in drug self-administration over time in chronic ment of drug dependence, medications are self-admin treatment studies, then reconciliation of findings from istered by the patient and medication delivery depends reports of acute and chronic treatment would be less on the patient's active participation and cooperation. As problematic. However, the inconsistencies in results ob a result, medication-induced reinforcing effects may im tained with the same treatment medication in the same prove patient compliance with the treatment regimen. species indicate the importance of these procedural dif In addition, the relative costs associated with medi ferences. This issue is further complicated by different cation-induced reinforcing effects must be weighed results obtained when cocaine or heroin doses are on against the costs of continued abuse of the target drug. the ascending or descending limb of the dose-effect Even if a medication produces reinforcing effects on its curve (see Figure 4). own, it may still be valuable in the treatment of drug dependence if it decreases self-administration of the tar get drug and produces fewer adverse effects than the TREATMENT MEDICATION EFFECTS ON target drug. COCAINE SELF-ADMINISTRATION Clinical Treatment of Cocaine Abuse Interpretation of Criteria for Medication Efficacy Currently there is no uniformly effective clinical treat From a clinical perspective, a decrease in drug self-ad ment for cocaine abuse and dependence. Treatment is ministration is the goal of treatment, and this is the cri complicated by the diversity of cocaine administration terion of efficacy used in inpatient evaluations of new patterns as well as concurrent abuse of opioids, alcohol, medications (Meyer and Mirin 1979; Mello and Mendel and other substances. Often pharmacotherapies that ap son 1980; Mello et al. 1981; 1982). However, in preclinical peared to be effective with cocaine abusers were not ef studies, this criterion has not been used consistently as fective with more severely ill persons who met DSM-III-R an index of potential medication effectiveness. Rather, criteria for cocaine dependence (see Mendelson and many investigators have interpreted an increase in drug Mello 1996 for review). Conversely, severely cocaine self-administration as evidence that the treatment med dependent men who also were heroin dependent ap ication was antagonizing the reinforcing effects of co peared more responsive to treatment with an opioid caine or heroin. This apparent contradiction in goals in mixed agonist-antagonist, buprenorphine, than opioid- 384 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6

dependent patients who used cocaine occasionally (see tigators, and the findings are often inconsistent (see Ta Mello et al. 1993c for review). Although a wide range of ble 1). These apparent inconsistencies reflect differences medications, including antidepressants, dopamine ago in procedure and in interpretation of findings as well as nists, antagonists and reuptake inhibitors, anticonvul variations in the receptor specificity of the dopamine sants and opioid agonists, antagonists, and mixed ago antagonists studied. As appreciation of the complexity nist-antagonists have been studied in clinical samples, of the dopamine system increased and compounds se the findings were often inconclusive (see Rawson et al. lective for dopamine receptor subtypes became avail 1991; Halikas et al. 1993; Tutton and Crayton 1993; Men able, it became possible to study the effects of specific delson and Mello 1996 for review). Inconsistent diag D1 and D2 dopamine receptor antagonists on cocaine nostic criteria for cocaine abuse and cocaine depen self-administration. dence, as well as controversy over appropriate criteria Most of the early studies examined the effects of for treatment effectiveness, contribute to the discrepant compounds with nonspecific dopaminergic activity results in the clinical literature. Traditionally, total drug such as chlorpromazine and haloperidol (Wilson and abstinence has been the accepted standard for success Schuster 1972; Herling and Woods 1980; de la Garza of treatment. Yet, this criterion is more stringent than is and Johanson 1982; Roberts et al. 1989). Often these applied in other areas of medicine. The clinical reality is mixed dopamine antagonists increased rather than de that most cocaine abusers and other substance abusers creased stimulant self-administration. These findings continue to use some drugs during maintenance treat were interpreted as evidence for antagonism of co ment, but a 30°/4, to 60% reduction in drug use may ben caine's reinforcing properties because reducing the unit efit the patient, as well as society (see Mello and Men dose of a reinforcing drug or substitution of saline also delson 1995; Mendelson and Mello 1996 for review). results in an initial increase in rates of drug self-admin istration. Many of the early studies administered single acute doses of nonselective dopamine antagonists and Cocaine Self-Administration Dose-Effect Curves did not evaluate treatment effects on alternative rein Cocaine maintains behavior leading to its self-adminis forcers such as food, so it was not possible to distin tration under a wide range of experimental conditions guish between rate altering effects of the treatment (Kelleher 1976; Spealman and Goldberg 1978; Griffiths drug and the enhancement or antagonism of the rein et al. 1979a; Spealman 1985; Winger 1988; Woods et al. forcing effects of cocaine (Table 1). Examples of some 1987). Under some second-order schedule conditions, selected studies are described to illustrate how these re monkeys responded for up to 1 hour for a single co search methods have evolved. caine injection (Goldberg et al. 1976). As shown in Fig One of the first studies of the effects of a selective D1 ure 2, the dose-effect curve relating cocaine dose per in antagonist (SCH 23390) and a selective D2 antagonist jection to either the number of drug injections per (pimozide) on the self-administration of cocaine and session or response rate has a typical inverted-U shape piribedil (a direct D2 agonist) by rhesus monkeys was with intermediate doses maintaining more drug self-ad reported by Woolverton in 1986. Each dopamine antago ministration behavior than either lower or higher doses. nist was studied across a broad dose range, and the ef Cocaine usually is classified as an indirect dopamine fects of single acute antagonist doses on a single rein agonist, and there is considerable evidence that co forcing dose of cocaine or piribedil were evaluated. The caine's reinforcing effects are mediated in part through traditional efficacy criterion of changes in responding its blockade of dopamine reuptake at presynaptic termi similar to the effects of reducing the unit dose of co nals (see Dackis and Gold 1985; Ritz et al. 1987; Ritz et caine was used as the basis for interpretation of the al. 1988; Kuhar et al. 1988, 1991; Johanson and Fischman findings. Since the D1 antagonist resulted in dose-re 1989; Spealman et al. 1992; Wise et al. 1995; Woolverton lated decreases in cocaine and piribedil self-administra and Johnson 1992 for review). A number of other struc tion, this was interpreted as a nonspecific decrease in turally distinct compounds that block dopamine re rates of responding. The D2 antagonist increased rates uptake are also reinforcing in animals (Woolverton et al. of both cocaine and piribedil self-administration, and 1984; Winger and Woods 1985; Bergman et al. 1989). Co because this effect was similar to the initial effects of caine also blocks the reuptake of serotonin and nor substituting saline for cocaine, these data were inter epinephrine, but these appear to be less important for preted as evidence for blockade of each drug's reinforc its reinforcing effects. ing effects. High doses of both dopamine antagonists produced cataleptic effects (Woolverton 1986). Subsequently, Woolverton and Virus (1989) reevalu Effects of Dopamine Antagonists on Cocaine ated the effects of the same D and D dopamine antag Self-Administration [Table 1] 1 2 tmists on self-administration of cocaine and an alterna The effects of dopamine antagonists on cocaine self-ad tive reinforcer, food, to determine the extent to which ministration have been examined by a number of inves- the behavioral effects of these antagonists were specific NECROPSYCHOPHARMACOLOGY 1996-VOL. 14, '\JO. 6 Evaluation of Treatments for Drug Abuse 385

to cocaine (Woolverton and Virus 1989). Each antago SCH 23390, there was a persistent sensitization to the nist was administered for the same number of sessions reinforcing effects of cocaine that lasted for at least 2 (4-10) that were required for drug-maintained respond weeks. Low doses of cocaine that did not maintain self ing to decline after saline substitution, and data were administration before SCH 23390 treatment, became re presented for the last 3 days of each treatment condi inforcing, that is, the ascending limb of the cocaine tion. Each antagonist was studied over a behaviorally dose-response curve was shifted to the left (Kleven and active but nontoxic dose range, and its effect on self-ad Woolverton 1990b). ministration of a single reinforcing dose of cocaine was With the development of longer-acting and more se examined. Each session consisted of a sequence of food lective dopamine antagonists, there has been additional availability for 15 minutes, a 15 minute timeout, cocaine support for the hypothesis that the reinforcing proper availability for 15 minutes, a 15 minute timeout, and ties of cocaine may be selectively reduced by antago food availability for 15 minutes. Cocaine and food were nism of specific dopamine receptors (Bergman et al. available under a FR30 schedule of reinforcement, and 1990; Corrigall and Coen 1991a; Hubner and Moreton there was a 2-minute timeout after delivery of each rein 1991; Bergman and Rosenzweig-Lipson 1992; Spealman forcer to minimize any direct rate-reducing effects of co et al. 1992; Caine and Koob 1994). Bergman and co caine. Under these conditions, there was an antagonist workers (1990) examined the effects of a long-acting D1 dose-related decrease in cocaine-maintained respond dopamine antagonist, SCH 39166, and of a D2 antago ing. However, these effects were not specific for cocaine, nist, eticlopride, on cocaine self-administration in squir because food maintained responding also decreased as rel monkeys (Bergman et al. 1990). This study differed the antagonist dose increased. Moreover, the initial re from many previous reports in that dopamine antago ductions in both cocaine and food self-administration nist effects were examined on an entire cocaine dose-ef were often followed by a return to baseline levels fect curve rather than on a single reinforcing dose of during chronic antagonist treatment suggesting that tol cocaine (see Table 1). Moreover, these D1 and D2 antag erance may have developed to these dopamine antago onists were examined chronically over 5 days. This pro nists (Woolverton and Virus 1989). These two pioneering cedure permitted examination of whether or not the co studies from the same laboratory have been described caine dose-effect curve was shifted to the right, in some detail because they illustrate the development suggesting antagonism of cocaine's reinforcing effects of more refined methods for examining the effects of by these dopamine antagonists (see Figure 4). Single potential treatment drugs on cocaine self-administra acute doses of the D1 antagonist SCH 39166 or the D2 tion. Although the investigators concluded that the de antagonist eticlopride were studied at each point in the creases in cocaine self-administration after D1 dopa cocaine dose-response curve. Responding for cocaine mine antagonist administration did not reflect blockade was maintained on a second-order FR 30, FI 10 minute of cocaine's reinforcing effects in both studies, the meth schedule of reinforcement and no alternative reinforcer odological improvements in the second study made this was studied. As shown in Figure 6, SCH 39166 shifted conclusion less open to alternative interpretations. the ascending limb of the cocaine self-administration The complexity of D1 antagonist effects on cocaine's dose-effect curve to the right in four of five monkeys, reinforcing properties was further illustrated in a study suggestive of surmountable antagonism of cocaine's re in which both selective attenuation of cocaine's rein inforcing effects (Bergman et al. 1990). Comparable ef forcing effects during SCH 23390 treatment and subse fects were obtained with eticlopride. quent enhancement following treatment were observed The authors concluded that these data were consis (Kleven and Woolverton 1990b). Because SCH 23390 tent with antagonism at the level of dopamine receptors has a relatively short duration of action, the effects of a (Bergman et al. 1990). However, there remains consider continuous infusion on cocaine and food-maintained able disagreement as to the capacity of dopamine recep responding were evaluated in rhesus monkeys under tor antagonists to antagonize competitively the rein conditions similar to those in the Woolverton and Virus forcing effects of cocaine (Roberts and Vickers 1984; study (1989). A range of doses of SCH 23390 was ad Caine and Koob 1994; Richardson et al. 1994; Winger ministered chronically over 5 to 13 days, the same pe 1994; Negus et al. 1996). In rats, low doses of the riod of time required for responding to decrease when dopamine D1 receptor antagonists SCH 23390 (5 mcg/ saline was substituted for cocaine. Baseline doses of co kg) and SCH 39166 (10 mcg/kg), but not A69024, selec caine were available for 2 weeks or more between suc tively decreased cocaine self-administration without cessive treatments. SCH 23390 infusion produced a se decreasing responding for food maintained on a multi lective decrease in cocaine-maintained responding in ple schedule (Caine and Koob 1994). Higher doses of two of the four monkeys. However, these effects dimin these D1 antagonists decreased both cocaine- and food ished over the period of observation, and both cocaine maintained responding. When only cocaine-maintained and food-maintained responding returned to pretreat responding was examined, treatment with these three ment baseline levels. Interestingly, after termination of D1 antagonists, as well as the D2 receptor antagonists Table 1. Summary of Preclinical Treatment Medication Evaluations in the Drug Self-Administration Model: Effects of <:.,J Q0 Dopamine Antagonists on Cocaine Self-Administration a, z Effect on Effect on Treatment Treatment Treatment Pretreatment Target Drug Target Other Other

Drug Dose Duration Interval Doses/Injection Drug Reinforcer Reinforcer Species Reference 0 D1 Dopamine receptor antagonists ;jCl 0. A69024 0.3125-2.5 Acute 15 or 30 min. Cocaine, 0.25 Increase and Food Decrease Rat Caine and \fl mg/kg SC dose mg decrease# Koob 1994 0 injections z depending on oql'O C schedule CJ) SCH 23390 0.003-0.30 Acute 30min. Cocaine, Decrease# Not tested Rhesus Woolverton mg/kg IV dose 0.03 or 0.1 injections monkey 1986 mg/kg SCH 23390 0.03-0.25 4-10 15min. Cocaine, 0.025 Decrease response Food Decrease Rhesus Woolverton mg/kg SC sessions or 0.05 mg/kg rate monkey and Virus 1989 SCH 23390 0.8-6.4 mg/ 5-13 days Continuous Cocaine, (l.025 Transient Food Transient Rhesus Kleven and kg/day IV infusion or 0.05 or 0.10 decrease decrease monkey Woolver- mg/kg response response ton 1990b rate rate SCH 23390 .1-30 µg/kg Acute 30min. Cocaine, 0.1-1.0 Low dose: Not tested - Rat Corrigall SC dose mg/kg increase# and Coen injections; high 1991b dose: decrease# of injections SCH 23390 2.0-20 µg/ Acute 30min. Cocaine, 0.75 Increase response Not tested Rat Hubner and kg SC dose mg/kg rate (FR5); Moreton z decrease 1991 tTl C"'. progressive :,:1 0 'Cl ratio breakpoint [.fJ ,< SCH 23390 0.05 mg/kg Acute 30 min. Cocaine, 0.125- Decrease Not tested Rat Lane et al. n IP dose 0.5 mg interreinforcer 1992 :r: 0 time 'Cl :r: SCH 23390 2.5-40 µg/ Acute 15 or 30 min. Cocaine, 0.25 Increase+ Food Decrease Rat Caine and :i,

kg SC dose mg decrease# Koob 1994 :i, injections n 0 depending on r- 0 schedule Cl ,< _. 0.01-0.1 5-21 days lOmin. Cocaine, dose Rightward Not tested Squirrel Bergman et \0 SCH 39166 and/ \0 mg/kg IM response or downward monkey al. 1990 'I' < 0.01-1.7 mg/ shifts in dose- 0 kg effect curve _.r response rate z (continued) 9 0-,

I I

'1 '1

\,.) \,.)

00 00

l'D l'D

> >

i= i=

..., ...,

Cl Cl "" ""

CJl CJl

"" ""

a" a"

P-l P-l

"" ""

>-l >-l

0 0

l'D l'D

[Jl [Jl

l'D l'D tTI tTI

g g

< <

g, g,

::l ::l 5· 5·

P-l P-l

2" 2"

-D -D

a--

"' "'

......

>,< >,< z z r r (:J (:J

r r

,.., ,.., >-

< <

0 0

a-- m m

;::: ;:::

,0 ,0 c') c') Q Q

:r; :r; 0 0

::c ::c

CF, CF,

.-< .-< 0 0 "" "" n n

n n

z z

,0 ,0

C C

(Jq (Jq

, ,

et et

and and

1994 1994

and and

1977 1977

Virus Virus

and and

1990 1990

(continued) (continued)

1989 1989

and and

1986 1986

1984 1984 Vickers Vickers

Wise Wise

Koob Koob

al. al.

Koob Koob

Roberts Roberts

Woolverton Woolverton

DeWitt DeWitt

Bergman Bergman

Winger Winger

Caine Caine

Winger Winger

Caine Caine

monkey monkey

Rhesus Rhesus

Rat Rat

Rhesus Rhesus

Rat Rat

Rhesus Rhesus

Rat Rat

Squirrel Squirrel

in in

rates rates

response response

curves; curves;

effect effect

dose- rates rates

shift shift

response response

curves; curves;

effect effect

dose- rates rates

shift shift

Decrease Decrease

Downward Downward

Decrease Decrease

2) 2)

tested tested

(group (group

Not Not

Food Food

Not Not

Food Food

Alfentanil Alfentanil

Not Not

Food Food

Alfentanil Alfentanil

on on

high high

rates rates

or or

and/ and/

or or

dose-

response response

decrease decrease

+ +

in in

curves; curves;

in in

curves; curves;

in in

curves; curves;

increase increase

doses doses

downward downward

injections injections

rates rates

# #

injections injections

doses: doses:

injections; injections;

increase# increase#

response response

depending depending schedule schedule

injections injections

decrease# decrease#

response response

effect effect

or or shifts shifts injections injections

downward downward shifts shifts response response

shifts shifts

depending depending schedule schedule

effect effect

decrease# decrease#

downward downward

effect effect

Increase# Increase#

Decrease Decrease

Low Low

Initial Initial

Rightward Rightward

Increase Increase

Rightward Rightward

Increase+ Increase+

mg/ mg/

0.05 0.05

mg/ mg/

or or

mg/kg mg/kg

mg mg

mg/kg mg/kg

0.025 0.025

kg kg

1 1

0.03-0.1 0.03-0.1

0.75 0.75

mg/kg mg/kg

0.25mg 0.25mg

kg kg

response, response,

0.01-1.7 0.01-1.7

dose dose mg/kg mg/kg

0.001-0.3 0.001-0.3

0.00032-0.10 0.00032-0.10

0.25 0.25

Cocaine, Cocaine,

min. min.

30 30

or or

min. min.

or or

hrs hrs

3 3

2 2

4 4

15 15

10 10 60min. 60min.

2 2

days days

dose dose

sessions sessions dose dose

dose dose

Acute Acute

4-15 4-15

Acute Acute

Acute Acute 5-21 5-21

antagoni;,~ antagoni;,~

IV IV

mg/kg mg/kg IM IM

mg/kg mg/kg

mg/kg mg/kg IP IP mg/kg mg/kg

µg/kg µg/kg TP TP

IM IM

µg/kg µg/kg

mg/kg mg/kg IV IV

SC SC IV IV

mg/kg mg/kg

SC SC

0.006-0.10 0.006-0.10

0.003-0.30 0.003-0.30

0.05-0.1 0.05-0.1

0.0625-0.50 0.0625-0.50

5-40 5-40

0.001-0.1 0.001-0.1

0.003-0.10 0.003-0.10

0.001-0.006 0.001-0.006

5-40 5-40

receptor receptor

39166 39166

Pimozide Pimozide

Pimol'.ide Pimol'.ide

Eticlopride Eticlopride

SCH SCH

SCH SCH D2Jlopamine D2Jlopamine (,IJ 00 00

Table 1. (continued) z

Effect on Effect on :?:

Treatment Treatment Treatment Pretreatment Target Drug Target Other Other 0 Drug Dose Duration Interval Doses/Injection Drug Reinforcer Reinforcer Species Reference "'::l p_. Raclopride 50-400 Acute 15 or 30 min. Cocaine, No change# Food Selective Rat Caine and (fl µ,g/kg dose 0.25 mg injections or decrease Koob 1994 0 decrease z SC ro depending on (Jq C schedule V, Spiperone 2.0-40.0 Acute 30min. Cocaine, Increase response Not tested Rat Hubner and µ,g/kg dose 0.75 mg/kg rate (FR5); Moreton SC decrease 1991 progressive ratio breakpoint Spiperone 3-30 Acute 60min. Cocaine, Increase# Not tested Rat Corrigall µ,g/kg dose 0.1-1.0 mg/kg injections and Coen SC 19916 Spiperone 5-40 Acute 15 or 30 min. Cocaine, Increase+ Food Decrease Rat Caine and µ,g/kg dose 0.25 mg decrease# Koob 1994 SC injections depending on schedule Sulpiride 5-20 Acute 3hrs Cocaine, Increase# Not tested Rat Roberts and mg/kg dose 0.75 mg/kg injections Vickers IP 1984 Other recq;,tor antagonists ztTl (+)-AJ 76 0.1-3.0 Acute 15min. Cocaine, Increase# Food (after Decrease Rhesus Vanover et c::: el mg/kg dose 0.03 or 0.10 injections session) monkey al. 1993 [fl IM mg/kg ,-< n (+)-AJ 76 1.88-30.0 Acute 30min. Cocaine, Decrease Not tested - Rat Richardson :r: 0 mg/kg dose 0.6mg progressive et al. 1993 :r: IP ratio breakpoint :i, (-)DS 121 7.5-15 Acute 30min. Cocaine, Decrease Food No effect Rat Smith eta!. :,:, 3:: dose 0.2 mg/kg/ progressive (group 2) 1995 :i, mg/kg n IP In). ratio breakpoint 0r' Alpha 0.01-0.40 Acute 2.5 hrs Cocaine, Low doses: Not tested Rat Ettenberg et 0 Cl flupenthixol mg/kg dose 0.75 mg/kg increase# al. 1982 ,-< ..... IP injections; high -D -D doses: decrease I °'< # injections 0 r-' (continued) ..... !'" z 9 °'

'° '°

00 00

'-" '-"

"' "'

ro ro C C

C C "' "'

CJ CJ

"' "'

0 0

"' "'

i i

ro ro

"' "'

Pl Pl

....., .....,

>+, >+,

0 0

::i ::i

0 0

0G 0G

;::t. ;::t.

Pl Pl

c c

< <

Pl Pl

tT1 tT1

a, a,

7'. 7'.

p p

,--' ,--' r r

0 0

'° '° -< -<

:z :z 1 1

C'J C'J

5 5

0 0 ,*'" ,*'" n n

$'. $'.

:;o :;o

::r: ::r:

0 0

::r: ::r: n n

-< -<

lfl lfl

0 0

:;o :;o

C C tn tn

z z

of of

in in

Col-

al. al.

Table Table

in in

1994 1994

and and

1995. 1995.

et et

cocaine cocaine

1994 1994

group group

respond

shown shown

al. al.

opioid opioid

conducted. conducted. on on

is is

and and

1995 1995

ton ton

Woolver

and and

of of

et et

1984 1984

Vickers Vickers

shown shown

Smith Smith

Vanover Vanover

was was

Richardson Richardson

Winger Winger

Roberts Roberts

1970 1970

another another

are are

opioid opioid

in in

effects effects

an an

study study

or or

antagonists antagonists

the the

or or

self-administration. self-administration.

the the

between between

drug-maintained drug-maintained

treatment treatment

of of

subjects subjects

monkey monkey

of of

monkey monkey

which which

cocaine cocaine

Rat Rat

Rhesus Rhesus

of of

Rat Rat

cocaine cocaine

rate rate

Rhesus Rhesus

dopamine dopamine

Rat Rat

in in

of of

published published

summarizes summarizes

of of on on

6 6

group group

duration duration

species species

Table Table

effects effects

papers papers

the the

injection injection

injections, injections,

opioids opioids

same same

the the

in in

of of

the the

of of

per per

and and

the the

in in

rates rates

response response

dose-effect dose-effect curves; curves;

Downward Downward

shift shift

dose dose

effects effects

number number

indicates indicates

9 9

The The

the the

tested tested

representative representative

4. 4.

summarizes summarizes

2) 2)

J J

either either

was was

self-administration. self-administration.

administration administration

in in

Column Column

tested tested

of of

tested tested

include include

Table Table

tested tested

Column Column

(group (group

in in

opioid opioid

summarizes summarizes

Not Not

change change

J\ftentanil J\ftentanil

blank. blank.

Not Not

route route

3 3

reinforcer reinforcer

tables tables

on on

is is

model. model.

of of

the the

8 8

shown shown

1i1ble 1i1ble

These These

is is

another another

direction direction

or or

Column Column

studied, studied,

not not

antagonists antagonists

paper. paper.

session session

lhe lhe or or

the the

drug drug

of of

tested, tested,

dose-effect dose-effect

opioid opioid

of of

breakpoint breakpoint ratio ratio

progressive progressive

self-administration self-administration

injections injections

ratio ratio

breakpoints breakpoints

decrease decrease

progressive progressive

injections injections

rates rates

response response

in in curves; curves;

downward downward shifts shifts

of of

injections injections

Decrease Decrease

was was

Decrease# Decrease# whether whether

terms terms

Increase# Increase#

end end

Rightward Rightward

Increase# Increase#

self-administration. self-administration.

in in

drug drug

treatment treatment

the the

effects effects

at at

the the

cocaine cocaine

in in

indicates indicates

reinforcer reinforcer

of of

self-administration self-administration

7 7

on on

0.1 0.1

drug drug

other other

doses doses or or

references references

mg/kg/ mg/kg/

mg/kg mg/kg

no no

Col1111111 Col1111111

mg mg

the the

of of

The The

agonists agonists

If If

inj. inj.

evaluations evaluations

0.2 0.2

summari..'.es summari..'.es

mg/kg mg/kg

0.03 0.03

of of

selt-administration selt-administration

mg/kg mg/kg

0.00032-0.30 0.00032-0.30

0.75 0.75

1. 1.

list list

5 5

Cocaine, Cocaine,

0.60 0.60

Cocaine, Cocaine,

the the

onset onset

partial partial

Table Table

opioid opioid

in in

breakpoint. breakpoint.

Column Column

reinforcer. reinforcer.

or or

and and

medication medication

in in

ratio ratio

drug drug

other other

appears appears

cocaine cocaine

agonists agonists

the the

on on

treatment treatment

hrs hrs

on on

30min. 30min.

60min. 60min.

0min. 0min.

60min. 60min.

citation citation

3 3

treatment treatment

self-administration. self-administration. progressive progressive

alphabetically alphabetically

drug drug

full full

drug drug

self-administration. self-administration.

the the

dopamine dopamine

the the

of of

preclinical preclinical

listed listed

in in

of of

cocaine cocaine

is is

of of

opioid opioid

and and

treatment treatment

on on

dose dose

treatment treatment

on on dose dose

dose dose

effects effects

drug drug

Acute Acute

the the

Acute Acute

changes changes

Acute Acute

Acutt> Acutt>

effects effects

the the

study, study,

of of the the

or or

of of

the the

administration administration

to to

effect effect

medications medications

curve, curve,

treatment treatment

effect effect

TM TM

the the

other other

summarizes summarizes

IP IP

mg/kg mg/kg

between between

mg/kg mg/kg IM IM

of of

summarize summarize

IV IV

mg/kg mg/kg

JP JP

2 2 mg/kg mg/kg

of of

agonist-antagonists agonist-antagonists

reference reference

7.5-30 7.5-30

0.1-10 0.1-10

2.0mg 2.0mg

0.003-0.10 0.003-0.10

0.025--0.10 0.025--0.10

dose-effect dose-effect

the the

describes describes

name name Table Table

tables tables

describes describes

6 6

interval interval

effects effects

opioid opioid

8 8

the the

drug drug

the the

The The

These These

the the

3. 3.

232 232

provides provides

mixed mixed

Column Column

table, table,

in in

Column Column

10 10

1-6: 1-6:

5. 5.

and and

each each

shift shift

decanoate decanoate

flupenthixol flupenthixol

ln ln

Tables Tables

(+)-UH (+)-UH

PD128483 PD128483

summarizes summarizes

Flupenthixol Flupenthixol

Cis-

Alpha Alpha

Column Column

subjects. subjects.

ing, ing,

1111111s 1111111s Column Column

onists onists

4 4 self-administration. self-administration. 390 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6

2.0 S-93

1.6 0---0 Cocaine (mg/kg/injection) 1.2 • 0.01 mg/kg ] SCH39166 0.8 -"' 0.o3 mg/kg Pre-Treatment

0.4 • 0.1 mg/kg /t:7 Saline Range Figure 6. Antagonism of cocaine 0.0 self-administration by a dopamine D1 "O receptor antagonist, SCH 39166, m C 1.2 S-78 1.0 individual monkeys. Cocaine self 0 (.) administration was maintained under Q) 1.0 (./) 0.8 a second-order FR30 FI 10 minute ,_ 0.8 schedule. Abscissae: Q) cocaine dose, log a. 0.6 scale; ordinates: rate of responding. 0.6 Cl) Points are means based on the last Q) 0.4 Cl) 0.4 three sessions at each dose of drug. C 0 The horizontal bar represents the a. 0.2 0.2 Cl) mean :±: SD values averaged for all Q) 0.0 0.0 periods of saline substitution in indi a: vidual subjects. From Bergman J, 1.0 Kamien JB, Spealman RD (1990): S-5 1.2 S-290 Antagonism of cocaine self-adminis 1.0 0.8 tered by selective D1 and D2 antago 0.8 nists. Behav Pharmacol 1:355-363. 0.6 Reprinted with permission. 0.6 0.4 0.4 0.2 0.2

0.0 0.0 0.01 0.1 0.01 0.1 Cocaine (mg/kg/injection)

eticlopride and spiperone, resulted in a dose-dependent sponding maintained by cocaine than by alfentanil, and increase in cocaine self-administration. These changes this was interpreted to suggest that cocaine antago in cocaine self-administration were similar to the effect nized the rate suppressant effects of these dopamine an of decreasing the cocaine unit dose, but response pat tagonists. Winger concluded that "there is little evi terns were different from those observed during saline dence to support the notion that DA antagonists acting substitution. Only the D2 antagonist raclopride had no on either the 0 1, the 0 2 or both 0 1 /02 sites produce a effect on cocaine self-administration under these condi competitive antagonism of the reinforcing effects of ei tions (Caine and Koob 1994). Caine and Koob con ther cocaine or alfentanil" (Winger 1994, p. 139). cluded that "finally, it remains to be demonstrated con Flupenthixol is a 0 1 /02 antagonist that produces an clusively that changes in self-administration behavior tidepressant effects clinically. Gawin and colleagues produced by dopamine antagonists can be attributed to (1989) have reported that chronic treatment with low changes in the reinforcing effects of cocaine indepen doses of flupenthixol decreases both cocaine craving dent of changes in other behavioral effects of cocaine and cocaine use in "crack" cocaine smokers. Chronic (e.g., rate altering effects)" (Caine and Koob 1994, p. 217). administration of cis-flupenthixol to rhesus monkeys Similar conclusions were reached by Winger (1994) over 10 days occasionally resulted in a selective de after studying the acute effects of the 0 1 antagonist crease in cocaine self-administration in comparison to a SCH 39166, the 0 2 antagonist eticlopride, and the saline treatment baseline with minimal effects on food mixed 0 1/02 antagonist cis-flupenthixol on the ascend self-administration (Negus et al. 1996). Selective de ing limb of the cocaine and the alfentanil dose-effect creases in cocaine self-administration occurred at rela curve in rhesus monkeys. Larger doses of these dopa tively low doses of flupenthixol (0.0032 mg/kg); how mine antagonists were required to suppress rates of re- ever, these effects were often transient and inconsistent NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6 Evaluation of Treatments for Drug Abuse 391

across monkeys. Higher doses (0.032 mg/kg) tended to opioid agonists and mixed agonist-antagonists. It also is reduce both cocaine and food-maintained responding. possible that cocaine-like drugs could increase compli Although dopamine D1/D2 receptor activity may con ance with behavioral treatment interventions (Balster et tribute to cocaine's reinforcing effects, cis-flupenthixol's al. 1992). A number of dopamine agonists and partial inconsistent effects on cocaine self-administration in agonists have been shown to maintain responding lead rhesus monkeys suggest that it may have limited clini ing to their self-administration. However, reinforcing cal usefulness (Negus et al. 1996). However, in rats, flu effects of treatment medications may be problematic be penthixol decreased progressive ratio breakpoints for cause these may contribute to abuse liability as well as cocaine (Richardson et al. 1994), suggesting an attenua facilitating compliance during chronic treatment. tion of cocaine's reinforcing effects. ( +)-AJ 76 is a dopamine antagonist that is reported to Reinforcing Effects of Dopamine Agonists. The rein be selective for dopamine autoreceptors and is thought forcing effects of dopamine agonists and partial ago to have weak stimulant effects and to increase dopa nists appear to reflect their relative selectivity for or effi mine metabolism (Vanover et al. 1993). In rhesus mon cacy at dopamine receptors. For example, a partial D1 keys, intermediate doses of (+)-AJ 76 increased cocaine receptor agonist (SKF 38393) failed to maintain drug self-administration with a concomitant decrease in food self-administration by rhesus monkeys (Woolverton et intake after the cocaine session. It was concluded that al. 1984), whereas a full D1 receptor agonist (SKF 81297) the changes in cocaine self-administration following was self-administered and its reinforcing effects were acute administration of ( + )-AJ 76 reflected at best a par attenuated by a D1 antagonist (SCH 39166) (Weed et al. tial blockade of cocaine's reinforcing effects (Vanover et 1993). Other D1 agonists (SKF 82958) and (SKF 77434) al. 1993). In rats, acute treatment with (+)-AJ 76 had lit also were self-administered by rats (Self and Stein tle effect on progressive ratio breakpoints over a dose 1992). Dopamine agonists with activity at the D2 recep range of 1.88 to 15 mg/kg, but breakpoints were signifi tor (apomorphine, quinpirole, piribedil and bromocrip cantly reduced at a high dose of 30 mg/kg (Richardson tine) are usually self-administered (Spealman et al. et al. 1993). Because progressive ratio breakpoints often 1992). A dopamine agonist (7-OH-DPAT) with high af are considered to be a sensitive measure of reinforcing finity for D3 receptors was self-administered by co efficacy, these results were consistent with the interpre caine-trained rats (Caine and Koob 1993) and rhesus tation that (+)-AJ 76 antagonized cocaine's reinforcing monkeys (Nader and Mach 1996). Curiously, drug-na effects. Moreover, because rats tended to increase rates ive monkeys could not be trained to self-administer of responding for the first six cocaine injections, these 7-OH-DPAT under the same conditions used to train co findings probably could not be attributed to sedation, caine self-administration, which suggests that this D3 but no alternative reinforcer was studied. agonist may have relatively low abuse liability (Nader In summary, dopamine antagonists have occasionally and Mach 1996). Despite accumulating evidence of the been reported to antagonize the reinforcing effects of importance of D1, D2, and D3 receptors in cocaine's re cocaine and shift the cocaine dose-effect curve to the inforcing effects, their complex interactions, coupled right. However, these cocaine antagonist effects often with the ongoing identification of additional dopamine are inconsistent across subjects, transient, and accompa receptor subtypes, suggest that there is unlikely to be nied by unacceptable adverse side effects. Usually dopa any simple explanation of the mechanisms by which co mine antagonist effects are not selective for cocaine, and caine maintains behavior leading to its self-administra behavior maintained by cocaine as well as by alterna tion (see Schwartz et al. 1992; Woolverton and Johnson tive reinforcers is decreased. Furthermore, these right 1992 for review). ward shifts are small, and self-administration of some The reinforcing properties of structurally dissimilar unit doses of cocaine is increased, as shown schemati drugs with mechanisms of action similar to those of co cally in Figure 4. An antagonist-induced increase in caine also have been examined. In one illustrative cocaine self-administration is not an optimal clinical study, cocaine self-administration was compared to be outcome. Taken together, these findings predict that havior maintained by several drugs that act as indirect dopamine antagonists may not have an ideal pharmaco dopamine agonists by inhibiting dopamine reuptake logical profile for the treatment of cocaine abuse. (Bergman et al. 1989). Self-administration of bupro pion, GBR 12909, methylphenidate, and nomifensine by squirrel monkeys was comparable to cocaine-main Effects of Dopamine Agonists and Partial Agonists tained responding across the dose range studied (Berg on Cocaine Self-Administration [Table 2] man et al. 1989). These findings were consistent with The rationale for treatment of cocaine abuse with co other reports that bupropion, methylphenidate, and caine-like drugs that have less abuse liability and fewer nomifensine were self-administered by rhesus monkeys toxic side effects than cocaine is similar to the rationale (Johanson and Schuster 1975; Winger and Woods 1985) advanced for the treatment of opioid dependence with and that GBR 12909 was self-administered by squirrel

(jJ

?°' $'.'.

0 (I) 0.. Cfl 0 z

11) i:: CJl

-< z m Al C:

.,, 0 n ,> 0 Vl :c .,, -< ,> -0 :c r 0

-0 < r 8 0 °' p z (JC/

°'

!'"

al.

1993

and

eta!.

Balster and

Balster

1993

(continued)

1990a

1995b

Koob

Woolverton

Reference

1979

and 1993

al.

1979

1973 Schuster 1993

1995b

and

1995b 1995a, al. 1995b

Kleven

Caine

Herling

Clowa

Mansbach

Skjoldager

Wilson

Herling

Mansbach

Glowa

Glowa Skjoldager

Glowa

Agonists

Partial

monkey

monkey monkey

monkey monkey Species

monkey

and

Rhesus

Rat

Rhesus Rhesus

Rhesus

Rhesus Rhesus

Rhesus

Rhesus Rhesus

Rhesus

Agonists

response rate

rate

response response rate

response rate

Effect

Reinforcer

Decrease

Decrease Decrease

Decrease

Dopamine

tested

Other Other

Effects

(group

Reinforcer

Food Not

Food

(group

Not

Food Food

Food Not Food

Food

Model:

Drug

Target

injections response

response

rate

Effect rate

response

response response rate

decrease injections rate response

rate

response rate

response response rate

response decrease rate response rate rate

Decrease

Decrease#

Decrease

Transient

Decrease#

Decrease

Transient

Decrease

Self-Administration

0.050

µg/

Drug

µg/kg

Drug

mg/kg

µg/kg

the

0.025 mg/kg

0.25 5.6-17

Target

0.001-0.03 mg/kg

30 0.2 10 kg

0.001-0.03 mg/kg 5.6-17

1-100

5.6-17

Doses/Injection

Cocaine,

in Cocaine,

Cocaine,

Cocaine, Cocaine,

Cocaine,

ad-

hrs

Evaluations

or30

Interval

ministration min.

infusion

min.

Pretreatment

min.

Continuous

Concurrent

30min.

0min.

5min.

0min.

5min. 30min. 30min.

Medication

or or

days

dose

dose dose 12

dose

4-5

Duration

Treatment

Acute

5-13 Acute

Acute

Acute Acute

Acute

Acute Acute

Acute

Treatment

mg/

me/

Preclinical

Dose

mg/kg IV mg/kg IV

mg/kg IM

mg/kg IV mg/kg IV

mg/kg IM mg/kg IM

kg IV mg/kg mg/kg IV

mg/kg IV mg/kg IV

Treatment

8-32 0.8-6.4

0.01-0.30

0.1-5.6

0.1-1.7

0.32-3.2

0.5-4.0

0.03-1.0

0.1-1

0.32-3.2 0.Dl-1.0

0.1-3.0

0.3-3.0

Summary

Self-Administration

Drug

12909

(cocaine

12909

12935

morphine

Treatment

analogue)

Cocaine

Table Bromocriptine

Apo

CFT

Cocaine

d-Amphetamine

cl-Amphetamine

GBR d-Amphetamine

GBR

GBR NEUROPSYCHOPHARMACOLOCY 19%-VOL. 14, l'\O. 6 Evaluation of Treatments for Drug Abuse 393

monkeys (Howell and Byrd 1991). Each drug was similar to cocaine in its rates and patterns of self-administra tion and yielded characteristic inverted U-shaped dose response curves with one exception. Mazindol was self administered by only half of the monkeys studied and produced a number of toxic side effects, including pro fuse salivation, self-mutilation, and distress calls (Berg man et al. 1989). In contrast to these findings in squirrel monkeys, mazindol was self-administered by rhesus monkeys (Wilson and Schuster 1976). A recent compari son of GBR 12909 and cocaine self-administration in rhesus monkeys found that GBR 12909 maintained lower rates of responding at higher unit doses than co caine (Skjoldager et al. 1993).

Effects of Dopamine Agonists on Cocaine Self-Admin istration. Treatment with cocaine and other dopam ine agonists has consistently decreased cocaine self-ad -0 0 ministration in animal models. However, these effects C r..:. often were not selective for cocaine when another rein forcer was available for comparison (see Table 2). For example, the effects of mazindol, a dopamine and nore pinephrine reuptake inhibitor, were compared to those of sertraline and fluoxetine, which inhibit the reuptake of serotonin (Kleven and Woolverton 1993). Cocaine and food self-administration were maintained on an FR 30 schedule. During treatment drug evaluations, each

bJ) t,C monkey was maintained on the dose of cocaine that bl) bJ) .:L "-.. E "-.. "-.. t,C LC) was at or near the peak of its dose-response curve. b.(; b.(; t--._ ::l.. ::l.. E Mazindol (0.4-3.2 mg/kg/24 hr), sertraline (0.1-8.0 :i,' C r0 ,3 0 ~--9 r0 C C 0 C r0 C C LC) LC) ·;e ·~ I LI'; mg/kg/24 hr) or fluoxetine (0.4-3.2 mg/kg/24 hr) was ... ·re N ,; ·r2 N ·~ 5 0 u '7 ci u r0 u V V 0 0 C ci C ci continuously infused through the second lumen of a ...-< u u cS 8 :.J double-lumen catheter for the same number of sessions required for responding to decrease after saline was substituted for cocaine. Doses of test drugs were given in an ascending order in the first monkey tested and presented in an irregular order in the other monkeys. Monkeys returned to saline-treatment baseline levels of

LC) cocaine and food-maintained responding for 1 week or more between successive doses of each treatment drug. Each drug decreased cocaine self-administration in a dose-related manner, but food self-administration also CJ :i, :i, Q.> CJ :i, 2 r.t: 1: '11 CJ 'J'J CJ CJ; CJ l/'l Q.> er, ;:l ;:l 0 ""5 C °5 C ;J C - 0 decreased over the same dose range (Kleven and Wool u -0 uV uv u -0 V -0 e -a <( <( <( <( <( <( verton 1993). Similar findings have been reported in studies where cocaine was administered to monkeys

"-.. given access to cocaine self-administration (see Table 2). bl) -..;; t,C 2: bl) be Ol) t,C LC) E~ E N~ The extent to which the nonselective effects of these ci "-.. 6 o- 0 c<'i"- ~'b.(; I OIJ> ·.c Ol) . ::2 I <:>CU monoamine uptake inhibitors on cocaine and food self J,~ E j_ en 86- I r0 LC) E c N ci ci C ci administration can be attributed to enhancement of the ' aversive effects of cocaine or anorectic effects on food maintained behavior could not be determined with cer tainty from these data (Kleven and Woolverton 1993). f-, f-, r0 0 <( -< Results from studies of the selective dopamine re C ::;, t--._ c.. P... :- r, ...:; v CJ:) LC) 0 Q uptake inhibitor GBR 12909 have been inconsistent C ·o.. t--._ .N .5 r0 ± ± c:: ::, r.i... u... across laboratories. Pretreatment with single doses of ::L 0 0 0 en en t'.. t'.. cocaine (0.32-3.2 mg/kg) or GBR 12909 (0.32-3.2 mg/ 394 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6

kg) had minimal and transient effects on cocaine self the cocaine dose-effect curve and decreased rates of re administration by rhesus monkeys (Skjoldager et al. sponding maintained by food. In the 7-OH-DPAT pre 1993). The greatest suppression occurred during the treatment paradigm, the effects on cocaine self-adminis first 25 minutes of cocaine self-administration sessions, tration were critically dependent on the position of the consistent with the relatively short duration of action of unit dose of cocaine on the cocaine dose-effect curve. each compound. Moreover, doses that were effective in 7-OH-OPAT pretreatment increased cocaine self-admin altering cocaine self-administration were just slightly istration at doses on the ascending limb of the dose below doses that had convulsant effects (Skjoldager et effect curve and decreased cocaine self-administration al. 1993). Skjoldager and coworkers concluded that GBR at doses on the descending limb of the dose-effect curve. 12909 enhanced cocaine's rate decreasing effects but did Food-maintained responding was uniformly decreased not modify its reinforcing effects (Skjoldager et al. under both conditions (Caine and Koob, 1995). 1993). In contrast, GBR 12909 was reported to decrease Partial dopamine agonists, which function as antago cocaine self-administration in rhesus monkeys respond nists at some doses, also decrease cocaine self-adminis ing for food and cocaine under an FR 30 schedule dur tration by squirrel monkeys. For example, pretreatment ing sessions composed of alternating cycles of food and with single doses of the partial 0 1 selective agonist SKF cocaine availability (Glowa et al. 1995a). Under these 38393 shifted the ascending limb of the cocaine dose-re conditions, intermediate doses of GBR 12909 (1-7 mg/ sponse curve to the right in comparison to saline control kg IV) produced a selective decrease in the self-admin treatment (Katz and Witkin 1992). SKF 38393 reduced istration of unit doses of cocaine located on the peak cocaine-maintained responding more than food-main and descending limb of the cocaine dose-effect curve. tained responding in separate groups of monkeys, GBR 12909 was more effective in reducing self-adminis which suggested that the effects of this 0 1 dopamine tration of lower rather than higher unit doses of cocaine agonist were selective for cocaine (Katz and Witkin (Glowa et al. 1995a). In subsequent studies, the effects 1992). These data (Figure 7) are a model study of acute of another dopamine reuptake inhibitor (GBR 12935) as treatment drug effects insofar as the effects of SKF well as CFT, a long-acting cocaine analog and d-am• 38393 were examined on the complete cocaine self-ad phetamine were evaluated on cocaine and food self-ad ministration dose-response curve, as well as on an al ministration under similar conditions (Glowa et al. ternative reinforcer. A dopamine 0 1 agonist with lim 1995b). Unlike GBR 12909, these compounds decreased ited agonist efficacy, SKF 75670, also resulted in a both cocaine- and food-maintained responding. Chronic rightward downshift (Bergman and Rosenzweig-Lip GBR 12909 (1.7 mg/kg) administration over 12 days re son 1992). This low-efficacy agonist was similar to the sulted in a selective and sustained decrease in cocaine 0 1 antagonist SCH 39166 in reducing cocaine's behav self-administration (Glowa et al. 1995b). Amantadine, ioral effects on several measures (Bergman and Rosen another dopamine reuptake inhibitor, failed to decrease zweig-Lipson 1992). cocaine self-administration during chronic treatment In summary, dopamine agonists and partial agonists and did not maintain self-administration behavior in occasionally have been reported to decrease cocaine baboons when substituted for cocaine (Sannerud and self-administration. However, these effects may be ac Griffiths 1988). companied by toxic side effects, including suppression Studies by Caine and Koob (1993, 1995) have examined of the self-administration of other reinforcers. In addi the effects of the 0 3 dopamine receptor-selective agonist tion, many dopamine agonists have reinforcing effects 7-OH-OPAT in rats responding for cocaine under sev that may contribute to their high abuse potential but eral schedules of reinforcement (Caine and Koob 1993, that may also facilitate patient compliance during treat 1995). Initial studies showed that concurrent adminis ment. Finally, some dopamine agonists, such as 7-OH tration of cocaine and 7-OH-OPAT decreased cocaine self OPAT, may actually enhance the reinforcing effects of administration by rats (Caine and Koob 1993). Moreover, cocaine. Partial dopamine agonists may present the the pattern of this decrease was similar to the pattern most promising profile for an effective medication with produced by increasing the unit dose of cocaine. These minimal side effects, because they mimic some effects of 7-OH-OPAT-induced changes in cocaine self-adminis cocaine without producing the toxic effects of full tration were interpreted as evidence that 7-OH-OPAT dopamine agonists or antagonists. increased the reinforcing effects of cocaine. Subsequent studies also suggested that 7-OH-OPAT pretreatment Effects of Opioid enhanced the reinforcing effects of cocaine (Caine and Antagonists, Agonists, and Koob 1995). For example, one group of rats was trained Mixed Agonist-Antagonists on Cocaine to respond for food and cocaine under a FR 5 /Timeout Self-Administration [Table 3] 2-minute schedule during sessions composed of alter nating food and cocaine cycles. Pretreatment with 0.4- Interactions between cocaine and opioids are poorly mg/kg SC of 7-OH-OPAT produced a leftward shift in understood. Because cocaine's reinforcing and discrimi- NECROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6 Evaluation of Treatments for Drug Abuse 395

1.0 120 C z 0.8 100 0 w u I- w c:( UJ a: 80 a: 0.6 ..J w 0. 0 a: 60 UJ I- w z UJ 0.4 0 z u 40 0 0. UJ *- w 0.2 20 a: i 0 -r 0.0 -r C 0.3 1.0 3.0 10.0 S 3.0 10.0 30.0 100.0 SKF 38393 (mg/kg) COCAINE (µg/kg/inj)

Figure 7. Effects of the partial D1 agonist SKF 38393 on behavior maintained by cocaine in squirrel monkeys. Left panel, effects of SKF 38393 on responding maintained under a FR schedule of either cocaine injection (17µg/kg/inj.) or food pre sentation (190 mg/presentation). Solid circles, cocaine-maintained responding. Each circle is the mean of four subjects with typically two determinations at each dose. Open circles, food-maintained responding. Each point is the mean of two subjects with typically two determinations at each dose. Points above C represent rates of responding during the control cocaine injection sessions that preceded test sessions for the subjects maintained on cocaine, and saline-control sessions for the sub jects maintained on food. The average control rates of responding(:+: SE) were 3.87 (:+: 0.54) and 1.22 (:+: 0.33) responses per second under the schedules of food presentation or cocaine injection, respectively. Right panel, effects of pretreatment with saline or 3.0 mg/kg SK 38393 on the cocaine dose-effect curve. Cocaine doses per injection (µg/kg) are shown on the abscissa and responses per second are shown on the ordinate. Solid circles, effects of cocaine dose/injection following pre treatment with saline. Each point is the mean of two subjects with typically two determinations at each dose.Open triangles, effects of cocaine dose/injection following pretreatment with 3.0 mg/kg SKF 38393. Each point is the mean of two subjects with typically two determinations at each dose. Oprn circle above S, average rate of responding maintained by saline injec tion. From Katz JL, Witkin JM (1992): Selective effects of the Di dopamine receptor agonist, SKF 38393, on behavior main tained by cocaine injection. Psychopharmacology 109:241-244. Reprinted with permission.

native stimulus effects are modulated by dopaminergic concurrent cocaine abuse, but lack of compliance by opi neural systems, the effects of opioid agonists and antag oid-dependent patients has limited naltrexone's clinical onists on cocaine self-administration behavior are diffi utility (National Research Council Committee on Clini cult to predict from what is known about the pharma cal Evaluation of Narcotic Antagonists 1978; Meyer and cology of either class of drugs alone. Cocaine often is Mirin 1979; Kosten et al. 1989b; Rosen and Kosten 1991). abused concurrently with opioids, and the combined ef An opioid mixed agonist-antagonist, buprenorphine, fects have been described anecdotally as enhancing the appears to be more effective in reducing combined co positive subjective effects of opioids, and decreasing caine and opioid abuse in severely drug-dependent pa aversive agitation produced by cocaine (Kosten et al. tients (Gastfriend et al. 1993; Mello et al. 1993c; Schot 1986, 1987; Rosen and Kosten 1991; Tutton and Crayton tenfeld et al. 1993; Mello and Mendelson 1996). 1993). One controlled clinical study suggests that mor phine-cocaine combinations result in a unique profile of Opioid Antagonists. Opioid antagonists have had in subjective effects that differ from those of either drug consistent effects on cocaine self-administration in pre alone (Foltin and Fischman 1992). Opioid agonists, an clinical studies (Table 3). Naltrexone, a mu-selective tagonists, and mixed agonist-antagonists have been opioid antagonist, has been studied with conflicting re used clinically to treat opioid abuse and polydrug sults in several species. In rats, acute pretreatment with abuse involving cocaine, often with inconsistent results naltrexone resulted in no change (Ettenberg et al. 1982), (Rosen and Kosten 1991; Tutton and Crayton 1993; Men decreases (Corrigal and Coen 1991a), or increases (Carroll delson and Mello 1996). Methadone, a mu-selective ag et al. 1986) in cocaine self-administration. Chronic nal onist, has not been consistently effective in reducing co trexone treatment in rhesus monkeys selectively de caine abuse by opioid-dependent patients, and if opioid creased cocaine self-administration in comparison to abuse persists, cocaine abuse may increase during treat food-maintained responding, but naltrexone's effects ment (Kosten et al. 1986, 1987, 1989a; Cushman 1988; were not dose-related (Mello et al. 1990b ). In a subse Chaisson et al. 1989; Dunteman et al. 1992). In early quent study, naltrexone had no effect on cocaine self-ad clinical trials, a mu-selective antagonist, naltrexone, ap ministration in rhesus monkeys studied under similar peared to be more effective than methadone in reducing conditions (Mello et al. 1993b). The absence of robust ef-

I I

z z

°' °' 9 9

< <

'"" '"" ,-0 ,-0

r< r<

0 0

,-0 ,-0

'"" '""

r-' r-'

-< -<

°' °' C, C,

0 0

:i, :i,

n n ;,:i ;,:i

2: 2:

:i, :i,

cj cj

0 0

:r: :r: -< -<

:r: :r:

0 0

n n

;}; ;};

;,:i ;,:i

'J) 'J)

C C ro ro

2 2

0.. 0..

::i ::i 0 0 "' "'

CfJ CfJ

§= §=

'.lJ '.lJ

2 2

?; ?;

a, a,

(JQ (JQ

al. al.

t't t't

al. al.

et et

al. al.

et et

al. al.

et et

ct ct

et et

and and

et et

Coen Coen

ct ct

Coen Coen

et et

t't t't

1995 1995

1982 1982

1971 1971

Vries Vries

(continued) (continued)

1992 1992

1995 1995

1995b 1995b

al. al.

1993b 1993b

1991a 1991a

and and 19906 19906

1986 1986

al. al.

1991a 1991a

and and

1971 1971

1978a 1978a

al. al.

Schuster Schuster

Reference Reference

Winger Winger

Glick Glick

Negus Negus

De De

Mello Mello

Corrigall Corrigall

Ettenberg Ettenberg

Carroll Carroll

Killian Killian

Corrigall Corrigall

Woods Woods

Goldberg Goldberg

monkt'y monkt'y

monkey monkey

Species Species

Rhesus Rhesus

Rat Rat

Rhesus Rhesus

Rat Rat

Rhesus Rhesus

Rat Rat

Rhesus Rhesus

Rat Rat

Rat Rat Rat Rat

Rhesus Rhesus

Self-Administration Self-Administration

5 5

on on

in in

Cocaine Cocaine

change change

Table Table change change

chang<' chang<'

Other Other

rate) rate)

(response (response

dose-

curve; curve;

shift shift effect effect

on on

Effect Effect

Rightward Rightward

Reinforcer Reinforcer

No No

See See

2) 2)

Opioids Opioids

2) 2)

of of

tested tested

tested tested tested tested

tested tested

Other Other

(group (group

Alfentanil Alfentanil

Not Not

Reinforcer Reinforcer

Food Food

Not Not

Food Food

Nicotine Nicotine

Food Food

Not Not

Not Not Not Not

Not Not

Not Not Heroin Heroin

Effects Effects

in in

Model: Model:

or or

# #

on on

in in

change change

Drug Drug

injections injections

change change

injections injections

change change

injections injections

change change

Target Target

change# change#

change change

(response (response

rate) rate)

dose-effect dose-effect

shift shift curve; curve;

injections injections

no no

in# in#

Effect Effect

# #

injections injections

rate rate

# #

injections injections

response response

Downward Downward

No No

Decrease Decrease \Jo \Jo

Decreast' Decreast'

No No

Decrease# Decrease#

No No

Increase# Increase#

No No

Decrease# Decrease#

No No

or or

mg/ mg/

Self-Administration Self-Administration

0.10 0.10

Drug Drug

0.3 0.3

mg/kg mg/kg

or or

mg/kg mg/kg

µg/kg µg/kg

mg/kg mg/kg

or or

mg/kg mg/kg

Drug Drug

mg/kg mg/kg

0.001-0.10 0.001-0.10

mg/kg mg/kg 0.4 0.4

0.00032-0.01 0.00032-0.01

kg kg

0.25-1.0 0.25-1.0

0.05 0.05

1.0 1.0 mg/kg mg/kg

0.1 0.1

0.05 0.05

200 200

0.75 0.75

0.3 0.3

0.o25mg/kg 0.o25mg/kg

0.G2mg/kg 0.G2mg/kg

Target Target

the the

Cocaine, Cocaine,

Cocaine, Cocaine, Doses/Injection Doses/Injection

Cocaine, Cocaine,

in in

min. min.

Evaluations Evaluations

Interval Interval

hrs hrs

30min. 30min.

2 2

120 120

30min. 30min.

120 120

l20min l20min

Pretreatment Pretreatment

20min. 20min.

30min. 30min.

0min. 0min.

20min. 20min.

0min. 0min.

Smin. Smin.

Medication Medication

days days

dose dose

dost' dost'

dose dose

Duration Duration

Acute Acute

10 10

Acute Acute

15 15

Acute Acute

Acute Acute Acute Acute

13-19 13-19 Acute Acute

Acute Acute

Treatment Treatment

IV IV

IP IP

IV IV

SC SC

mg/ mg/

IV IV

mg/ mg/

µg/ µg/

IV IV

IP IP

IM IM

mg/kg mg/kg

Dose Dose

mg/kg mg/kg

SC SC kg kg

kg kg

IV IV

mg/kg mg/kg

kg kg IP IP

Preclinical Preclinical

Treatment Treatment Treatment

0.Dl-1.0 0.Dl-1.0

10 10

0.1-3.2 0.1-3.2

0.03-10 0.03-10

0.40 0.40

0.1-10.0 0.1-10.0

0.32-2.30 0.32-2.30

1-4000 1-4000

0.5-2 0.5-2

0.1-10.0 0.1-10.0

1.0 1.0

0.1-10 0.1-10

0.033.0 0.033.0

of of

Summary Summary

antagonists antagonists

3. 3.

Drug Drug

phimine phimine

Treatment Treatment

Quadazocine Quadazocine

Nor-binaltor-

Naltrindole Naltrindole

Naltrcxone Naltrcxone

Naltrexone Naltrexone

Naloxone Naloxone

Naloxonc Naloxonc

NaloxonL· NaloxonL·

Nalorphine Nalorphine

Opioid Opioid Table Table

"I "I

s,C s,C

(Jl (Jl

ro ro

C C o" o"

CJ CJ

.. ..

C C

.. ..

0 0

;:;. ;:;.

(Jl (Jl

ro ro

Pl Pl ro ro g g

.. .. -l -l

...., ...., 0 0

::i ::i

(J(;/ (J(;/ g- Pl Pl

Pl Pl

c c

< < trl trl

I I

CT' CT'

z z

' '

0 0 < <

"" ""

CT' CT'

>< ><

8 8

_.,. _.,.

r< r<

0 0

:,.,. :,.,.

Al Al n n

3:: 3::

:c :c '-:l '-:l

:,.,. :,.,. 0 0

:c :c

>< >< n n

'-:l '-:l

Al Al

[fl [fl

0 0

tT1 tT1

C: C:

z z

al. al.

and and

et et

al. al.

and and

et et

1992 1992

et et

(continued) (continued)

1973 1973

Schuster Schuster

1992 1992

1993d 1993d

1995 1995

1993b 1993b

1993a 1993a

Lac Lac

1992 1992

1990b 1990b

1989 1989

Wilson Wilson

Winger Winger

Mello Mello

Lukas Lukas

Mello Mello

Carroll Carroll

Winger Winger

Mello Mello

Carroll Carroll

Mello Mello

monkey monkey

Rhesus Rhesus

Rat Rat

Rhesus Rhesus

2) 2)

in in shift shift

dose-

change change

decrease decrease

curve curve

dose- effect effect

shift shift

dependent dependent

decrease decrease

(group (group

curve curve

in in effect effect

down down

deliveries deliveries

drug drug

decrease decrease

Downward Downward

Dose-

Transient Transient

Decrease Decrease

No No

Selective Selective

Decrease# Decrease#

Transient Transient

2) 2)

and and

tested tested

non-

(group (group

solution solution

saccharin saccharin

contingent contingent

(group (group

PCP PCP

ethanol ethanol

Not Not

Alfentanil Alfentanil

Food Food

Glucose Glucose

Alfentanil Alfentanil

Saccharin Saccharin

Food Food

in in

rug rug

in in

shift shift

injections injections

dose-effect dose-effect curve curve

shift shift

injections injections

dose-effect dose-effect curve;# curve;#

injections injections

# # decrease decrease

dose-effect dose-effect curve curve

deliveries deliveries

injections injections

Decrease# Decrease#

Downward Downward

Decrease# Decrease#

Right Right

Decrease# Decrease#

Transient Transient

Down Down

Decrease# Decrease# Decreased Decreased

Decrease# Decrease#

IV IV 2 2

or or

0.2 0.2

0.001-

IV IV

0.05-

IV IV

0.001-

0.05 0.05

0.05-

IV IV

0.2 0.2

0.10 0.10

0.05 0.05

0.10 0.10

dose-

base, base,

mg/kg mg/kg

mg/kg/ mg/kg/

mg/kg mg/kg

or or

mg/kg mg/kg

Cl.1 Cl.1

0.10 0.10

0.30 0.30

mg/kg/inj. mg/kg/inj.

inj. inj.

0.10 0.10

continuous continuous access access

mg/kg mg/kg

response, response,

0.001-0.10 0.001-0.10

smoked smoked

mg/kg mg/kg

0.10 0.10

mg/kg mg/kg

0.05 0.05

mg/kg mg/kg

0.05 0.05

Cocaine, Cocaine,

Cocaine Cocaine

Cocaine, Cocaine,

min. min.

5min. 5min.

10 10

120 120

Omin. Omin.

30 30

120min. 120min.

30 30

120 120

1/ 1/

hrs hrs or or

days days

dose dose

days days

dose dose

days days

72 72

48 48

days days days days

days days

dose dose

days days

Acute Acute

10 10

5 5

10 10

60 60

5 5

Acute Acute

5 5

30-120 30-120

15 15

IV IV

mg/ mg/

IV IV

mg/ mg/

agonist-antagonists agonist-antagonists

IM IM

mg/kg mg/kg

IV IV

mg/kg mg/kg

TV TV

mg/kg mg/kg

IV IV

mg/kg mg/kg

kg kg

IV IV

mg/kg mg/kg

TV TV

IM IM

mg/kg mg/kg

IV IV kg kg

mg/kg mg/kg

IV IV

mg/kg mg/kg

0.25-8.0 0.25-8.0

0.001-0.32 0.001-0.32

0.03-1.0 0.03-1.0

0.1-1.0 0.1-1.0

0.40 0.40

0.1-0.40 0.1-0.40

0.001-3.2 0.001-3.2

0.03-0.80 0.03-0.80

0.32 0.32

0.237-0.70 0.237-0.70

0.4-0.7 0.4-0.7

mixed mixed

and and

agonists agonists

Morphine Morphine

Heroin Heroin

Butorphanol Butorphanol

Buprenorphine Buprenorphine

Buprenorphine Buprenorphine Opioid Opioid 398 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6

fects of naltrexone on cocaine self-administration is con sistent with findings from most preclinical evaluations of mu-selective opioid antagonists. Earlier studies of pretreatment with the short-acting opioid antagonists naloxone or nalorphine also found no suppression of cocaine self-administration in rhesus monkeys (Goldberg et al. 1971; Woods and Schuster 1971; Killian et al. 1978a). In addition to mu-receptor selective opioid antago nists, the effects of a delta receptor selective opioid an tagonist, naltrindole, on cocaine self-administration have also been examined in rhesus monkeys (Negus et al. 1995b) and in rats (de Vries et al. 1995). In rats, naltrin dole had no effects on cocaine self-administration (de Vries et al. 1995). In rhesus monkeys, at least one dose of naltrindole (0.1-3.2 mg/kg IV) administered over 10 days decreased self-administration of a maximally rein .... forcing dose of cocaine (0.01 mg/kg/inj.) QJ in three of .... u QJ .... four subjects. Food self-administration was unaffected ..c:: 0 by ... - -0 naltrindole, which suggested that decreases in co 0 ·=QJ 0 i:=: £ caine-maintained responding did not result from a non selective decrease in operant behavior. However, these effects did not occur at all naltrindole doses or in all monkeys. An intermediate dose of naltrindole usually was more effective than either lower or higher doses. Fi nally, the cocaine-antagonist actions of naltrindole were not replicable and were not monotonically related to the naltrindole dose. Although naltrindole may modulate the reinforcing effects of cocaine, these results suggest that delta opioid receptors play a minimal and inconsis tent role in the reinforcing effects of cocaine in rhesus monkeys (Negus et al. 1995b).

Opioid Agonists. The effects of mu opioid agonists, including methadone, on cocaine self-administration have received relatively little attention in preclinical studies (Table 3). Both morphine and heroin reduced cocaine self-administration in rhesus monkeys, but these effects were not selective (Wilson and Schuster 1973; Winger et al. 1992). Kappa- and delta-receptor se lective opioid agonists also may influence cocaine self administration, and there has been increasing interest in 'EQJ 0= s ·.;: examining ... spiradoline se lectively decreased cocaine self-administration without affecting water-maintained responding in rats (Glick et al. 1995). Cocaine self-administration remained signifi cantly reduced for at least 2 days, and in some instances up to 6 days after a single kappa-agonist treatment (Glick et al. 1995). These findings indicate that kappa NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6 Evaluation of Treatments for Drug Abuse 399

agonists can decrease cocaine's reinforcing effects, per our own (Mello et al. 1990b, 1992, 1993a, 1993b). Recent haps by inhibiting cocaine-related increases in extracellu studies in rhesus monkeys indicate that buprenorphine lar dopamine (Glick et al. 1995). These drug self-admin treatment modifies cocaine self-administration on both istration data are consistent with evidence that kappa the ascending and descending limbs of the cocaine agonists antagonize cocaine's behavioral effects in rats dose-effect curve. Daily buprenorphine treatment (0.10 (Heidbreder et al. 1993, 1995; Crawford et al. 1995) as mg/kg/ day) shifted the cocaine dose-effect curve well as cocaine's discriminative stimulus effects in (0.001-0.30 mg/kg/inj.) downward and approximately squirrel monkeys (Spealman and Bergman 1992, 1994). 1 log unit to the right with minimal effects on food Delta agonists and cocaine have similar reinforcing maintained responding (Lukas et al. 1995; see Table 3). effects in drug self-administration procedures Qenck et Because the eventual clinical application of bu al. 1987; Shippenberg et al. 1987; Devine and Wise 1994) prenorphine for drug abuse treatment involves chronic and similar neurochemical effects on extracellular dopa maintenance, we subsequently asked whether tolerance mine levels (Spanagel et al. 1990; Langoni et al. 1991). developed to buprenorphine's effects on cocaine-main However, the effects of delta-receptor selective agonists tained responding. We examined the effects of 30 to 120 on cocaine self-administration are unknown. Delta ago days of buprenorphine treatment (0.32 mg/kg/ day) on nists may be useful as maintenance drugs for cocaine cocaine and food self-administration in six rhesus mon abuse, but no preclinical evaluations have been re keys (Mello et al. 1992). As shown in Figure 8, during ported at this time. Currently, two systemically active the first 15 days of buprenorphine treatment, cocaine delta agonists are available for study, BW373U86, which self-administration decreased by 60 percent below the has mu and delta activity, and SNC 80, a more selective saline treatment baseline level and remained sup delta agonist (see Bilsky et al. 1995; Negus and Picker pressed by 70 to 94 percent over the period of observa 1996 for review). tion. Food-maintained responding was also signifi cantly suppressed at the beginning of buprenorphine Opioid Mixed Agonist-Antagonists. These drugs have treatment, but it gradually returned to and significantly both opioid agonist and antagonist effects, and bu exceeded saline treatment baseline levels. When saline prenorphine, butorphanol, and nalbuphine have been treatment was substituted for buprenorphine, cocaine evaluated in preclinical studies of cocaine self-adminis self-administration resumed, and food self-administra tration (see Table 3). Of these, only buprenorphine has tion remained above pre-buprenorphine baseline levels selective effects on cocaine-maintained responding across (Figure 8). The resumption of cocaine self-administra a broad dose range. Buprenorphine is currently under tion after buprenorphine treatment ended indicates that evaluation for the clinical treatment of opioid abuse (Se the significant decrease in cocaine self-administration gal and Schuster 1995), and it combines the characteris was due to buprenorphine treatment and not to other tics of mu-opioid agonists such as methadone with uncontrolled variables. These data indicate that toler long-acting opioid-antagonist effects similar to those of ance did not develop to buprenorphine' s effects on co naltrexone. This combination makes buprenorphine caine self-administration by rhesus monkeys during safer than opioid agonists alone because the risk for le 100 to 120 days of daily treatment (Mello et al. 1992). thal overdose is minimized (Lewis et al. 1983, 1995). Bu One implication of the finding that buprenorphine prenorphine antagonizes the acute subjective and phys can selectively reduce both opioid and cocaine self-ad iological effects of opioids under many conditions, and ministration by rhesus monkeys is that it may be useful its agonist properties appear to increase patient compli for the treatment of polydrug abuse involving concur ance (see Mello et al. 1993c; Mello and Mendelson 1995 rent opioid and cocaine use. Recent clinical reports indi for review). cate that buprenorphine significantly reduces both co In rhesus monkeys, daily buprenorphine treatment caine and opioid abuse as well as needle use and needle (0.237-0.70 mg/kg/ day) significantly reduced cocaine sharing by polydrug abusers who meet DSM-III-R crite self-administration below saline treatment baseline lev ria for opioid and cocaine dependence (Gastfriend et al. els with minimal effects on food self-administration 1993; Gastfriend et al. 1995; Schottenfeld et al. 1993; see (Mello et al. 1989, 1990b). Food intake during the ses Mello et al. 1993c; Mello and Mendelson, 1995). More sion immediately after buprenorphine treatment was over, in controlled clinical studies, buprenorphine re not suppressed in comparison to saline treatment, and duced the positive subjective effects of speedball combi animals did not appear sedated. We concluded that bu nations of cocaine and morphine (Foltin and Fischman prenorphine treatment selectively reduced cocaine 1995). Further studies of the concordance between clini maintained responding but did not produce a general cal and preclinical studies of buprenorphine's effects on ized suppression of operant behavior (Mello et al. 1989; polydrug abuse will be facilitated by the development 1990b). These findings were subsequently confirmed of a primate model of the simultaneous self-administra and extended in other laboratories (Carroll and Lac tion of heroin and cocaine, i.e., "speedballs"; (Mello et 1992; Carroll et al. 1992; Winger et al. 1992), as well as al. 1995). Preliminary findings from ongoing studies 400 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6

suggest that buprenorphine significantly decreases low antagonism of cocaine's reinforcing properties (Wilson dose cocaine and high-dose heroin combinations more and Schuster 1972; Herling and Woods 1980; Woolver effectively than it decreases speedballs containing high ton and Balster 1981; de la Garza and Johanson 1982; doses of cocaine and low or high doses of heroin (Mello Roberts and Vickers 1984). More recently, the acute ef et al. 1996a). fects of chlorpromazine were reevaluated and compared with the effects of two anti-anxiety drugs, buspirone and gepirone. Gepirone had no effect on cocaine Effects of Other Medications on Cocaine self-admin istration and decreased food self-administration Self-Administration [Table 4] (Gold and Balster 1992). Acute administration of both chlor In addition to compounds with primarily dopaminergic promazine and buspirone resulted in a dose-dependent or opioid activity, the effects on cocaine self-administra increase in cocaine self-administration with no change tion of a wide variety of other compounds with diverse in food self-administration (Gold and Balster 1992). mechanisms of action also have been evaluated. The ef Subsequently, the effects of chronic treatment with bus fects of medications that are used for other indications on pirone and gepirone were examined for 10 days. Nei cocaine self-administration are summarized in Table 4. ther buspirone nor gepirone had any consistent effects For the most part, these medications did not selectively re on cocaine or food self-administration during daily duce cocaine self-administration at doses that did not con treatment (Gold and Balster 1992). This study illustrates currently disrupt food self-administration or produce the importance of comparing acute and chronic treat severe side effects. These data are concordant with find ment regimens rather than drawing conclusions about ings from other behavioral models. The effects of a number potential therapeutic efficacy on the basis of acute med of compounds on cocaine-induced changes in locomotor ication administration. activity and schedule-induced responding as well as other An antidepressant, desipramine, has been used to behaviors have recently been reviewed (see Witkin, 1994). treat cocaine abuse and related affective disorders As noted earlier, before the development of com (Gawin and Kleber 1984), but there remains consider pounds selective for specific dopamine receptor sub able disagreement about the efficacy of desipramine types, clinically available medications with nonspecific therapy based on subsequent clinical evaluations (Weiss dopaminergic activity, used for the treatment of psycho 1988; Fischman et al. 1990; Halikas et al. 1993; Mendel sis, were examined in cocaine self-administration mod son and Mello 1996). In rhesus monkeys, continuous in els. These compounds often increased cocaine self-ad fusion of desipramine over 21 days or more had no ministration, and this was interpreted as evidence for effect on responding maintained by cocaine or an alter-

Figure 8. The effects of 30 to 120 Saline Buprenorphine 0.32 mg/kg/day Saline days of daily buprenorphine treat- 80 120 ment (0.32 mg/kg/day) on cocaine and food self-administration. Days of Cocaine• treatment are shown on the abscissa Food 0 and cocaine injections and food pel- f 100 "Tl >, 60 0 lets per day are shown on the left and cu 0 right ordinates, respectively. Each C a. data U) f .,, point for cocaine injections (solid -C + 80 !. circles) and food pellets (open circles) 0 ;: i' during the pre-buprenorphine saline (.) 40 Q) re.- control period is the average ::+:: S.E. of C ·c Cl) 4 monkeys over 15 days. The first 100 Q) ? 60 '< days of buprenorphine treatment are C an average of data from 4 monkeys, cu + (.) and days 101-120 are an average of 0 20 (.) data from 3 monkeys. These data 40 f were adapted from Mello NK; Lukas SE; Kamien JB; Mendelson JH; Cone EJ (1992): The effects of chronic buprenorphine treatment on cocaine 0 20 and food self-administration by 15 15 30 45 60 75 105 120 90 15 rhesus monkeys. J. Pharmacol Exp Elapsed Time (Days) Ther 260(3):1185-1193. NEUROPSYCHOPHARMACOLOGY 19%-VOL. 14, t"sO. 6 Evaluation of Treatments for Drug Abuse 401

native reinforcer, food, at doses up to 10 times higher have affinity for dopaminergic, cholinergic, serotoner than those used in clinical treatment of cocaine abuse gic, and GABAergic receptor sites (Vanover et al. 1993). (Kleven and Woolverton 1990b ). Similar findings were Acute administration of clozapine to rhesus monkeys obtained when daily doses of desipramine were given increased response rates for cocaine at doses on the de to rhesus monkeys for periods of 5 to 30 days (Mello et scending limb of the cocaine dose-response function al. 1990a). Cocaine injections increased or did not change, without affecting the pattern of responding. This effect and food injections also increased or did not change. was similar to decreasing the unit dose of cocaine, but These preclinical data are consistent with the clinical lit response patterns were not the same as those seen when erature on desipramine treatment, where both stimulation saline was substituted for cocaine. Although an alterna of cocaine use and inconsistent or incomplete attenua tive reinforcer was not available, the effects of clozapine tion of cocaine abuse during desipramine maintenance on food intake subsequent to cocaine self-administra have been reported (see Mello et al. 1990a for review). tion was measured, and food intake was decreased in The antidepressant fluoxetine also has been used clin two of the four monkeys at doses that increased cocaine ically for the treatment of cocaine dependence in heroin self-administration (Vanover et al. 1993). In rats, acute abusers (Pollack and Rosenbaum 1991; Batki et al. clozapine treatment decreased cocaine self-administra 1993). The initially encouraging reports have not been tion (Roberts and Vickers 1984). confirmed in placebo-controlled double-blind trials Ibogaine, an hallucinogen derived from the iboga (Grabowski et al. 1995). Fluoxetine selectively reduced plant indigenous to Africa, is a recent addition to the list cocaine self-administration with no change in food self of potential pharmacotherapies for cocaine abuse. There administration in rats (Carroll et al. 1990b ). Suggestive has been considerable attention to ibogaine in the popu evidence that fluoxetine may decrease cocaine's rein lar press, but controlled clinical trials have not been forcing properties was shown by a progressive ratio conducted, and the anecdotal results reported to date evaluation in which fluoxetine dose-dependently de are, at best, inconclusive. In preclinical evaluations in creased the number of responses rats emitted for a sin rats, acute administration of ibogaine decreased cocaine gle dose of cocaine (Richardson and Roberts 1991 ). self-administration (Cappendijk and Dzolijic 1993). The anticonvulsant carbamazepine has been used for However, when ibogaine's effects on cocaine-main the clinical treatment of cocaine abuse with primarily tained responding were compared with its effects on negative results (see Cornish et al. 1995; Kranzler et al. food, there was a nonselective decrease in both food 1995; Montoya et al. 1995; Mendelson and Mello 1996 and cocaine self-administration (Dworkin et al. 1995). for review). Its clinical usefulness is limited by a num These investigators suggested caution in the use of ber of adverse side effects ranging from dizziness, nau ibogaine for clinical treatment of drug abuse because of sea, and vomiting to aplastic anemia and agranulocyto its adverse motor effects and evidence of cerebellar neu sis (Mendelson and Mello 1996). The rationale for the rotoxicity (Dworkin et al. 1995). use of an anticonvulsant to treat cocaine abuse was based on the observation that carbamazepine reduces cocaine-induced seizures in a "kindling" model in rats TREATMENT MEDICATION (Halikas et al. 1993; Post et al. 1993) and the assumption EFFECTS ON OPIOID SELF-ADMINISTRATION that cocaine craving and withdrawal signs and symp Clinical Treatment of Opioid Abuse toms may reflect a progressive increase in neuronal sen with Opioid Antagonists sitivity. Dietary carbamazepine over 5 days reduced co caine self-administration as well as administration of It is reasonable to assume that a medication that antago food and a glucose saccharine solution (Carroll et al. nizes the subjective and physiological effects of opioids 1990a). Seizures, convulsions, and apparent overdose would reduce opioid self-administration by humans deaths were observed at the highest dose of carba and animals. The compelling logic of this approach led mazepine in combination with cocaine. It was con to the development of naltrexone, a prototypic long-act cluded that the reduction in cocaine self-administration ing mu-opioid receptor antagonist (Blumberg and Day probably reflected the toxic and aversive effects of this ton 1974). In controlled clinical trials, naltrexone re compound (Carroll et al. 1990a). Similar conclusions duced the subjective and physiologic effects of opioids were reached by Sharpe and coworkers in 1992. In that in opioid-dependent men (Martin et al. 19736). Naltrex study, carbamazepine decreased both cocaine and food one also reduced heroin self-administration by heroin self-administration at doses that did not induce severe dependent men studied on a clinical research ward motor incoordination (Sharpe et al. 1992). The nonselec (Meyer and Mirin 1979; Mello et al. 1981). When nal tive and toxic effects reported in these preclinical mod trexone was first introduced into clinical trials, it ap els are concordant with recent clinical evaluations (see peared to be an ideal pharmacotherapy in terms of its Mendelson and Mello 1996 for review). safety, absence of significant side effects, and capacity to Clozapinr, an antipsychotic medication, appears to antagonize opioid effects for 24 to 48 hours (Martin et 0 "" N z

0 Pl

[fl 0.. 2 0

11) C. Vl

(Jq

C Al

Jl -< 0 n :r: 0 :r: Al "' :,, :::::: n :,, 0 r -<

'° ::0 r < 8 0 °' ..... z p

°'

,.,.

and

al. 1984 1992

1984

1992

and

and et

Balster

and

Cocaine

(continued)

Reference

1988

Griffiths Balster

1990a

1992

1972

Schuster

Woodsl980 Vickers

Balster and 1981

Vickers

Sannerud

Gold

Carroll

Sharpe

Wilson

Herling

Roberts

Woolverton

Gold

Roberts

Indications

monkey

Species monkey

monkey

Baboon

Rhesus

Rat

Rhesus

Rat

Rhesus

Rat

Other

for

de- -

change

Other

change

intake

intake response rate

Effect

response rate

or crease choices

Reinforcer

Decrease

Medications

(after

tested

Other tested

session)

saccharin solution

Effects

session)

Reinforcer

Food

Glucose-

Food

Not

Food

Not

Food

Not

Model:

day

high

high rate

rate

high

choices

decrease

Drug

Target

doses:

Effect

doses:

change#

injections

change

injections

injections/

increase; doses:

# injections

response

increase; doses:

# response

injections increase

increase# injections;

doses: #

injections

Low

Decrease#

Decrease

Low

Decrease

Increase# No

Low

Decrease#

Self-Administration

Drug

0.32

0.02

0.2 1.0 Drug

0.1-0.2

0.01-

0.75 100

0.02

0.75

mg/kg

the

mg/kg

in mg/kg

Target

0.05

mg/kg

0.3

µg/kg mg/kg

0.05

mg/kg

Doses/Injection

Cocaine,

Evaluations

Interval

infusion

min.

infusion

Pretreatment

Continuous

15min.

5min.

0min.

5min. Continuous 15

Medication

days

8-10 dose

days

food

doseor5 days

dose days

dose

Duration dose

21-60

Acute

2-4 Acute

Acute Acute

6-10 Acute

Acute

Treatment

160

mg/

Dose

mg/

Preclinical

IV mg/kg

mg/kg IV

mg/kg PO kg

IM mg/kg

mg/kg IM kg

kg/day IV

IV mg/kg

mg/kg

of 0.01-1.0

80,130

7or15mg/

0.125-8.0

0.03-1.0

0.5-2.0 1-8

0.03-0.20

2.0-12.5

Summary

Drug

psychotic)

Treatment Treatment Treatment

(anti-

Parkinsonian agent)

(antianxiety)

(anti-

convulsant) (anti-

convulsant)

(anti

(antipsychotic) (anti

(antipsychotic)

Table

Self-Administration

Amantadine

Buspirone

Carbamazepine

Chlorpromazine

Clozapine

<;,J <;,J

0 0

(1) (1)

I I

"' "'

i: i:

......

i: i: a a

......

o' o'

"' "'

g g

(1) (1)

I') I')

::i ::i

...... (1) (1)

>-l >-l

0 0

......

0 0

2" 2" ;:t. ;:t.

I') I')

tT1 tT1

I') I')

< <

r r cr-

< < z z

r" r"

!J !J

a, a,

0 0

sD sD

>--( >--(

::0 ::0 0 0

3::: 3:::

>, >,

C1 C1

>--( >--(

:r: :r:

:,:i :,:i 0 0

>, >, --0 --0

:r: :r:

n n

tT1 tT1

oq oq

--0 --0

n n CJ) CJ)

0 0 0 0

C: C: :,:i :,:i

z z

,*'" ,*'"

al. al.

et et al. al.

and and

et et

1994 1994

et et

and and

1990 1990

et et

and and

and and Garza Garza

al. al.

Balster Balster

1990a 1990a

la la

(continued) (continued)

1984 1984

Vickers Vickers

et et

1982 1982

and and

& &

Johanson Johanson

1981 1981

1992 1992

Balster Balster

1989 1989

1991 1991

Roberts Roberts

and and

1990b 1990b

Falk Falk 1990a 1990a

n n

Woolverto Woolverto 1993 1993

Richardson Richardson

Roberts Roberts

de de

Woolverton Woolverton

Porrino Porrino

Richardson Richardson

Gold Gold

Kleven Kleven Mello Mello

Tang Tang

Carroll Carroll

Vanover Vanover

monkey monkey

Rat Rat

Rhesus Rhesus

Rat Rat

Rhesus Rhesus

Rat Rat

Rhesus Rhesus

or or

intake intake

change change

effect effect

or or

choice choice

decrease decrease

intake intake

rate rate

change change response response no no

change change no no

No No

Decrease Decrease

No No

Increase Increase

Decrease Decrease

2) 2)

(after (after

tested tested

solution solution

session) session)

(group (group

and and

saccharin saccharin

session) session)

Not Not

Food Food

Not Not

f'ood f'ood

Not Not

Food Food

Clucose Clucose

Food Food

or or

rate rate

# #

dose: dose:

doses: doses:

dose: dose:

change change

doses: doses:

change# change#

change change

effect effect

ratio ratio

injections injections progressive progressive

breakpoint breakpoint

decrease decrease

injections injections

decrease# decrease#

injections injections

high high

increase; increase;

increase increase

choice choice

injections injections

ratio ratio

progressive progressive

breakpoint breakpoint

injections injections

no no

injections injections response response

high high

injections; injections;

increase# increase# decrease# decrease#

Increase# Increase#

Low Low

Decrease Decrease

No No

Decrease# Decrease#

No No Decrease Decrease

No No

lncrease lncrease

Low Low

or or

oral oral

0.60 0.60

0.75 0.75

0.1 0.1

100 100

1.0 1.0

mg/kg mg/kg

0.02 0.02

1.5 1.5

0.2 0.2

0.16 0.16 0.05 0.05

0.025 0.025

0.03 0.03

mg/kg mg/kg

0.05 0.05

mg mg

mg/kg mg/kg

1-1g/kg 1-1g/kg

mg/kg mg/kg

0.05 0.05

mg/kg mg/kg

mg/ml mg/ml

0.10 0.10

or or

0.10 0.10

Cocaine, Cocaine,

Cocaine, Cocaine, Cocaine, Cocaine,

Cocaine, Cocaine,

min. min.

infusion infusion

0min. 0min.

60min. 60min.

5min. 5min.

Continuous Continuous

15 15

30min. 30min.

90min. 90min.

Continuous Continuous

30 30

or or

days days

dose dose

21 21

8-10 8-10 dose dose

days days

dose dose

30 30

dose dose

days days

Acute Acute

6-10 6-10

Acute Acute

2: 2:

25 25

5 5

Acute Acute

IP IP

10 10

1-1g/ 1-1g/

mg/ mg/

IV IV

or or

IP IP

5 5

mg/kg mg/kg

IM IM

mg/kg mg/kg

TM TM

kg/day kg/day

IV IV

mg/kg mg/kg

kg kg

IV IV

kg kg

mg/kg mg/kg

IV IV

mg/kg mg/kg

day day mg/kg/ mg/kg/

IM IM

mg/kg mg/kg

2.5mgIM 2.5mgIM

0.05-0.2 0.05-0.2

0.01-0.08 0.01-0.08

25-400 25-400

0.03-1.0 0.03-1.0

2.5-10 2.5-10

2.5-20 2.5-20

2 2

2.5, 2.5,

0.8-12.8 0.8-12.8 0.56-10.0 0.56-10.0

0.03-3.0 0.03-3.0

(antipsychotic) (antipsychotic)

(antidepressant) (antidepressant)

(antianxiety) (antianxiety)

(antidepressant) (antidepressant)

(antidepressant) (antidepressant) (antidepressant) (antidepressant)

Haloperidol Haloperidol

Fluoxetine Fluoxetine

Gepirone Gepirone

Desipramine Desipramine Desipramine Desipramine

Desipramine Desipramine

(antipsychotic) (antipsychotic) Clozapine Clozapine .i:- .i:-

0 z

(E.. tl) 0 ;:J 0...

CJ) 0 I

C z rt, Vl

':J z -< Al c::

m Al Cf) n

::r: 0 ':J 0 ::r: -< n -D ::0 < 0 r 0 >---' 8 p z °' °'

a,:;

!'"

and

al.

and

1989

1995

1979

al. Dzolijic 1993

and

al.

1973

1992

Schuster 1973

Schuster

al.

and French 1973 1995 Schuster

1984

Vickers 1973 Schuster

Reference

Roberts

Dworkin

Cappendijk

Wilson

Lane

Herling Wilson

Wilson

Bourland

Roberts

Wilson

Species monkey

monkey monkey

monkey

Rat

Rhesus

Rhesus Rat

Rhesus

Rat

Rhesus

doses:

Other

doses: decrease response increase; high

rate

Effect

Decrease

Low

tested

tested tested

tested

Other

(separate

group)

(group

Reinforcer Reinforcer

Not

Not Food

Not

food

Not

break-

doses:

break-

doses:

Drug

doses:

Target

change#

Effect

point

progressive ratio injections

injections

increase;

decrease response rate injections high

injections;

progressive decrease ratio injections point injections increase;

decrease#

high

Decrease

Decrease#

Low

Decrease#

Increase#

Low

1.2

Drug

0.33

0.2

0.125-

0.2

0.2 0.3-0.9

0.25

0.75

0.2

mg/inj.

mg/kg

mg/inj. Target

mg/kg

0.5

mg/kg

µg/kg

mg/kg

0.50

mg/kg

Doses/Injection

Cocaine

Cocaine,

min.

Interval

min.

Pretreatment

30min.

60-90

30min.

5min.

0min.

5min.

60min.

5min.

dose

days

and

dose

dose and6-ll days

dose

Duration

Treatment

Acute

mg/

mg/ mg/

mg/

II'

IP IM

mg/

Dose

mg/kg IP

mg/kg IV kg

mg/kg IM kg kg

mg/kg IM

Treatment

10-40 0.025-0.05

1.0-50 40-80

0.001-1.0

2.5-30

3.0-17.8

2-12

2.5-15

2.5-10

0.005-0.40

(anti-

(continued)

Drug

barbital

Treatment

(antipsychotic)

(antiemetic) depressant)

(sedative)

(anorectic)

(antipsychotic)

Table

Haloperidol

lbogaine

Ibogaine

Imipraminc

Pen

Ondansetron Phenmetrazine l'entobarbital

Remoxipride

Thioridazinc

Trifluoperazine NEUROPSYCHOPHAR:VIACOLOGY 1996-VOL. 14, NO. 6 Evaluation of Treatments for Drug Abuse 405

al. 1973b; Julius and Renault 1976; Verebey et al. 1976). Effects of Opioid Antagonists on Opioid Yet, naltrexone proved to be relatively ineffective in Self-Administration [Table 5] outpatient clinical trials, because it was difficult to re tain opioid-dependent patients in naltrexone treatment Opioid antagonists include such compounds as nalox programs (see National Research Council 1978; Meyer one, naltrexone, nalorphine, and quadazocine, and the and Mirin 1979; Schecter 1980 for review). Today, nal effects of opioid antagonists on opioid self-administra trexone is an effective treatment for highly motivated tion have been extensively evaluated in preclinical opioid-dependent patients, usually medical profession models of opioid abuse. Most of these studies have ex als, who must remain opioid-free to retain clinical licen amined only a single unit dose of the opioid agonist sure (Jaffe and Martin 1990). The failure of naltrexone as (Weeks and Collins 1964, 1976; Goldberg et al. 1971; a wide-scale treatment for opioid addiction is instruc Woods et al. 1975; Griffiths et al. 1976; Killian et al. tive for the development of future pharmacotherapies 1978a, 1978b; Ettenberg et al. 1982; Koob et al. 1984; Ne for drug abuse treatment. Although naltrexone antago gus et al. 1993a). However, the effect of an opioid antag nized opioid effects, it had no positive mood-modulat onist such as naltrexone on opioid self-administration ing effects, and it could be abruptly discontinued without critically depends on the unit dose of opioid used to adverse consequences. It appeared that the pharma maintain responding. For example, Weeks and Collins cological blockade of opioid effects, without other ago trained morphine-dependent female rats to respond for nist properties, was not an effective treatment for most 10-mg/kg/injection morphine under a FR 10 schedule opioid-dependent patients, even in the context of a that was in effect 24 hours/ day (Weeks and Collins multimodality treatment program. 1964, 1976). Initial experiments revealed that decreasing the unit dose of morphine resulted in an increase in the number of injections per day, indicating that the 10- Opioid Self-Administration Dose-Effect Curves mg/kg/ injection unit dose was on the descending limb Experimental animals will self-administer a wide range of the morphine dose-effect curve. Continuous infusion of opioid agonists, including morphine, heroin and co for 24 hours with the opioid antagonists nalorphine deine, and synthetic opioid analgesics such as alfenta (0.006-1.0 mg/kg/hour; Weeks and Collins 1964) and nil, methadone and hydromorphone (Weeks and Col naloxone (0.056-0.56 mg/kg/ day; Weeks and Collins lins 1964; Woods and Schuster 1968; Woods and 1976) produced a dose-dependent increase in the number Schuster 1970; Deneau et al. 1969; Downs and Woods of morphine injections per day, an effect that mimicked 1974; Jones and Prada 1977; Sanchez-Ramos and a decrease in the unit dose of morphine. Schuster 1977; Harrigan and Downs 1978; Herling 1981; Harrigan and Downs (1978), in contrast, trained Young et al. 1981; Mello et al. 1983; Negus et al. 1993b). rhesus monkeys to self-administer morphine (8.0 µg/kg/ There are at least three major types of opioid receptors, injection), d-amphetamine (4.0 µg/kg/injection), and sa the mu-, delta-, and kappa-opioid receptors (Martin et line on alternate days (3 days morphine, 2 days d-am• al. 1976; Lord et al. 1977; Mansour et al. 1987), and, for phetamine, 2 days saline). The drug solutions were avail the most part, the opioid agonists that have been used able under a FR 1 schedule of drug delivery during 15- to maintain opioid self-administration are selective for minute sessions every 4 hours (8 sessions per day). Under mu receptors. In general, the dose-effect curve relating these conditions, continuous infusion with the antago dose per injection of these opioids to some metric of nist naltrexone (0.25-10.0 µg/kg/hour) for 4 weeks pro drug self-administration (i.e., response rate, injections duced a selective and dose-dependent decrease in mor per session) displays an inverted-U shape, with inter phine self-administration, an effect opposite to that mediate doses maintaining more self-administration obtained by Weeks and Collins (1964, 1976). However, behavior than either lower or higher doses (see Figure in monkeys, the 8.0 µg/kg/inj. unit dose of morphine was 2). As discussed earlier, the slopes of the ascending and located at the peak of the ascending limb of the mor descending limbs of the dose-effect curve, as well as the phine dose-effect curve, such that decreases in the unit position of the dose-effect curve on the abscissa, can dose of morphine led to a decrease in morphine self-ad vary dramatically as a function of several experimental ministration (Harrigan and Downs 1978). Thus, in both parameters, including the type of opioid serving as the studies the effects of opioid antagonist treatment on mor reinforcer and the schedule of drug delivery. As a result, phine self-administration mimicked the effects of decreas it is important to consider the underlying opioid dose ing the unit dose of morphine. In addition, Harrigan and effect curve when interpreting the results produced by Downs (1978) reported that naltrexone treatment did pretreatments with various potential treatment medica not alter cl-amphetamine self-administration, suggest tions. The following discussion of test drug effects on ing that effects of naltrexone were selective for opioid opioid dose-effect curves considers three types of po self-administration. Other studies also have reported a tential treatment medications: opioid antagonists, opi selective effect of opioid antagonists on opioid self-ad oid agonists and nonopioids. ministration (Ettenberg et al. 1982; Winger et al. 1992). .... 0a, Table 5. Summary of Preclinical Treatment Medication Evaluations in the Drug Self-Administration Model: Effects of Opioid Antagonists on Opioid Self-Administration z

Effect on Effect on

Treatment Treatment Treatment Pretreatment Target Drug Target Other Other 0 Drug Dose Duration Interval Doses/Injection Drug Reinforcer Reinforcer Species Reference ll) ::i p,. Beta-funaltrex- 5.0-20.0 Acute 20 hrs Heroin, 60 Increase# Not tested Rat Negus et al. Cfl amme mg/kg dose µ,g/kg injections 19936 0 SC z l'tl Nalorphine 0.006-1.0 20-28 hrs Continuous Morphine, 10 Increase# Not tested Rat Weeks and '.]q C mg/kg/ infusion mg/kg injections Collins en hr IV 1964 Nalorphine 3-3,000 Acute 'imin. Morphine, 100 Low doses: Cocaine See Table 3 Rhesus Goldberg et µ,g/kg dose µ,g/kg increase; (group 2) monkey al. 1971 lV and 5 high doses: days decrease# injections Naloxonazine 7.5-30 mg/ Acute 20 hrs Heroin, 60 Increase# Not tested Rat Negus et al. kg!V dose µ,g/kg injections 1993b Naloxone 1-100 µ,g/ Acute 5min. Morphine, 100 Low doses: Not tested Rhesus Coldberg et kg IV dose µ,g/kg increase; monkey al. 1971 high doses: decrease# injections Naloxone 56-560 µ,g/ 20-28 hrs Continuous Morphine, 10 Increase# Not tested - Rat Weeks and kg/day infusion mg/kg injections Collins lV 1976 Naloxone 0.09-0.75 1 day Continuous I Iernin, 0.32 or Increase# Food Decrease# Baboon Griffiths et z tT1 mg/kg/ infusion 0.96 mg/kg heroin food al. 1976 C choices choices 7l day IV 0 u Naloxone 0.01-4.0 13-19 days 5min. Heroin, 6 µ,g/kg Low doses: Cocaine See Table 3 Rhesus Killian et al. CJ') -( mg/kg increase; monkey 1978a n :r: lM high doses: 0 "O decrease# :r: injections >- Naloxone 0.05-30 Acute 15min. Heroin, 60 Low doses: Not tested Rat Koob et al. mg/kg dose µ,g/kg increase; 1984 n>- 0 IP high doses: r decrease# 8-( injections Naltrexone 0.25-10 4 weeks Continuous Morphine, 8 Decrease# Mctham- No change Rhesus Harrigan '° '°°' µ,g/kg/ infusion µ,g/kg injections phet- monkey and .k hr IV amine Downs 0r 1978 ,*"' (continued) z p °'

-..:i -..:i

""' ""'

0 0

r1l r1l

[Jl [Jl

C C

u u

......

C C

......

0 0

"' "'

a a r1l r1l

g g

......

r1l r1l I)) I))

......

-l -l

::i ::i 0 0

5· 5·

oq oq

2" 2"

I)) I))

< < trl trl

9 9

°' °' z z

0 0

< <

'° '° "' "'

'f 'f

-< -<

r r

R R

0 0 ,,. ,,.

n n

$'. $'.

;;c ;;c

,,. ,,.

"iCI "iCI

:r: :r:

::c ::c

0 0 -< -<

en en

"iCI "iCI

el el n n

C: C: tTl tTl

z z

al. al.

ct ct

al. al.

et et

al. al.

et et

1982 1982

1992 1992

1989 1989

Woods Woods

and and

1995 1995

1993b 1993b

1984 1984

al. al.

1981 1981

Downs Downs

and and

1978 1978

Downs Downs

and and

Winger Winger

Bertalmio Bertalmio

Glick Glick

Negus Negus

Koob Koob

Ettenberg Ettenberg

Herling Herling

Harrigan Harrigan

monkey monkey

Rhesus Rhesus

Rat Rat

Rhesus Rhesus

3 3

Table Table

change change

rate rate

response response

reported reported

See See

Increase Increase

Not Not

No No

2) 2)

tested tested

(group (group

amphetamine amine

amine amine

phet-

Cocaine Cocaine

Not Not

Cocaine Cocaine

Food Food

Meth-

Metharn-

and and

dose-

and and

dose-

curve curve

dose-

curve curve

dose-

in in

curve curve in in

curve curve

in in

change change

injections) injections)

change change

injections) injections)

rate) rate) (response (response

shift shift effect effect

rate) rate) (response (response

effect effect

shift shift

injections injections

rate) rate)

(response (response

shift shift effect effect

(# (#

downward downward

shift shift effect effect

downward downward

(# (#

shift shift effect effect

downward downward

Rightward Rightward

l\o l\o

No No

Increase# Increase#

Rightward Rightward

0.03-

0.32-

0.04 0.04

8-800 8-800

60 60

0.25-64 0.25-64

µ.g/kg µ.g/kg

100 100

320 320

mg/kg mg/kg

µ.g/kg µ.g/kg

625 625

Alfentanil, Alfentanil,

Morphine, Morphine,

Heroin, Heroin,

Codeine, Codeine,

Heroin, Heroin,

Morphine, Morphine,

min. min.

hrs hrs

infusion infusion

30 30

30-60 30-60

2 2

15min. 15min.

30min. 30min.

15 15

Continuous Continuous

dose dose

weeks weeks

dose dose

weeks weeks

Acute Acute

4 4

6-8 6-8

IV IV

mg/kg mg/kg

IV IV

mg/kg mg/kg

SC SC

mg/kg mg/kg

IV IV

mg/kg mg/kg

IP IP

mg/kg mg/kg

IP IP

mg/kg mg/kg

IM IM

mg/kg mg/kg

µ.g/kg/ µ.g/kg/

hrIV hrIV

µ.g/kg/ µ.g/kg/

hr hr

0.01-1.0 0.01-1.0

0.1-1.0 0.1-1.0

10 10

5.0-10 5.0-10

1.0-17 1.0-17

0.05-30 0.05-30

0.01-10 0.01-10

0.003-0.32 0.003-0.32

1.0-20 1.0-20

0.25-20 0.25-20

phimine phimine

Quadazocine Quadazocine

Quadazocine Quadazocine Nor-binaltor-

Nor-binaltor-

Naltrindole Naltrindole

Naltrexone Naltrexone

Naltrcxone Naltrcxone

Naltrexone Naltrexone

Naltrexonc Naltrexonc Naltrexone Naltrexone 408 N.K. Mello and S.S. .\Jegus NEUROPSYCHOPHAR'\1ACOLOGY 1996-VOL. 14, NO. 6

When evaluated against complete opioid agonist ministration of alfentanil by rhesus monkeys (Bertalmio dose-€ffect curves, opioid antagonists have been reported and Woods 1987; 1989; Winger et al. 1992). It is not clear to produce both rightward and downward shifts. Harri whether these different results (i.e., rightward vs. gan and Downs (1978) also evaluated the effects of con downward shifts) reflect differences in the nature of the tinuous naltrexone infusion on the self-administration antagonists or differences in the procedures used to of a range of morphine doses (2.0--625 µg/kg/injection). evaluate antagonist effects. Chronic infusion of a relatively low dose of naltrexone Naltrexone, quadazocine, and most antagonists that (1 µg/kg/hour) produced an approximately two-fold have been evaluated for their effects on opioid self-ad rightward shift in the morphine dose-effect curve and a ministration are selective for mu-opioid receptors. The small (i.e., less than 10'¼,) decrease in the maximum ef effects of mu-, delta-, and kappa-selective antagonists fect. Chronic infusion of a higher dose of naltrexone (10 on heroin self-administration were compared in rats µg/kg/hour) shifted the entire morphine curve down trained to self-administer 0.06 mg/kg/injection of her ward so that no dose of morphine maintained rates of oin under a FR 5 schedule during daily, 3-hour sessions self-administration different from those maintained by (Negus et al. 1993a). Initial experiments indicated that saline. As shown in Figure 9, similar results have been this unit dose of heroin (0.06 mg/kg/inj.) lay on the de obtained with chronic and continuous naltrexone infu scending limb of the heroin dose-effect curve. The mu sion on the dose-effect curve for heroin self-administra selective antagonist ~-funaltrexamine produced the ex tion in rhesus monkeys (Harrigan and Downs 1981), pected, dose-dependent increases in heroin self-admin with low doses of naltrexone shifting the dose-effect istration, and high doses of ~-funaltrexamine produced curve to the right and higher doses of naltrexone shift extinction patterns of responding. These effects resem ing the heroin dose-effect curve rightward and down bled the effects of decreasing the unit dose of heroin ward. Finally, Herling and Woods (1980) reported that and suggested that ~-funaltrexamine antagonized the acute IM naltrexone injections produced rightward and reinforcing effects of heroin. The delta-selective antago downward shifts in the dose-effect curve for codeine nist naltrindole also produced a dose-dependent in self-administration in rhesus monkeys. In contrast to crease in the self-administration of 0.06 mg/kg/inj. of these effects of naltrexone, acute pretreatment with the heroin; however, relatively high doses of naltrindole opioid antagonist quadazocine produced only parallel were required to produce these effects, suggesting that rightward shifts in the dose-effect curve for the self-ad- naltrindole may have increased heroin self-administra-

-0- Heroin Alone -0- Heroin Alone ----+--- + 1 µg/kg/hr naltrexone + 4 mg/kg/day methadone ____._ + 5 µg/kg/hr naltrexone -----____._ + 12 mg/kg/day methadone ---- + 20 µg/kg/hr naltrexone ---- + 24 mg/kg/day methadone 50 50

40 40

U) C 30 30 0 .:; u G.I 20 20 'i:

10 10

0 0 .1 10 100 .1 10 100

Unit Dose Heroin (mg/kg/inj)

Figure 9. Self-administration of heroin by rhesus monkeys during chronic infusion with either the opioid antagonist nal trexone (left panel) or the opioid agonist methadone (right panel). Abscissae: Unit dose in mg/kg/inj (log scale). Ordinates: Injections during each access period. Monkeys responded for drugs under a FR 1 (continuous reinforcement) schedule with 15-minute access periods occurring every 4 hours. Each point in the left panel shows mean data from four monkeys. Each point in the right panel shows data from one monkey. Adapted from Harrigan and Downs (1981): Pharmacological evalua tion of narcotic antagonist delivery systems in rhesus monkeys. NIDA Research Monograph 28. Washington, DC: Govern ment Printing Office, pp. 77-92. NECROl'SYCHOPHARMACOLOCY 1996-VOL. 14, \JO. 6 Evaluation of Treatments for Drug Abuse 409

hon by acting at mu receptors. The kappa-selective an minimal capacity to induce physiological dependence tagonist nor-binaltorphimine had no effect on heroin in man (Jasinski et al. 1978; Houde 1979; Lewis, 1995). self-administration (Negus et al. 1993b). These data Assessments of buprenorphine for treatment of opioid suggest that the reinforcing properties of heroin are pri abuse were first conducted in inpatient clinical studies marily mediated by mu-opioid receptors. and subsequently in preclinical drug self-administra In summary, low doses of mu-selective opioid antago tion models. In 1978, Jasinski and coworkers reported nists typically produce rightward shifts in opioid ago that buprenorphine antagonized the physiological and nist dose-effect curves. Higher antagonist doses may subjective effects of morphine (60-120 mg/day) for up produce further rightward shifts or may shift opioid to 29.5 hours in abstinent opioid-dependent men (Jasin dose-effect curves downward. When the effects of an ski et al. 1978). Subsequent inpatient studies indicated tagonists are evaluated against single unit doses of an that buprenorphine (8 mg/ day SC) significantly re opioid agonist, either increases or decreases in opioid duced heroin self-administration by heroin-dependent self-administration may be observed depending on the men who had abused heroin for almost 10 years (Mello position of the agonist dose along the dose-effect curve. and Mendelson 1980; Mello et al. 1982). These findings emphasize the importance of character In 1992 the first controlled outpatient comparison of izing the agonist dose-effect curve as a basis for inter buprenorphine and methadone maintenance treatment preting the effects of test drugs. There is no evidence to for opioid dependence was reported (Johnson et al. suggest that the effects of opioid antagonists on opioid 1992). In a randomized, double-blind, parallel group self-administration change appreciably during chronic study, it was found that buprenorphine (8 mg/ day SL) treatment, although this issue has not been examined was as effective as 60 mg/ day of methadone and supe systematically. The effects of opioid antagonists on opi rior to 20 mg/ day of methadone on two outcome mea oid self-administration are usually selective in that opi sures: patient retention and opioid-free urines (Johnson oid antagonists are more effective in altering opioid et al. 1992). During the first 2 weeks of treatment, opi self-administration than in altering self-administration oid-positive urines decreased more rapidly in buprenor of nonopioid reinforcers. Thus, opioid antagonists ap phine-treated patients than in methadone-treated pa pear to alter opioid self-administration while producing tients, but remained 47% and 56% opioid positive over few other adverse effects. However, opioid antagonists the 17 weeks of treatment. A subsequent comparison of may precipitate opioid withdrawal signs and symp lower doses of buprenorphine (2 and 6 mg) with metha toms in people physically dependent on opioids. done (35 and 65 mg) found that methadone was supe rior in reducing opioid use and in treatment retention (Kosten et al. 1993). Currently, buprenorphine Clinical Treatment of Opioid Abuse is the only opioid mixed agonist-antagonist with Opioid Agonists that is being evaluated for the treatment of opioid self-administration in large Among the alternatives to opioid antagonists for treat scale clinical trials. A significant reduction in opioid ment of opioid dependence are opioid agonists and opi self-administration has been reported in clinical studies oid mixed agonist-antagonists. Opioid agonists such as as well as preclinical evaluations. Recent preclinical and methadone and 1-a-acetyl methadol (LAAM) produce clinical reports have indicated that buprenorphine also opioid-like subjective effects and attenuate responses to reduces cocaine self-administration by rhesus monkeys heroin and other opioids (Dole and Nyswander 1965; and concurrent cocaine abuse by opioid-dependent pa Jaffe and Senay 1971; Jaffe et al. 1972; Martin et al. tients (Kosten et al. 1989a, 1989b; Mello et al. 1989, 1973a; Judson and Goldstein 1979; Jones and Prada 1990b; Gastfriend et al. 1993; Schottenfeld et al. 1993). 1975; see Jaffe and Martin 1990 for review). Methadone These findings were discussed earlier in the section on is the major pharmacotherapy available for drug abuse opioid effects on cocaine self-administration (see Mello treatment, and its clinical utility is well established et al. 1993c; Mello and Mendelson 1995 for review). (Dole 1988). More recently, partial mu agonists or opi oid mixed agonist-antagonists such as buprenorphine Effects of Opioid Agonists on Opioid have been evaluated for treatment of opioid depen Self-Administration [Table 6] dence (see Rosen and Kosten 1991; Mello et al. 1993c; Segal and Schuster 1995 for review). Buprenorphine an Many studies since the mid-1970s have evaluated the tagonizes opioid effects but also has agonist properties effects of opioid agonist treatments on opioid self-ad that make it more acceptable to patients than an opioid ministration (Weeks and Collins 1964; Griffiths et al. antagonist such as naltrexone. Buprenorphine's opioid 1976; Jones and Prada 1977; Stretch 1977; Wurster et al. antagonist properties confer a greater margin of safety 1977; Harrigan and Downs 1981; Lukas et al. 1981; than opioid agonists, [i.e., overdose did not occur at 10 Mello et al. 1983; Koob et al. 1986; Negus et al. 1992; times the analgesic dose (Banks 1979; Lewis et al. 1983)]. Winger et al. 1992; see Table 6). In an early study, Weeks Buprenorphine has potent analgesic properties with and Collins (1964) examined the effects of etonitazene, ,!>, .... z 0

0. 0 Ill

::i

[Jl 0 z ro

C [JJ

(JQ

-0 C: -< ;,:J 0 (fl ::r: "Cl n ,,. 0 ;,:J ::r: ,,.

n 0 r -<

,-..; "' "' 8 0 < ['" °'

°'

al.

eta!.

and

1976

Downs

and 1981

1983

1992

1964 Collins 1992

1981

Reference 1964

Collins al.

Prada 1977 Downs and

1981

1983

(continued)

Harrigan

Mello

Winger

Weeks Winger

Lukas

Weeks

Griffiths

Harrigan Jones

Mello

monkey

Species

monkey

Rhesus

Rat

Rhesus

Rat

Baboon

Dog

Rhesus

Table

change pellets

Table

reported

increase or

Effect

food choices

reported

food

pellets

Reinforcer

Not

See

Increase#

Not

Decrease#

tested

Other Other

amphetamine

(group

amphetamine

Reinforcer

Meth-

Food

Cocaine

Not

Cocaine

Not

Not Food

Not Meth-

Food

( # (

dose-

curve(#

curve

Drug

Target

Effect

change

shift injections) effect injections

shift (response effect rate)

injections

shift (response effect rate)

injections

choices decrease heroin injections

shift effect

injections)

Downward

Decrease#

Downward

Decrease#

Downward

Decrease#

Temporary Downward No

Self-Administration

mg/

Opioid

mg/ or

2-625

0.32-

Drug

0.01

Hydro-

0.32

0.25-64

0.01

mg/kg

µg/kg

Target

µg/kg

kg morphone

0.02

mg/kg

0.96

Doses/Injection kg µg/kg

Hydromor- phone,0.02 0.02mg/kg mg/kg

Morphine,

Heroin,

Alfentanil,

Morphine,

Alfentanil,

Morphine, Morphine,

Heroin,

Morphine,

Heroin,

Agonist-Antagonists

Interval

min.

infusion

infusion infusion

Pretreatment

Continuous

120min.

30min.

Continuous 10min.

0min.

Continuous

Continuous 120

Opioid

Mixed

days

davs each

days each

(5 days

and dose)

dose

months

dose

day

Duration

day (5 at

dose) weeks

weeks

Treatment

4weeks

Acute

12 1

10-11

6months

Agonists

mg/ or

mg/

Dose

mg/kg mg/

mg/

mg/kg/ 0.040

hr mg/kg/ 0.789

day IV mg/kg

kg/hr

IV mg/kg

intra- gastric IV kg/hr,

IV kg/day kg/day

IV kg/day IV

11.86

mg/kg IV

Treatment

Opioid

0.020

0.014-

0.001-3.2

1-32

0.001-0.32

1--4

8.3 7--48.4

4-24

0.179-

Effects

Drug

Treatment

Table

Buprenorphine

Codeine Heroin

LAAM

Meperidine

Methadone Methadone

Methadone

Methadone NEUROPSYCHOPHARMACOLOCY 1996-VOL. 14, NO. 6 Evaluation of Treatments for Drug Abuse 411

morphine, codeine, and meperidine on morphine self -;;; rii ...... administration by rats. Morphine-dependent female (J Lf") (l/ Lf") -><: 0-- -><: 0-- rats were trained to respond to 10 mg/kg/inj. of mor u ,.....0-- u ,.....0-- [5 G phine, a unit dose on the descending limb of the mor phine dose-effect curve. Adding etonitazene (10 µg/ ml) to the rats' drinking water for 24 hours decreased the number of injections per day to approximately 30% of

bO control levels. Similarly, continuous infusion of mor 0 0 phine, meperidine, or codeine (1-32 mg/kg/hour) for 24 hours produced a dose-dependent decrease in mor phine self-administration (Weeks and Collins 1964). Al though these data suggest that treatment with an opioid agonist may alter the reinforcing effects of morphine, the selectivity of these effects was not evaluated. It is also possible that the opioid agonists may have pro duced a nonselective disruption of behavior that im paired the subjects' ability to emit an operant response. More recent studies have addressed this issue by em ploying procedures that examine test drug effects on self-administration of both opioid and nonopioid rein forcers. For example, Griffiths, Brady, and colleagues (Griffiths et al. 1976, 1981; Wurster et al. 1977) devel cf) '11c C oped a choice procedure in which baboons selected

<: '-><: ·-C ------bC ..c: "- ..c: .3 :i tion decreased while food self-administration increased 0.. ·:=:C ------bO ..., cc e,. ~' C N C E C.:oL. E E Jc'. C') indicates that the decrease in heroin self-administration ;;:;: ;;:;: :c <1'. did not occur because the baboons were unable to re spond. Rather, these results suggest that continuous in fusion with either morphine or methadone selectively decreased the reinforcing effects of heroin. Further more, these findings suggest that tolerance did not de velop to the opioid-induced decreases in heroin self-ad 0 ministration during chronic opioid maintenance. This last conclusion contrasts with a study by Jones and

•J; >. Prada (1977) reporting that tolerance did develop cf) c:; -><: :.., c., v within 2 weeks to opioid-induced decreases in mor >, c., .8 !: CJ ;::J v 3: "i u "O phine self-administration in dogs. N ['-.. <1'. In a related study, the effectiveness of buprenorphine and methadone in reducing the self-administration of heroin (0.01-0.02) and dilaudid (hydromorphone; 0.02 mg/kg/inj.) by rhesus monkeys was compared (Mello et al. 1983). Drugs and food were available in alternat ing daily sessions, and each drug or food delivery re quired an average of 64 responses under a second-order FR 4 (VR16:S) schedule. Chronic daily administration of the partial mu agonist buprenorphine (0.014-0.79 mg/ '.l.l '.l.l :.., '.l.l c., .5 kg/ day) for 6 to 8 months selectively decreased self-ad C C ..c: .5 .5 0.. c 00 ministration of heroin and dilaudid while having little i: ..c: ..c: ;::J v 00 0.. 0.. 0.. c:; "T ..., ..., ..D o' effect on food self-administration. Again, these results 0 Q :::: ""2 ·=0.. Lf") ;;:;: ;;:;: z cf. ::J suggest that the decrease in opioid self-administration 412 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6

did not reflect a nonselective suppression of all behav doses of drug was available for self-administration. iors, but rather a selective decrease in the reinforcing ef Thus, a complete dose-effect curve could be examined fects of the opioid agonist. Interestingly, chronic treat during each daily session. Acute pretreatment with the ment with methadone (0.179-11.86 mg/kg/day) did opioid agonists buprenorphine, heroin, and nalbuphine not consistently decrease opioid self-administration in all produced a dose-dependent downward shift in both this study, although food self-administration was sup the alfentanil and cocaine dose-effect curves. Buprenor pressed. The relative ineffectiveness of methadone in phine displayed a potency difference in producing this study may have resulted in part from the treatment these effects, with doses as low as 0.003 mg/kg effec regimen in which methadone doses were gradually in tively suppressing the alfentanil dose-effect curve, creased over a period of 4 months. However, other whereas a dose of 3.2 mg/kg buprenorphine was re studies also have reported behavioral toxicity resulting quired to suppress the cocaine dose-effect curve. Thus, from administration of methadone doses that were suf low doses of buprenorphine selectively decreased alfen ficient to decrease opioid self-administration Gones and tanil self-administration. Both nalbuphine and heroin, Prada 1971; Harrigan and Downs 1981; see the follow in contrast, were equipotent and nonselective in de ing sections in this review). creasing alfentanil and cocaine self-administration. As In all the studies discussed, the effects of opioid ago in the study by Harrigan and Downs (1981), lower unit nists were evaluated on a single unit dose per injection doses of alfentanil were sometimes more sensitive than of the self-administered opioid. Other studies have ex higher unit doses of alfentanil to the effects of opioid amined the effects of opioid agonists on the self-admin agonist pretreatments. istration of a range of opioid unit doses. For example, For the most part, the opioid agonists that have been Harrigan and Downs (1981) trained rhesus monkeys to evaluated for their effects on opioid self-administration self-administer either heroin (0.25-64 µg/kg/inj.) or are selective for mu receptors, although compounds morphine (2-625 µg/kg/inj.) under a FR 1 schedule such as buprenorphine and nalbuphine also bind with during 15-minute sessions occurring every 4 hours (to high affinity to kappa- and delta-opioid receptors (e.g. tal of eight sessions per day). Days during which an Wood et al. 1981). The effect of more selective kappa opioid was available for self-administration alternated and delta-opioid receptor agonists on opioid self-ad with days during which either d-amphetamine or saline ministration has received little attention, although rela was available. Figure 9 shows that continuous infusion tively low doses of the kappa agonist U50,488 (Koob et of methadone (4-24 mg/kg/day for at least 3 weeks) al. 1986) and the delta agonist DPDPE (Negus et al. decreased the self-administration of low to intermedi 1992) did not alter heroin self-administration in rats. ate unit doses of heroin located on the ascending limb High doses of kappa and delta selective agonists do or peak of the heroin dose-effect curve; however, self suppress rates of responding maintained by other rein administration of high unit doses located on the de forcers such as food (e.g. Negus et al. 1993b ), so it is rea scending limb of the opioid dose-effect curve was not sonable to predict that they would also decrease re decreased. Similar effects were obtained in experiments sponding maintained by opioid agonists. Recently, two examining the effects of chronic buprenorphine (0.48- kappa agonists, spiradoline and U50,488, have been re 0.96 mg/kg/ day) on morphine self-administration. ported to produce dose-related (2.5 to 10 mg/kg) and Overall, chronic treatment with methadone or bu selective decreases in morphine self-administration by prenorphine produced a downward shift in the opioid rats (Glick et al. 1995). The effects of US0,488 on mor dose-effect curves. Unfortunately, the effects of metha phine-maintained responding were fully antagonized done and buprenorphine on d-amphetamine self-ad by the kappa antagonist norbinaltorphimine (10 mg/ ministration were not reported, so it is not clear kg) (Glick et al. 1995). A more thorough evaluation of whether these opioid agonists selectively decreased these compounds is warranted. opioid self-administration. However, the investigators 111 summary, opioid agonists, and particularly mu-se did note that doses of methadone that suppressed her lective opioid agonists, have generally been reported to oin self-administration produced "severe debilitation decrease opioid self-administration and to shift opioid and depression," suggesting that the effects of metha dose-effect curves downward. In some cases, the mag done on heroin self-administration may have been non nitude of these decreases in drug self-administration selective (Harrigan and Downs 1981). appears to depend on the unit dose of opioid-maintain In another study examining opioid agonist effects on ing behavior. Specifically, opioid agonist pretreatments of entire opioid dose-effect curves, Winger and coworkers ten are more effective in decreasing self-administration (1992) trained two groups of rhesus monkeys to re of lower unit doses located on the ascending limb or spond for either cocaine (1-100 µg/kg/inj.) or alfenta peak of the opioid dose effect-curve than in decreasing nil (0.03-3 µg/kg/inj.) under a FR 30 schedule. Each of self-administration of higher unit doses located on the the two daily sessions was composed of four cycles, and descending limb of the dose-effect curve. Again, these during each cycle either saline or one of several unit findings illustrate the importance of characterizing the NEUROPSYCHOPHARMACOLOCY 1YY6-VOL. 14, NO. 6 Evaluation of Treatments for Drug Abuse 413

agonist dose-effect curve as a basis for interpreting the teins (G proteins) that functionally couple the opioid effects of test drugs. During chronic treatment, toler receptor to the enzyme. These G proteins can be inacti ance may develop to opioid-induced decreases in opi vated by pertussis toxin, suggesting that pertussis oid self-administration. In some studies, the effects of toxin may be able to block the effects of opioid ago opioid agonists on opioid self-administration have been nists. Consistent with this hypothesis, intracranial ad shown to be selective; in other cases, opioid agonist-in ministration of pertussis toxin into the nucleus accum duced decreases in opioid self-administration are ac bens was recently reported to produce a small companied by decreases in the self-administration of rightward shift in the dose-effect curves for both her nonopioid reinforcers or by overt behavioral toxicity. oin and cocaine self-administration (Self et al. 1994). Interestingly, methadone, the drug most often used clin Although pertussis toxin certainly is not a candidate ically for the treatment of opioid dependence, has for the clinical treatment of opioid dependence, these proven to be a particularly toxic drug at doses that sup preliminary findings do suggest that drugs targeted at press opioid self-administration in animal models. The intracellular signal transduction mechanisms may be partial mu-agonist buprenorphine, in contrast, has con able to alter opioid self-administration. sistently been reported to produce a selective suppres At an intercellular level, it has been hypothesized sion of opioid self-administration at doses that produce that the mesolimbic dopamine system forms a final few other behavioral effects. common pathway for the reinforcing effects of many abused drugs, including opioids (Bozarth 1986; DiChi ara and Imperato 1988a). Several studies have tested this hypothesis by examining the effects of dopamine Effects of Nonopioids on Opioid antagonists on opioid self-administration Self-Administration (Glick and Cox 1975; Stretch 1977; Ettenberg et al. 1982; Van Ree Opioid agonists and antagonists presumably alter opi and Ramsey 1987; Gerber and Wise 1989; Higgins et al. oid self-administration by competing with the self-ad 1994; Winger 1994). In general, dopamine antagonists ministered opioid for opioid receptors. Compounds decrease opioid self-administration, but only at doses that do not act directly on opioid receptors, but that that also produce overt sedation and decrease behavior may nonetheless modulate opioid activity, also have maintained by nonopioid reinforcers. These findings ar been evaluated for their effects on opioid self-adminis gue against a prominent role for dopamine in opioid tration. For example, active immunization of rhesus self-administration and suggest that dopamine antago monkeys with a morphine/protein conjugate (mor nists will not be useful in the treatment of opioid abuse. phine-6-hemisuccinate-bovine serum albumin) induces Other compounds reported to alter opioid self-ad the production of antibodies that bind both morphine ministration include the cholinergic antagonist atropine and heroin and selectively shift the dose-effect curve (Glick and Cox 1975), the neuropeptide oxytocin (Ko for heroin, but not for cocaine, to the right (Bonese et al. vacs and Van Ree 1985), the hallucinogen ibogaine 1974). Moreover, the antibodies from immunized mon (Glick et al. 1991; Dworkin et al. 1995), the calcium keys can be isolated and infused into non-immunized channel blockers isradipine and nimodipine (Kuzmin et monkeys to produce similarly selective effects on her al. 1992), and the indirect serotonin agonist dexfenflu oin self-administration (Killian et al. 1978b ). Presum ramine (Higgins et al. 1993). However, these studies ably, these morphine antibodies function by binding used only a single unit dose of the self-administered morphine or heroin in plasma and preventing the drug opioid. In addition, most of these studies did not dem from reaching receptors involved in self-administration. onstrate selectivity, although ibogaine (Dworkin et al. In contrast to receptor antagonists, however, antibodies 1995) and dexfenfluramine (Wang et al. 1995) were display high selectivity for the antigen, and morphine found to produce a selective decrease in opioid self-ad antibodies would not be expected to alter self-adminis ministration relative to food self-administration. Fi tration of structurally distinct opioids such as metha nally, the effects of chronic treatment with these drugs done or alfentanil. are unknown, although Wang and coworkers (1995) Other compounds that do not affect binding of the found that dexfenfluramine produced only a transient self-administered opioid to the receptor may alter the decrease in heroin self-administration during chronic intracellular or intercellular consequences of opioid-re dexfenfluramine treatment. As a result, additional re ceptor activation that culminate in reinforcement of search is necessary to extend these provocative results opioid self-administration behavior. At the intracellu and further characterize the effects of these compounds lar level, for example, activation of mu-opioid recep on opioid self-administration. Interestingly, the alpha2 tors has been demonstrated to inhibit the enzyme ade adrenoreceptor agonist clonidine has not been evalu nylate cyclase and decrease intracellular levels of the ated in a preclinical model of opioid abuse, although second messenger cAMP. This inhibition of adenylate this drug attenuates some signs of opioid withdrawal cyclase is mediated by guanine nucleotide binding pro- and is one of the principal pharmacotherapies em- 414 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6

ployed in the treatment of opioid dependence (Charney tion to these general considerations, several procedural et al. 1986; Stine and Kosten, 1992; Gold 1993). issues also have an important effect on the interpreta tion of data obtained.

TOWARD IMPROVED PRECLINICAL MODELS FOR MEDICATION EVALUATION Selectivity of Medication Effects One of the challenges of preclinical medication evalua Preclinical studies of novel compounds are an im tions is to design a study that permits an unambiguous portant first step in identifying potentially effective assessment of the relative selectivity of medication ef pharmacotherapies for eventual clinical evaluation and fects on drug self-administration. Because a decrease in suggesting promising directions for medications devel drug self-administration may reflect sedation or toxicity opment. Animal models of drug self-administration rather than a decrease in drug reinforcing efficacy, it is simulate one basic behavioral endpoint of human drug important to evaluate treatment drug effects on re abuse that cac1 be affected by pharmacotherapies. The sponding maintained by an alternative reinforcer such clinical relevance of this model is suggested by the de as food. Moreover, food should be available under the gree of concordance between clinical and preclinical same conditions as drug self-administration and mea studies of drugs used for the treatment of opioid depen sures of operant response-contingent food self-adminis dence and polydrug abuse involving cocaine and opi tration are preferable to post-session food intake. Evalu oid dependence described throughout this review ation of treatment medication effects on drug and food (Mello 1991; Mello et al. 1993c). Howe\'er, many of the self-administration in the same subject is preferable to compounds examined in preclinical studies are not ap using different groups. proved for phase II clinical testing, and parallel clinical and preclinical studies have not been conducted. Ac Treatment Medication Dose Range cordingly, the cross-validation of findings from clinical and preclinical models is an ongoing process (see Fig Several doses of the treatment medication should be ure 1). Many of the preclinical studies of pharmacologi studied to establish the range over which it is effective cal antagonists considered in this review were origi and to ensure that behaviorally active doses are exam nally designed to investigate mechanisms of drug ined. An optimal design would evaluate the effects of action rather than to predict treatment medication effi several doses of the treatment drug on the self-adminis cacy. The degree of similarity of data obtained under tration of several unit doses of cocaine or heroin. How different experimental conditions has sometimes been ever, the amount of experimental time and resources re obscured by a great diversity of study designs and basic quired to study all these dose conditions make it procedures, as well as variations in criteria for interpre impractical for an initial evaluation of new treatment tation of medication efficacy. medications. A pragmatic alternative is to evaluate sev Review of this complex literature has led us to con eral doses of the treatment drug in combination with clude that several related methodological issues should one reinforcing dose of cocaine or heroin that is on the be considered in the design of future studies. An ideal ascending limb of the dose-effect curve. If an alterna preclinical evaluation of potential treatment medica tive reinforcer is also studied to ensure that apparent tions should include the determination of complete treatment effects do not reflect sedation or toxicity, this dose-effect curves for the target drug of abuse and the more spartan paradigm can yield useful information. evaluation of chronic administration of several doses of The effects of the most promising treatment medica the treatment medication on several unit doses of the tions on the complete cocaine or heroin dose-effect abused drug. It is important to recapture baseline levels curve can be evaluated subsequently. of drug self-administration between each successive ex posure to the treatment medication to ensure that intra Stability of Medication Effects venous catheters are patent and that cumulative effects of treatment drugs do not complicate the interpretation Because clinical treatment of substance abuse involves of findings. In addition, the profile of interactions be chronic administration of a pharmacotherapy, it is im tween the treatment medication and the abused drug portant to evaluate the stability of medication effects should be examined to identify possible toxic effects as over time in preclinical models of drug self-adminis well as reinforcing effects of the drug combination. Al tration. Medication-induced changes in drug self-ad though some agonist effects appear to be important for ministration may not be reliably detected in a single patient acceptance of a drug abuse treatment medica test session. Several days of observation are necessary tion (Mello et al. 1993c; Mendelson and Mello 1996), to determine whether treatment medication effects in clinical usefulness may be limited if a pharmacotherapy crease or decrease over time. There have been many re is highly reinforcing and likely to be abused. In addi- ports of changes in medication effects during chronic NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6 Evaluation of Treatments for Drug Abuse 415

administration on cocaine self-administration (Wool Criteria of Medication Effectiveness verton and Virus 1989; Kleven and Woolverton 1990b; Gold and Balster 1992; Negus and Picker 1996; Negus Careful attention to each of the issues described will de et al. 19956) and opioid self-administration (Jones and termine, in part, the appropriate criteria for treatment Prada 1977; Wang et al. 1995). Moreover, when several efficacy. We have discussed the ambiguities of interpre drug self-administration sessions are run each day, tation that may result when efficacy is defined as an in changes in drug-maintained responding do not always crease in drug-maintained responding similar to that occur during the first session after administration of observed during substitution of saline or a minimally the treatment drug (Negus et al. 1995a). In clinical reinforcing low dose of the abused drug. However, it is medicine, some pharmacotherapies such as antide important to recognize that this problem occurs pri pressants require chronic exposure before effective re marily when the treatment drug is evaluated on only duction of symptoms occurs. Medications for drug one occasion. When chronic administration of the treat abuse treatment may also require repeated administra ment drug is evaluated, an initial increase in drug tion over time to become fully effective. Alternatively, maintained responding may not be sustained. More the development of tolerance to medication effects on over, equating high levels of saline self-administration drug self-administration could compromise its clinical with drug medication efficacy ignores the fact that sub usefulness. stitution of saline or very low doses of the abused drug It is also important to determine the degree to which do not invariably result in an increase in injection-main medication effects persist after treatment has been dis tained responding. For example, after an animal has continued. In clinical practice, an ideal therapeutic out been exposed to saline extinction procedures on multi come would include enduring beneficial effects of the ple occasions, the substitution of saline for cocaine or treatment medication so that the patient could eventu heroin may be followed by a rapid decrease in saline ally become drug-free. In the preclinical model, the time self-administration. This qualification is relevant to the required for drug self-administration to return to base interpretation of findings from all acute treatment stud line levels is one measure of the persistence of treat ies where increases in drug self-administration were in ment medication effects. However, it is also possible terpreted as evidence for antagonism of the drug's rein that chronic exposure to the treatment medication forcing properties (see Tables 1-6). might result in an increased sensitivity to the reinforc It is intuitively obvious that a decrease in drug self ing effects of the abused drug (Kleven and Woolverton administration is the goal of clinical treatment, but in 19906). Redetermination of drug dose-effect curves at preclinical evaluations, medications may have different the conclusion of treatment can discriminate between effects depending on the position of the unit dose on the these alternatives. drug self-administration dose-effect curve. For exam ple, antagonism of the reinforcing effects of cocaine or heroin decreases the self-administration of some drug doses (i.e., doses on the ascending limb of the dose-ef Drug Dose-Effect Curves fect curve) while increasing self-administration of other The importance of evaluating medication effects on doses (i.e., doses on the descending limb; see Figure 4, dose-effect curves for the abused drug has been em left panel). Although considerable experimental effort phasized throughout this review. The interpretation of has been devoted to analysis of the antagonist effects of findings depends, in part, on the position of the unit drugs, the clinical implications of the antagonist-related dose of cocaine or heroin on the drug dose-effect curve. increases in drug self-administration are somewhat If the unit dose of cocaine or heroin is on the ascending problematic. An ideal treatment medication would de limb of the dose-effect curve (see Figure 4), a decrease crease drug self-administration across a wide range of in drug self-administration during treatment is consis doses on both the ascending and descending limbs of tent with a decrease in the drug's reinforcing properties. the dose-effect curve (see Figure 4, center panel). This However, if the unit dose is on the descending limb of preclinical outcome corresponds to a decrease in hu the dose-effect curve, a decrease in drug self-adminis man drug abuse. Moreover, because human drug.abus tration may reflect an interaction between the direct ers seldom have accurate information about the purity rate-reducing effects of the abused drug and the treat of illicit drugs, it is important to develop pharmacother ment medication. When the entire drug dose-effect apies that reduce drug self-administration across a curve is examined, a competitive antagonist would shift broad range of drug doses. the ascending limb of the curve to the right. But, the ef Awareness of the distinction between assessment of fect of an antagonist on the descending limb of the treatment medication effects on drug self-administra dose-effect curve would depend on whether or not the tion and analysis of the pharmacological mechanisms of rate-reducing effects of high doses of cocaine or heroin drug interactions may help to reconcile some apparent were antagonized (see Woods et al. 1987 for review). inconsistencies in the preclinical literature. Often, these 416 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOCY 1996-VOL. 14, NO. 6

may be sequential steps in the evaluation process; for Bergman J, Kamien JB, Spealman RD (1990): Antagonism of example, identification of a potentially effective treat cocaine self-administration by selective dopamine D1 ment medication may precede analysis of its pharmaco and D2 antagonists. Behav Pharmacol 1:355-363 logical mechanisms of action. Alternatively, analysis of Bertalmio AJ, Woods JH (1987): Differentiation between mu pharmacological mechanisms of drug interactions may and kappa receptor-mediated effects in opioid drug dis crimination: Apparent PA analysis. J Pharmacol Exp identify compounds with specific receptor antagonist or 2 Ther 243:591-597 agonist activities that are promising candidates for pre clinical evaluation. However, it is important not to limit Bertalmio AJ, Woods JH (1989): Reinforcing effect of alfenta nil is mediated by mu opioid receptors: Apparent PA2 the search for new treatment medications to a narrow analysis. J Pharmacol Exp Ther 251:455-460 range of predictable agonist-antagonist interactions Biden JR Jr (1989): Pharmacotherapy: A strategy for the within pharmacologically defined drug classes. Effec 1990's. A Staff Report. Washington, DC, U.S. Govern tive treatment of polydrug abuse involving drugs from ment Printing Office, p 38 several pharmacological classes probably will require Bilsky EJ, Calderon SN, Wang T, Bernstein RN, Davis P, less categorical approaches to medications develop Hruby J, McNuitt RW, Rothman RB, Rice K, Porreca F ment. The availability of animal models of polydrug (1995): SNC 80, A selective, nonpeptidic and systemi self-administration should facilitate the evaluation of cally active, opioid delta agonist. J Pharmacol Exp Ther the next generation of treatment medications (Mello et 273:359-366 al. 1995, 1996). Although many issues still remain to be Blumberg H, Dayton HB (1974): Naloxone, naltrexone, and resolved, recent advances in and collaboration among related noroxymorphones. In Braude MC, Harris LS, researchers in behavioral science, pharmacology, neuro May EL, Smith JP, Villarreal JE (eds), Narcotic Antago nists: Advances in Biochemical Psychopharmacology. biology, and medicinal chemistry provide a basis for the New York, Raven Press, pp 33-43 discovery of more effective medications for substance Bonese KF, Wainer BH, Fitch FW, Rothberg RM, Schuster CR abuse treatment and relapse prevention. (1974): Changes in heroin self-administration by a rhesus monkey after morphine immunization. Nature 252:708-710 Bourland JA, French ED (1995): Effects of remoxipride, an REFERENCES atypical antipsychotic, on cocaine self-administration in the rat using fixed- and progressive-ratio schedules of Adams EH, Durell J (1984): Cocaine: A growing public reinforcement. Drug and Alcohol Dependence 40:111- health problem. In Grabowski J (ed), Cocaine: Pharma 114 cology, Effects, and Treatment of Abuse, NIDA Research Bozarth MA (1986): Neural basis of psychomotor stimulant Monograph 50. Washington DC, U.S. Government and opioid reward: Evidence suggesting the involve Printing Office, pp 9-14 ment of a common dopaminergic system. Behav Brain Aigner TG, Balster RL (1978): Choice behavior in rhesus Res 22:107-116 monkeys: Cocaine versus food. Science 201:534-535 Brady JV, Lukas SE (eds) (1984): Testing Drugs for Physical Balster RL (1991): Drug abuse potential evaluation in ani Dependence Potential and Abuse Liability, NIDA mals. Br J Addict 86:1549-1558 Research Monograph 52. Washington, DC, U.S. Govern ment Printing Office Balster RL, Mansbach RS, Gold L, Harris LS (1992): Preclini cal methods for the development of pharmacotherapies Brady JV, Griffiths RR, Hienz RD, Ator NA, Lukas SE, Lamb for cocaine abuse. In Harris LS (ed), Problems of Drug RJ (1987): Assessing drugs for abuse liability and depen Dependence 1991, NIDA Research Monograph 119. dence potential in laboratory primates. In Bozarth MA Washington, DC, U.S. Government Printing Office, pp (ed), Methods of Assessing the Reinforcing Properties of 161-164 Abused Drugs. New York, Springer Verlag, pp 45-86 Banks CD (1979): Overdose of buprenorphine: Case report. Byrd LD (1981): Quantitation in behavioral pharmacology. In NZ Med J 89:255-256 Thompson T, Dews PB, McKim WA (eds), Advances in Batki SL, Manfredi LB, Jacob IP, Jones RT (1993): Fluoxetine Behavioral Pharmacology. New York, Academic Press, for cocaine dependence in methadone maintenance: pp 75-90 Quantitative plasma and urine cocaine/benzoylecgo Caine SB, Koob GF (1993): Modulation of cocaine self nine concentrations. J Clin Psychopharmacol 13:243-250 administration in the rat through dopamine recep Bergman J, Rosenzweig-Lipson S (1992): Cocaine-antagonist tors. Science 260:1814-1816

effects of limited-efficacy D1 agonists. In Harris LS (ed), Caine SB, Koob GF (1994): Effects of dopamine Di and D2 Problems of Drug Dependence 1991, NIDA Research antagonists on cocaine self-administration under differ '.V!onograph 119. Washington, D.C., U.S. Government ent schedules of reinforcement in the rat. J Pharmacol Printing Office, pp 185-189 Exp Ther 270:209-218 Bergman J, Madras BK, Johnson SE, Spealman RD (1989): Caine SB, Koob CF (1995): Pretreatment with the dopamine Effects of cocaine and related drugs in nonhuman pri agonist 7-OH-DPAT shifts the cocaine self-administra mates. III. Self-administration by squirrel monkeys. J tion dose-effect function to the left under different Pharmacol Exp Ther 251:150-155 schedules in the rat. Behav Pharmacol 6:333-347 NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6 Evaluation of Treatments for Drug Abuse 417

Cappendijk LT, Dzolijic MR (1993): Inhibitory effects of de la Garza R, Johanson CE (1982): Effects of haloperidol and ibogaine on cocaine self-administration in rats. Eur J physostigmine on self-administration of local anesthet Pharmacol 241:261-265 ics. Pharmacol Biochem Behav 17:1295-1299 Carroll ME, Lac ST (1992): Effects of buprenorphine on self Deneau GA, Yanagita T, Seevers MH (1969): Self-administra administration of cocaine and a non-drug reinforcer in tion of psychoactive substances by the monkey. Psy rats. Psychopharmacology 106:439-446 chopharmacologia 61 :30-48 Carroll ME, Lac ST, Walker MJ, Kragh R, Newman T (1986): Devine DP, Wise RA (1994): Self-administration of morphine, Effects of naltrexone on intravenous cocaine self-admin DAMGO and DPDPE into the ventral tegmental area of istration in rats during food satiation and deprivation. J rats. J Neurosci 14:1978-1984 Pharmacol Exp Ther 238:1-7 de Vries TJ, Babovic-Vuksanovic D, Elmer G, Shippenberg TS Carroll ME, Lac ST, Asencio M, Kragh R (1990a): Fluoxetine (1995): Lack of involvement of o-opioid receptors in reduces intravenous cocaine self-administration in rats. mediating the rewarding effects of cocaine. Psychophar Pharmacol Biochem Behav 35:237-244 macology 120:442-448 Carroll ME, Lac ST, Asencio M, Halikas JA, Kragh R (1990b): DeWitt H, Wise RA (1977): Blockade of cocaine reinforce Effects of carbamazepine on self-administration of intra ment in rats with the dopamine receptor blocker pimozide, venouslv delivered cocaine in rats. Pharmacol Biochem but not with the noradrenergic blockers phentolamine Behav 37:551-556 or phenoxybenzamine. Canad J Psychol 31:195-203 Carroll ME, Carmona GN, May SA, Buzalsky S, Larson C Dews PB (1958): Studies on behavior. IV. Stimulant actions of (1992): Buprenorphine's effects on self-administration of methamphetamine. J Pharmacol Exp Ther 122:137-147 smoked cocaine base and orally delivered phencyclid Dews PB (1981): History and present status of rate-depen ine, ethanol and saccarin in rhesus monkevs. J Pharma- dency investigations. In Thompson T, Dews PB, McKim col Exp Ther 261:26-33 ' WA (eds), Advances in Behavioral Pharmacology. New Centers for Disease Control and Prevention (February, 1993): York, Academic Press, pp 111-118 Facts about Women and HIV/ AIDS. Atlanta, GA, CDC DiChiara G, Imperato A (1988a): Drugs abused by humans Chaisson RE, Bacchetti P, Osmond D, Brodie B, Sande MA. preferentially increase synaptic dopamine concentra Moss AR (1989): Cocaine use and HIV infection in intra tions in the mesolimbic system of freely moving rats. venous drug users in San Francisco. JAMA 261:561-565 Proc Natl Acad Sci US A. 85:5274-5278 Charney DS, Heninger GR, Kleber HD (1986): The combined DiChiara G, Imperato A (1988b): Opposite effects of mu and use of clonidine and naltrexone as a rapid, safe and kappa opiate agonists on dopamine release in the effective treatment for abrupt withdrawal from metha nucleus accumbens and in the dorsal caudate of freely done. Am J Psychiatry 143:831-837 moving rats. J Pharmacol Exp Ther 244:1067-1080 Condelli WS, Fairbank JA, Dennis ML, Rachal JV (1991): Dole VP (1988): Implications of methadone maintenance for Cocaine use by clients in methadone programs: Signifi theories of narcotic addiction. JAMA 260:3025-3029 cance, scope, and behavioral interventions. J Subst Dole VP, Nyswander M (1965): A medical treatment for Abuse Treat 8:203-212 diacetylmorphine (heroin) addiction: A clinical trial Cornish JW, Maany I, Fudala PJ, Neal S, Poole SA, Volpicelli with methadone hydrochloride. JAMA 193:646-650 P, O'Brien CP (1995): Carbamazepine treatment for Donahoe RM, Falek A (1988): Neuroimmunomodulation bv cocaine dependence. Drug Alcohol Depend 38:221-227 opiates and other drugs of abuse: Relationship to HIV Corrigall WA, Coen KM (1991a): Cocaine self-administration infection and AIDS. In Bridge TP, Mirsky FA, Goodwin FK (eds), Psychological, Neuropsychiatric and Sub is increased by both D1 and D2 dopamine antagonists. Pharmacol Biochem Behav 39:799-802 stance Abuse Aspects of AIDS. New York, Raven Press, pp 145-157 Corrigall WA, Coen KM (1991b): Opiate antagonists reduce cocaine but not nicotine self-administration. Psvcho- Downs DA, Woods JH (1974): Codeine- and cocaine-rein pharmacology 104:167-170 ' forced responding in rhesus monkeys: Effects of dose on response rates under a fixed-ratio schedule. J Pharmacol Crawford CA, McDougall SA, Bolanos CA, Hall S, Berger SP Exp Ther 191:179-188 (1995): The effects of the kappa agonist U-50,488 on cocaine-induced conditioned and unconditioned behav Dunteman GH, Condelli WS, Fairbank JA (1992): Predicting iors and Fos immuoreactivity. Psychopharmacology cocaine use among methadone patients: Analysis of 120:392-399 findings from a national study. Hosp Community Psy chiatry 43:608-611 Cushman P Jr. (1988): Cocaine use in a population of drug abusers on methadone. Hosp Community Psychiatry Dworkin SI, Gleeson S, Meloni D, Koves TR, Martin TJ 39:1205-1207 (1995): Effects of ibogaine on responding maintained by food, cocaine and heroin reinforcement in rats. Psycho Dackis CA, Gold MS (1985): New concepts in cocaine addic pharmacology 117:257-261 tion: The dopamine depletion hypothesis. Neurosci Biobehav Rev 9:469-477 Epidemiologic Trends in Drug Abuse (1995): Volume I: High lights and Executive Summary NIH Publication 95- Abuse DAWN (1993): Preliminary Estimates from the Drug 3990. National Institute on Drug Abuse, NIH, PHS, Warning Network. Substance Abuse and Mental Health USDHHS Services Administration Office of Applied Studies, Advance Report Number 2. US Dept of Health and Ettenberg A, Pettit HO, Bloom FE, Koob GF (1982): Heroin Human Services and cocaine intravenous self-administration in rats: 418 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6

Mediation by separate neural systems. Psychopharma Glowa JR, Wojnicki FHE, Matecka D, Bacher JD, Mansbach cology (Berlin) 78:204-209 RS, Balster RL, Rice K (1995a): Effects of dopamine reuptake inhibitors on food- and cocaine-maintained Federal Strategy for Drug Abuse and Drug Traffic Preven responding: I. dependence on unit dose of cocaine. Exp tion (1973). Washington, DC, Strategy Council on Drug 3:219-231 Abuse Clin Psychopharmacol Matecka D, Rice K, Rothman RB Ferster CB, Skinner BF (1957): Schedules of Reinforcement. Glowa JR, Wojnicki FHE, of dopamine reuptake inhibitors on New York, Appleton-Century-Crofts (1995b ): Effects food- and cocaine-maintained responding: II. Compari Fischman MW, Foltin RW, Nestadt G, Pearlson GD (1990): sons with other drugs and repeated administrations. Effects of desipramine maintenance on cocaine self Exp Clin Psychopharmacol 3:232-239 administration by humans. J Pharmacol Exp Ther 253: 760-770 Gold LH, Balster RL (1992): Effects of buspirone and gepirone on i.v. cocaine self-administration in rhesus Foltin RW, Fischman MW (1992): The cardiovascular and monkeys. Psychopharmacology (Berl) 108:289-294 subjective effects of intravenous cocaine and morphine combinations in humans. J Pharmacol Exp Ther 261: Gold MS (1993): Opiate addiction and the locus coeruleus. 623-632 The clinical utility of clonidine, naltrexone, methadone and buprenorphine. Psychiatr Clin North Am 16:61-73 Foltin RW, Fischman MW (1995): Interaction of buprenor phine with cocaine-morphine combinations. Exp Clin Goldberg SR, Woods JH, Schuster CR (1971): Nalorphine Psychopharmacol 3:261-269 induced changes in morphine self-administration in rhesus monkeys. Pharmacol Exp Ther 176:464--471 Gastfriend DR, Mendelson JH, Mello NK, Teoh SK, Reif S (1993a): Buprenorphine pharmacotherapy for concur Goldberg SR, Morse WH, Goldberg DM (1976): Behavior rent heroin and cocaine dependence. Am J Addict 2:269- maintained under a second order schedule by intramus 278 cular injection of morphine or cocaine in rhesus mon keys. J Pharmacol Exp Ther 199:278-286 Gastfriend DR, Wapler M, Teoh SK, Reif S, Mendelson JH, Mello NK (1993b): Effects of buprenorphine on needle Grabowski J, Rhoades H, Elk R, Schmitz J, Davis C, Creson sharing, drug use and drug craving in men with com D, Kirby K (1995): Fluoxetine is ineffective for treatment bined heroin and cocaine dependence. In Harris LS (ed), of cocaine dependence or concurrent opiate and cocaine Problems of Drug Dependence 1992, NIDA Research dependence: Two placebo-controlled, double-blind tri Monograph 132. Washington, DC, U.S. Government als. J Clin Psychopharmacol 15:163-174 Printing Office, p 312 Griffiths RR, Balster RL (1979): Opioids: Similarity between Gastfriend DR, Mendelson JH, Mello NK, Teoh SK, Reif S evaluations of subjective effects and animal self-admin (1995): Long-term effects of buprenorphine for treat istration results. Clin Pharmacol Ther 25:611-617 ment of combined opiate and cocaine dependence. In Griffiths RR, Brady J, Bradford L (1979a): Predicting the Harris LS (ed), Problems of Drug Dependence 1994, abuse liability of drugs with animal drug self-adminis NIDA Research Monograph 153. Washington, DC, U.S. tration procedure: Psychomotor stimulants and halluci Government Printing Office, p 9 nogens. In Thompson T, Dews PB (eds), Advances in Gawin FH (1991): Cocaine addiction: Psychology and neuro Behavioral Pharmacology. New York, Academic Press, physiology. Science 251: 1580-1586 pp 163-208 Gawin FH, Ellinwood EH (1988): Cocaine and other stimu Griffiths RR, Bradford LD, Brady JV (1979b): Progressive lants: Actions, abuse and treatment. N Engl J Med 318: ratio and fixed-ratio schedules of cocaine-maintained 1173-1182 responding in baboons. Psychopharmacology 65:125-136 Ga win FH, Kleber HD (1984): Cocaine abuse treatment, open Griffiths RR, Wurster RM, Brady JV (1976): Discrete-trial pilot trial with desipramine and lithium carbonate. Arch choice procedure: Effects of naloxone and methadone Gen Psychiatry 41:903-909 on choice between food and heroin. Pharmacol Rev 27: 357-365 Gawin FH, Allen D, Humblestone B (1989): Outpatient treat ment of "crack" cocaine smoking with flupenthixol Griffiths RR, Bigelow G, Henningfield J (1980): Similarities in decanoate. Arch Gen Psychiatry 46:322-325 animal and human drug taking behavior. In Mello NK (ed), Advances in Substance Abuse, Behavioral and Bio Gerber GJ, Wise RA (1989): Pharmacological regulation of logical Research. Greenwich, CT, JAI Press, pp 1-90 intravenous cocaine and heroin self-administration in rats: A variable dose paradigm. Pharmacol Biochem Griffiths RR, Wurster RM, Brady JV (1981): Choice between Behav 32:527-531 food and heroin: effects of morphine, naloxone and sec obarbital. J Exp Anal Behav 35:335-351 Glick SD, Cox RD (1975): Dopaminergic and cholinergic influences on morphine self-administration in rats. Res Halikas J, Nurgent S, Crosby R, Carlson G (1993): Survey of Commun Chem Pathol Pharmacol 12:17-24 pharmacotherapies used in the treatment of cocaine abuse. J Addictive Dis 12:129-139 Glick SD, Rossman K, Steindorf S, Maisonneuve IM, Carlson JN (1991): Effects and after effects of ibogaine on mor Harrigan SE, Downs DA (1978): Continuous intravenous phine self-administration in rats. Eur J Pharmacol naltrexone effects on morphine self-administration in 195:341-345 rhesus monkeys. J Pharmacol Exp Ther 204:481--486 Glick SD, Maisonneuve JM, Raucci J, Archer S (1995): Kappa Harrigan SE, Downs DA (1981): Pharmacological evaluation opioid inhibition of morphine and cocaine self-adminis of narcotic antagonist delivery systems in rhesus mon tration in rats. Brain Res 681:147-152 keys. In Willette RE, Barnett G (eds), Narcotic Antago- NEUROPSYCHOPHARMACOLOCY 1996-VOL. 14, NO. 6 Evaluation of Treatments for Drug Abuse 419

nists: Naltrexone Pharmacochemistry and Sustained drug reinforcers: Cocaine and methylphenidate in the Release Preparations, NIDA Research Monograph 28. rhesus monkey. J Pharmacol Exp Ther 193:676-688 Washington, DC, U.S. Government Printing Office, pp Johanson CE, Fischman MW (1989): The pharmacology of 77-92 cocaine related to its abuse. Pharmacol Rev 41:3-52 Heidbreder CA, Goldberg SR, Shippenberg TS (1993): The Johnson RE, Jaffe JH, Fudala PJ (1992): A controlled trial of kappa-opioid receptor agonist U-69593 attenuates buprenorphine treatment for opioid dependence. JAMA cocaine-induced behavioral sensitization in the rat. 267:2750-2755 Brain Res 616:335-338 Jones BE, Prada JA (1975): Drug seeking behavior during Heidbreder CA, Babovic-Vuksanovic D, Shoaib M, Shippen methadone maintenance. Psychopharmacologia (Berl) berg T (1995): Development of behavioral sensitization 41:7-10 to cocaine: Influence of kappa opioid receptor agonists. J Pharmacol Exp Ther 275:150-163 Jones BE, Prada JA (1977): Effects of methadone and mor phine maintenance on drug-seeking behavior in the Herling S (1981): Naltrexone effects on food and codeine dog. Psychopharmacology 54:109-112 maintained responding in rhesus monkeys. Eur J Phar macol 73:41-49 Judson BA, Goldstein A (1979): Levo-alpha-acetyl-methadol (LAAM) in the treatment of heroin addicts. I. Dosage Herling S, Woods JH (1980): Chlorpromazine effects on schedule for induction and stabilization. Drug Alcohol cocaine-reinforced responding in rhesus monkeys: Depend 4:461-466 Reciprocal modification of rate-altering effects of drugs. J Pharmacol Exp Ther 214:354-361 Julius D, Renault P (eds) (1976): Narcotic Antagonists: Nal trexone Progress Report, NIDA Research Monograph 9. Herling S, Downs DA, Woods JH (1979): Cocaine, d-amphet Washington, DC, U.S. Government Printing Office amine, and pentobarbital effects on responding main tained by food or cocaine in rhesus monkeys. Psych Katz JL, Witkin JM (1992): Selective effects of the Di dopa opharmacology 64:261-269 mine receptor agonist, SKF 38393, on behavior main tained by cocaine injection. Psychopharmacology Higgins GA, Wang Y, Sellers EM (1993): Preliminary findings 109:241-244 with the indirect 5-HT agonist dexfenfluramine on her oin discrimination and self-administration in rats. Phar Kelleher RT (1976): Characteristics of behavior controlled by macol Biochem Behav 45:963--966 scheduled injections of drugs. In Kelleher RT, Goldberg SR, Krasnegor NA (eds), Control of Drug Taking Behav Higgins GA, Joharchi N, Wang Y, Corrigall WA, Sellers EM ior by Schedules of Reinforcement. Baltimore, Williams (1994): The CCKA receptor antagonist devazepide does & Wilkins, pp 307-323 not modify opioid self-administration or drug discrimi nation: Comparison with the dopamine antagonist halo Kelleher RT, Morse WH (1968): Determinants of the specificity peridol. Brain Res 640:246-254 of behavioral effects of drugs. Ergeb Physiologie 60:1-56 Houde RW (1979): Analgesic effectiveness of the narcotic Killian AK, Bonese K, Schuster CR (1978a): The effects of agonist-antagonists. Br J Clin Pharmacol 7 (suppl 3): naloxone on behavior maintained by cocaine and heroin 297-308 injections in the rhesus monkey. Drug Alcohol Depend 3:243-251 Howell LL, Byrd LD (1991): Characterization of the effects of cocaine and GBR 12909, a dopamine uptake inhibitor, Killian A, Bonese K, Rothberg RM, Wainer BH, Schuster CR on behavior in the squirrel monkey. J Pharmacol Exp (1978b): Effects of passive immunization against mor Ther 258:178-185 phine on heroin self-administration. Pharmacol Bio chem Behav 9:347-352 Hubner CB, Moreton JE (1991): Effects of selective Di and D2 dopamine antagonists on cocaine self-administration in Kleven MS, Woolverton WL (1990a): Effects of bromocriptine the rat. Psychopharmacology 105:151-156 and desipramine on behavior maintained bv cocaine or food presentation in rhesus monkeys. Psy~hopharma Jaffe J_H, Martin WR (1990): Opioid analgesics and antago cology (Berl) 101 :208-13 nists. In Gilman AG, Rall TW, Nies AS, Taylor P (eds), The Pharmacological Basis of Therapeutics. New York, Kleven MS, Woolverton WL (1990b): Effects of continuous Pergamon Press, pp 485-521 infusions of SCH 23390 on cocaine- or food-maintained behavior in rhesus monkeys. Behav Pharmacol l:365-374 Jaffe JH, Senay EC (1971): .'\1ethadone and 1-methadvl ace tate: Use· in management of narcotic addicts. JAMA Kleven MS, Woolverton WL (1993): Effects of three mono 216:1303-1305 amine uptake inhibitors on behavior maintained by cocaine or food presentation in rhesus monkeys. Drug Jaffe JH, Senay EC, Schuster CR, Renault PR, Smith B, Alcohol Depend 31:149-158 DiMenza S (1972): Methadvl acetate vs methadone A double-blind study in heroi~ users. JAMA 222:437-442 Koob GF, Pettit HO, Ettenberg A, Bloom FE (1984): Effects of opiate antagonists and their quaternary derivatives on Jasinski DR, Pevnick JS, Griffith JD (1978): Human pharma heroin self-administration in the rat. J Pharmacol Exp cology and abuse potential of the analgesic buprenor Ther 229:481-486 phine. Arch Gen Psvchiatrv 35:601-616 Koob GF, Vaccarino FJ, Amalric M, Bloom FE (1986): Neuro Jenck F, Gratton A, Wise RA (1987): Opioid receptor sub chemical substrates for opioid reinforcement. In Brown types associated with ventral tegmental facilitation of RM, Clouet DH, Friedman DP (eds), Opioid Receptor lateral hypothalamic brain stimulation reward. Brain Subtypes and Brain Function, 1\JIDA Research Mono Res 423:34-38 graph 71. Washington, DC, U.S. Government Printing Johanson CE, Schuster CR (1975): A choice procedure for Office, pp 147-164 420 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6

Kosten TR, Rounsaville BJ, Gawin FH, Kleber HD (1986): (eds), Buprenorphine: Combatting Drug Abuse with a Cocaine abuse among opioid addicts: Demographic and Unique Opioid. New York, Wiley-Liss, pp 151-163 diagnostic factors in treatment. Am J Drug Alcohol Abuse 12:1-16 Lewis JW, Rance MJ, Sanger DJ (1983): The pharmacology and abuse potential of buprenorphine: A new antago Kosten TR, Schumann B, Wright D, Carney MK, Gawin FH nist analgesic. In Mello NK (ed), Advances in Substance (1987): A preliminary study of desipramine in the treat Abuse: Behavioral and Biological Research. Greenwich, ment of cocaine abuse in methadone maintenance CT, JAI Press, pp 103-154 patients. J Clin Psychiatry 48:442-444 Longoni R, Spina L, Mulas A, Carboni E, Garau L, Melchiorri Kosten TR, Kleber HD, Morgan C (1989a): Treatment of 2 P, DiChiara G (1991): [D-Ala ]Deltorphin II: D1-depen cocaine abuse with buprenorphine. Biol Psychiatry dent stereotypes and stimulation of dopamine release in 26:170-172 nucleus accumbens. Neuroscience 11: 1565-1576 Kosten TR, Kleber HD, Morgan C (1989b): Role of opioid Lord JAH, Waterfield AA, Hughes J, Kosterlitz HW (1977): antagonists in treating intravenous cocaine abuse. Life Endogenous opioid peptides: Multiple agonists and Sci 44:887-892 receptors. Nature 267:495-499 Kosten TR, Shottenfeld R, Ziedonis D, Falcioni J (1993): Lukas SE, Moreton JD, Khazan l\i (1981): Effects of levo Buprenorphine versus methadone maintenance for opi alpha-acetylmethadol (LAAM) on morphine self-admini oid dependence. J Nerv Ment Dis 18:358-364 stration in the rat. Psychopharmacology 73:12-16 Kovacs CL, Van Ree JM (1985): Behaviorally active oxytocin Lukas SE, Mello NK, Drieze JM, Mendelson JH (1995): fragments simultaneously attenuate heroin self-admin Buprenorphine-induced alterations of cocaine's rein istration and tolerance in rats. Life Sci 37:1895-1900 forcing properties in rhesus monkey: A dose-response analysis. Drug Alcohol Depend 40:87-98 Kozel NJ, Adams EH (1986): Epidemiology of drug abuse: An overview. Science 34:970-974 Madras BK, Fahey MA, Bergman J, Canfield DR, Spealman RD (1989): Effects of cocaine and related drugs in non Kranzler HR, Bauer DH, Klinghoffer V (1995): Carbam human primates. I. [3H] cocaine binding sites in cau azepine treatment of cocaine dependence: A placebo date-putamen. J Pharmacol Exp Ther 251:131-141 controlled trial. Drug Alcohol Depend 38:203-211 Maisonneuve IM, Archer S, Glick SD (1994): U50,488 aK opi Kreek MJ (1987): Multiple drug abuse patterns and medical oid receptor agonist, attenuates cocaine-induced increases consequences. In Meltzer H (ed), Psychopharmacology, in extracellular dopamine in the nucleus accumbens of the Third Generation of Progress. New York, Raven rats. Neurosci Lett 181:57-60 Press, pp 1597-1604 Mansbach RS, Balster RL (1993): Effects of mazindol on Kuhar MJ, Ritz MC, Sharkey J (1988): Cocaine receptors on behavior maintained or occasioned by cocaine. Drug dopamine transporters mediate cocaine reinforced Alcohol Depend 31:183-191 behavior. In Clouet D, Asghar K, Brown R (eds), Mecha nisms of Cocaine Abuse and Toxicity, NIDA Research Mansour A, Khachaturian H, Lewis ME, Aki! H, Watson SJ Monograph 88. Washington, DC, U.S. Government (1987): Autoradiographic differentiation of mu, delta Printing Office, pp 14-22 and kappa opioid receptors in rat forebrain and mid brain. J Neurosci 7:2445-2464 Kuhar MJ, Ritz MC, Boja JW (1991): The dopamine hypothe Martin WR, Jasinski DR (1977): Assessment of the abuse sis of the reinforcing properties of cocaine. Trends Neu rosci 14:299-302 potential of narcotic analgesics in animals. In Martin WR (ed), Drug Addiction I: Handbook of Experimental Kuzmin A, Zvartau E, Gessa CL, Martellotta MC, Fratta W Pharmacology. Berlin, Springer-Verlag, pp 159-196 (1992): Calcium antagonists isradipine and nimodipine Martin WR, Eades CG, Thompson JA, Huppler RE, Gilbert suppress cocaine and morphine intravenous self-admin PE (1976): The effects of morphine- and nalorphine-like istration in drug-naive mice. Pharmacol Biochem Behav drugs in the non-dependent and morphine-dependent 41:497-500 dog. J Pharmacol Exp Ther 197:517-532 Lamb RJ, Preston KL, Schindler CW, Meisch RA, Davis F, Martin WR, Jasinski DR, Haertzen CA, Kay DC, Jones BE, Katz JL, Henningfield JE, Goldberg SR (1991): The rein Mansky PA, Carpenter RW (1973a): Methadone-A re forcing and subjective effects of morphine in post evaluation. Arch Gen Psychiatry 28:286-295 addicts: A dose-response study. J Pharmacol Exp Ther 259: 1165-1173 Martin WR, Jasinski DR, Mansky PA (1973b): Naltrexone, an antagonist for the treatm~nt of heroin dependence Lane JD, Pickering CL, Hooper ML, Fagan K, Tyers MB, effects in man. Arch Gen Psychiatry 28:784-791 Emmett-Oglesby MW (1992): Failure of ondansetron to block the discriminative or reinforcing stimulus effects Mello NK (1979): Animal models of alcoholism: Progress and of cocaine in the rat. Drug Alcohol Depend 30:151-162 prospects. In Davidson R (ed), Modification of Patho logical Behavior, Experimental Analysis of Etiology and Lewis JW (1974): Ring C-Bridged derivatives of thebaine and Therapy. New York, Gardner Press, pp 273-333 oripavine. In Braude MC, Harris LS, May EL, Smith JP, Villarreal JE (eds), Narcotic Antagonists. Advances in Mello NK (1991): Pre-clinical evaluation of the effects of Biochemical Psychopharmacology. New York, Raven buprenorphine, naltrexone and desipramine on cocaine Press, pp 123-136 self-administration. In Harris LS (ed), Problems of Drug Dependence 1990, NIDA Research Monograph No. 105, Lewis JW (1995): Clinical pharmacology of buprenorphine in DHHS Pub!. No. (ADM)91-1753. Washington, DC, U.S. relation to its use as an analgesic. In Cowan A, Lewis JW Government Printing Office, pp 189-195 NEUROPSYCHOPHARMACOLOCY 19%-VOL. 1-l, J\iO. 6 Evaluation of Treatments for Drug Abuse 421

Mello NK (1992): Behavioral strategies for the evaluation of (1993d): Acute effects of cocaine on anterior pituitary new pharmacotherapies for drug abuse treatment. In hormones in male and female rhesus monkeys. J Phar Harris LS (ed), Problems of Drug Dependence 1991, macol Exp Ther 266:804-811 NIDA Research Monograph No. 119, DHHS (ADM) 92- Mello NK, Negus SS, Lamas X, Lukas SE, Mendelson JH 1888. Washington, DC, U.S. Government Printing Office, (1996): Buprenorphine's pp 150-154 effects on "speedball" combina tions of cocaine and heroin. In Harris LS (ed), Problems of Mello NK, Mendelson JH (1980): Buprenorphine suppresses Drug Dependence 1995, NIDA Research Monograph heroin use by heroin addicts. Science 27:657-659 No. 162, NIH Pub. No. 96-4116. Rockville, MD, NIDA, Mello NK, Mendelson JH (1992): Primate studies of the DHHS, p. 170 behavioral pharmacology of buprenorphine. In Blaine Mello NK, Negus SS, Lukas SE, Mendelson JH, Sholar JW, JD, Snyder M (eds), Buprenorphine: An Alternative Drieze J (1995): A primate model of polydrug abuse: Treatment for Opioid Dependence, NIDA Research Cocaine and heroin combinations. JPharmacol Exp Ther Monograph No. 121, DHHS (ADM)92-1912. Washing 274: 1325-1337 ton, DC, U.S. Government Printing Office, pp 61-100 Mendelson JH, Mello NK (1994): Cocaine and other com Mello NK, Mendelson JH (1995): Buprenorphine treatment monly abused drugs. In Isselbacher KH, Braunwald E, of cocaine and heroin abuse. In Cowan A, Lewis JW Wilson JD, Martin JB, Fauci AS, Kasper DL (eds), Harri (eds), Buprenorphine: Combatting Drug Abuse with a son's Principles of Internal Medicine, ed 13, New York, Unique Opioid. New York, Wiley-Liss, pp 241-287 McGraw-Hill, pp 2429-2433 Mello NK, Mendelson JH, Bree MP (1981): Naltrexone effects Mendelson JH, Mello NK (1996): Management of cocaine on morphine and food self-administration in morphine abuse and dependence. N Eng J Med 334:965-972 dependent rhesus monkeys. J Pharmacol Exp Ther Meyer RE, Mirin 218:550-557 SM (1979): The Heroin Stimulus. New York, , Plenum Press Mello NK, Mendelson JH, Kuehnle JC (1982): Buprenorphine Montoya ID, Levin FR, Fudala PJ, Gorelick DA (1995): Dou effects on human heroin self-administration. Pharma J ble-blind col Exp Ther 223:30-39 comparison of carbamazepine and placebo for treatment of cocaine dependence. Drug Alcohol Depend Mello NK, Bree MP, Mendelson JH (1983): Comparison of 28:213-219 buprenorphine and methadone effects on opiate self Musto DF (1973): The American Disease: Origins of Narcotic administration in primates. J Pharmacol Exp Ther 225:378-386 Control. New Haven, Yale University Press Nader MA, Mach RH (in press): Self-administration of the Mello NK, Mendelson JH, Bree MP, Lukas SE (1989): dopamine D agonist 7-OH-DPAT in rhesus Buprenorphine suppresses cocaine self-administration 3 monkeys is modified by previous cocaine exposure. by rhesus monkeys. Science 245:859-862 Psychopharma cology Mello NK, Lukas SE, Bree MP, Mendelson JH (1990a): National Commission on Acquired Immune Deficiency Syn Desipramine effects on cocaine self-administration by drome (1991): Fifth Interim Report to the President rhesus monkeys. Drug Alcohol Depend 26:103-116 and the Congress: The Twin Epidemics of Substance Use and Mello NK, Mendelson JH, Bree MP, Lukas SE (1990b): HIV. Washington, DC, National Commission on AIDS. Buprenorphine and naltrexone effects on cocaine self National Research Council Committee on Clinical Evalua administration by rhesus monkey. J Pharmacol Exp Ther tion of 254:926-939 Narcotic Antagonists (1978): Clinical evaluation of naltrexone treatment of opioid-dependent individu Mello NK, Lukas SE, Kamien JB, Mendelson JH, Drieze J, als. Arch Gen Psychiatry 35:335-340 Cone EJ (1992): The effects of chronic buprenorphine Negus SS, Mello NK, lamas X, Mendelson JH (1996): Acute treatment on cocaine and food self-administration by and chronic effects of flupenthixol on the discriminative rhesus monkeys. J Pharmacol Exp Ther 260:1185-1193 , stimulus and reinforcing effects of cocaine in rhesus Mello NK, Kamien JB, Lukas SE, Mendelson JH, Drieze JM, monkeys. J Pharmacol Exp Ther (in press) Sholar JW (1993a): Effects of intermittent buprenorphine Negus SS, Picker MJ (1996): BW37U86: A non-peptidic, sys administration on cocaine self-administration by rhesus temically-active delta opioid agonist. monkeys. J Pharmacol Exp Ther 264:530-541 , CNS Drug Reviews 2:68-90 Mello NK, Lukas SE, Mendelson JH, Drieze J (1993b): Nal Negus SS, Weinger MB, Henriksen SJ, Koob GF (1992): Effect trexone-buprenorphine interactions: Effects on cocaine of mu, delta and kappa opioid agonists on heroin-main self-administration. Neuropsychopharmacology 9:211- tained 224 responding in the rat. In Harris LS (ed), Problems of Drug Dependence 1991, NIDA Research Monograph Mello NK, Mendelson JH, Lukas SE, Gastfriend D, Teoh SK, No. 119, DHHS Pub. No. (ADM) 92-1888. Rockville, Holman BL (1993c): Buprenorphine treatment of opiate MD, U.S. Government Printing Office, p 410 and cocaine abuse: Clinical and preclinical studies. Har Negus SS, Burke TF, Medzihradsky vard Rev Psychiatry 1:168-183 F, Woods JH (1993a): Effects of opioid agonists selective for mu, kappa and Mello NK, Kamien JB, Lukas SE, Drieze J, Mendelson JH delta opioid receptors on schedule-controlled respond (1993d): The effects of nalbuphine and butorphanol ing in rhesus monkeys: Antagonism by quadazocine. J treatment on cocaine and food self-administration by Pharmacol Exp Ther 267:896-903 rhesus monkeys. Neuropsychopharmacology 8:45-55 · Negus SS, Henriksen SJ, Mattox A, Pasternak GW, Porto Mello NK, Sarnyai Z, Mendelson JH, Drieze JM, Kelly M ghese PS, Takemori AE, Weinger MB, Koob GF (1993b): 422 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6

Effect of opioid agonists selective for mu, delta and Sanchez-Ramos JR, Schuster CR (1977): Second-order sched kappa opioid receptors on the reinforcing effects of her ules of intravenous drug self-administration in rhesus oin in rats. J Pharmacol Exp Ther 265:1245-1252 monkeys. Pharmacol Biochem Behav 7:443-450 Negus SS, Butelman ER, Chang KJ, DeCosta BR, Winger G, Sannerud CA, Griffiths RR (1988): Amantadine: Evaluation Woods JH (1994): Behavioral effects of the systemically of reinforcing properties and effect on cocaine self active delta opioid agonist BW373U86 in rhesus mon administration in baboons. Drug Alcohol Depend keys. J Pharmacol Exp Ther 270:1025-1034 21:195-202 Negus SS, Mello NK, Lukas SE, Mendelson JH (1995a): Diur Schecter A (1980): The role of narcotic antagonists in the nal patterns of cocaine and heroin self-administration in rehabilitation of opiate addicts: A review of naltrexone. rhesus monkeys responding under a schedule of multi Am J Drug Alcohol Abuse 7: 1-18 ple daily sessions. Behav Pharmacol 6:763-765. Schottenfeld RS, Pakes J, Ziedonis D, Kosten TR (1993): Negus SS, Mello NK, Portoghese PS, Lukas SE, Mendelson Buprenorphine: Dose-related effects on cocaine and opi JH (1995b ): Role of delta opioid receptors in the reinforc oid use in cocaine-abusing opioid-dependent humans. ing and discriminative stimulus effects of cocaine in Biol Psychiatry 3:66-74 rhesus monkeys. J Pharmacol Exp Ther 273:1245-1256 Schuster CR, Johanson CE (1974): The use of animal models Pillai R, Nair BS, Watson RR (1991): AIDS, drugs of abuse for the study of drug abuse. In Gibbons RJ et al. (eds), and the immune system: A complex immunotoxicologi Research Advances in Alcohol and Drug Problems. New cal network. Arch Toxicol 65:609-617 York, Wiley, pp 1-31 Pollack MH, Rosenbaum JF (1991): Fluoxetine treatment of Schwartz JC, Giros B, Martres MP, Sokoloff P (1992): The cocaine abuse in heroin addicts. J Clin Psychiatry 52:31-33 dopamine receptor family: Molecular biology and phar Porrino LJ, Ritz MC Goodman NL, Sharpe LG, Kuhar MJ, macology. Semin Neurosci 4:99-108 Goldberg SR (1989): Differential effects of the pharmaco Segal DL, Schuster CR (1995): Buprenorphine: What interests logical manipulation of serotonin systems of cocaine the National Institute on Drug Abuse? In Cowan A, and amphetamine self-administration in rats. Life Sci Lewis JW (eds), Buprenorphine: Combatting Drug 45:1529-1535 Abuse with a Unique Opioid. New York, Wiley-Liss, pp Post RM, Weiss SRB, Aigner TG (1993): Carbamazepine in 309-320 the treament of cocaine abuse. In Korenman S, Barchas J Seiden LS, Balster RL (1985): Behavioral Pharmacology: The (eds), Biological Basis of Substance Abuse. New York, Current Status. New York, Alan R. Liss Oxford University Press, pp 443-462 Self OW, Stein L (1992): The D1 agonists SKF 82958 and SKF Rawson RA, Obert JL, McCann MJ, Castro FG, Ling W 77434 are self-administered by rats. Brain Res 582:349-352 (1991): Cocaine abuse treatment: A review of current strategies. J Subst Abuse 3:457-491 Self OW, Nestler EJ, Stein L (1994): Role of G-proteins in her oin and cocaine reinforcement. J Neurosci 14:6239-6247 Richardson NR, Roberts DCS (1991): Fluoxetine pretreat ment reduces breaking points on a progressive ratio Sharpe LG, Jaffe JH, Katz JL (1992): Carbamazepine pro schedule reinforced by intravenous cocaine self-admin duces nonspecific effects on cocaine self-administration istration in the rat. Life Sci 49:833-840 in rats. Life Sci 51:PL13-18 Richardson NR, Piercey MF, Svensson K, Collins RJ, Myers Shippenberg TS, Bals-Kubik R, Herz A (1987): Motivational JE, Roberts DCS (1993): Antagonism of cocaine self properties of opioids: Evidence that an activation of administration by the preferential dopamine autorecep &-receptors mediates reinforcement properties. Brain tor antagonist (+)-AJ 76. Brain Res 619:15-21 Res 436:234-239 Richardson NR, Smith AM, Roberts DCS (1994): A single Skjoldager P, Winger G, Woods JH (1991): Analysis of fixed injection of either flupenthixol decanoate or haloperidol ratio behavior maintained by drug reinforcers. J Exp decanoate produces long-term changes in cocaine self Anal Behav 56:331-343 administration in rats. Drug Alcohol Depend 36:23-25 Skjoldager P, Winger G, Woods JH (1993): Effects of GBR Ritz MC, Lamb RJ, Goldberg SR, Kuhar MJ (1987): Cocaine 12909 and cocaine on cocaine-maintained behavior in receptors on dopamine transporters are related to self rhesus monkeys. Drug Alcohol Depend 33:31-39 administration of cocaine. Science 237:1219-1223 Smith A, Piercey M, Roberts DCS (1995): Effects of (-)-DS 121 Ritz MC, Lamb RJ, Goldberg SR, Kuhar MJ (1988): Cocaine and (+)-UH 232 on cocaine self-administration in rats. self-administration appears to be mediated by dopa Psychopharmacology 120:93-98 mine uptake inhibition. Prog Neuropsychopharmacol Spanagel R, Herz A, Shippenberg TS (1990): The effects of Biol Psychiatryl2:233-239 opioid peptides on dopamine release in the nucleus Roberts DCS, Vickers G (1984): Atypical neuroleptics accumbens: An in vivo microdialysis study. J Neuro increase self-administration of cocaine: An evaluation of chem 55: 1734 a behavioural screen for antipsychotic activity. Psycho Spealman RD (1985): Environmental factors determining the pharmacology 82: 135-139 control of behavior by drugs. In Seiden LS, Balster RL Roberts DCS, Loh EA, Vickers G (1989): Self-administration (eds), Behavioral Pharmacology: The Current Status. of cocaine on a progressive ratio schedule in rats: Dose New York, Alan R. Liss, pp 23-38 response relationship and effect of haloperidol pretreat Spealman RD (1992): Use of cocaine-discrimination tech ment. Psychopharmacology 97:535-538 niques for preclinical evaluation of candidate therapeu Rosen MI, Kosten TR (1991): Buprenorphine: Beyond metha tics for cocaine dependence. In Harris LS (ed), Problems done? Hosp Community Psychiatry 42:347-349 of Drug Dependence 1991, NIDA Research Monograph NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6 Evaluation of Treatments for Drug Abuse 423

No. 119, DHHS Publ. No. (ADM) 92-1888. Washington, Weeks JR, Collins RJ (1976): Changes in morphine self DC, U.S. Government Printing Office, pp 175-179 administration in rats induced by prostaglandin El and Spealman RD, Goldberg SR (1978): Drug self-administration naloxone. Prostaglandins 12:11-19 by laboratory animals: Control by schedules of rein Weiss RD (1988): Relapse to cocaine abuse after initiating forcement. Annu Rev Pharmacol Toxicol 18:313-339 desipramine treatment. JAMA 260:2545-2546 Spealman RD, Bergman J (1992): Modulation of the discrimi- Wilson MC, Schuster CR (1972): The effects of chlorprom native-stimulus effects of cocaine by mu and kappa opi azine on psychomotor stimulant self-administration in oids. J Pharmacol Exp Ther 261:607-615 the rhesus monkey. Psychopharmacologia 26:115-126 Spealman RD, Bergman J (1994): Opioid modulation of the Wilson MC, Schuster CR (1973): The effects of stimulants and discriminative stimulus effects of cocaine: Comparison depressants on cocaine self-administration behavior in of µ, K and o agonists in squirrel monkeys discriminat the rhesus monkey. Psychopharmacologia 31:291-304 ing low doses of cocaine. Behav Pharmacol 5:21-31 Wilson MC, Schuster CR (1976): Mazindol self-administra- Spealman RD, Bergman J, Madras BK, Kamien JB, Melia KF tion in the rhesus monkey. Pharmacol Biochem Behav (1992): Role of D1 and D2 dopamine receptors in the 4:207-210 behavioral effects of cocaine. Neurochem Int 20:147S- 152S Winger G (1988): Pharmacological modifications of cocaine and opioid self-administration. In Clouet DH, Asghar K, Stine SM, Kosten TR (1992): Use of drug combinations in Brown RM (eds), Mechanisms of Cocaine Abuse and treatment of opioid withdrawal. J Clin Psychopharma Toxicity, NIDA Research Monograph No. 88. Washing col 12:203-209 ton, DC, U.S. Government Printing Office, pp 125-136 Stretch R (1977): Discrete-trial control of cocaine self-injec Winger G (1994): Dopamine antagonist effects on behavior tion behaviour in squirrel monkeys: Effects of mor maintained by cocaine and alfentanil in rhesus mon phine, naloxone and chlorpromazine. Can J Physiol keys. Behav Pharmacol 5:141-152 Pharmacol 55:778-790 Winger G, Woods JH (1985): Comparison of fixed-ratio and Tang M, Falk JL (1990): Schedule-induced oral self-adminis progressive-ratio schedules on maintenance of stimu tration of cocaine and ethanol solutions: Lack of effect of lant drug-reinforced responding. Drug Alcohol Depend chronic desipramine. Drug Alcohol Depend 25:21-25 15:123-130 Thompson T, Schuster CR (1964): Morphine self-administra Winger G, Skjoldager P, Woods JH (1992): Effects of tion, food reinforced, and avoidance behaviors in rhesus buprenorphine and other opioid agonists and antago monkeys. Psychopharmacologia 5:87-94 nists on alfentanil- and cocaine-reinforced responding Thompson T, Unna KR (1977): Predicting Dependence Lia in rhesus monkeys. J Pharmacol Exp Ther 261:311-317 bility of Stimulant and Depressant Drugs. Baltimore, Wise RA, Newton P, Leeb K, Burnette B, Pocock D, Justice J Jr University Park Press (1995): Fluctuations in nucleus accumbens dopamine Tutton CS, Crayton JW (1993): Current pharmacotherapies concentration during intravenous cocaine self-adminis for cocaine abuse: A review. JAddict Dis 12:109-127 tration in rats. Psychopharmacology 120:10-20 Van Ree JM, Ramsey N (1987): The dopamine hypothesis of Witkin JM (1994): Pharmacotherapy of cocaine abuse: Pre opiate reward challenged. Eur J Pharmacol 134:239-243 clinical development. Neurosci Biobehav Rev 18:121- Vanover KE, Woolverton WL (1994): Behavioral effects of the 142 dopamine autoreceptor agonist PD 128483 alone and in Wood PL, Charleson SE, Lane E, Hudgin RL (1981): Multiple combination with cocaine. J Pharmacol Exp Ther 270: opioid receptors: Differential binding ofµ, K and o ago 1049-1056 nists. Neuropharmacolog 20:1215-1220 Vanover KE, Piercey MF, Woolverton WL (1993): Evaluation Woods JH, Schuster CR (1968): Reinforcement properties of of the reinforcing and discriminative stimulus effects of morphine, cocaine and SPA as a function of unit dose. cocaine in combination with (+) -AJ76 or clozapine. J Internat J Addict 3:231-237 Pharmacol Exp Ther 266:780-789 Woods JH, Schuster CR (1970): Regulation of drug self Verebey K, Volavka J, Mule SJ, Resnick RB (1976): Naltrex administration. In Harris RT, Mcisaac WM, Schuster CR one: Disposition, metabolism and effects after acute and (eds), Advances in Mental Science II. Drug Dependence. chronic dosing. Clin Pharmacol Ther 20:315-328 Austin, TX, Texas Research Institute of Mental Sciences, Wang Y, Joharchi N, Fletcher PJ, Sellers EM, Higgins GA pp 158-169 (1995): Further studies to examine the nature of dexfen Woods JH, Schuster CR (1971): Opiates as reinforcing stim fluramine-induced suppression of heroin self-adminis uli. In Thompson T, Pickens R (eds), Stimulus Properties tration. Psychopharmacology 120:134-141 of Drugs. New York, Apple-Century-Crofts, pp 163-175 Weed MR, Vanover KE, Woolverton WL (1993): Reinforcing Woods JH, Downs DA, Carney J (1975): Behavioral functions effect of the D1 dopamine agonist SKF 81297 in rhesus of narcotic antagonists: Response drug contingencies. monkeys. Psychopharmacology 113:51-52 Fed Proc 34:1777-1784 Weeks JR (1962): Experimental morphine addiction: Method Woods JH, Winger G, France CF (1987): Reinforcing and dis for automatic intravenous injection in unrestrained rats. criminative stimulus effects of cocaine: Analysis of Science 138:143-144 pharmacological mechanisms. In Fisher S, Raskin A, Weeks JR, Collins RJ (1964): Factors affecting voluntary mor Uhlenhuth EH (eds), Cocaine: Clinical and Behavioral phine intake in self-maintained addicted rats. Psycho Aspects. New York, Oxford University Press, pp 21-65 pharmacologia 6:267-279 Woolverton WL (1986): Effects of a D1 and D2 dopamine 424 N.K. Mello and S.S. Negus NEUROPSYCHOPHARMACOLOGY 1996-VOL. 14, NO. 6

antagonist in the self-administration of cocaine and pir D2 dopamine antagonist on behavior maintained by ibedil bv rhesus monkeys. Pharmacol Biochem Behav cocaine or food. Pharmacol Biochem Behav 32:691-697 24:351-535 . Woolverton WL, Goldberg LI, Ginos JZ (1984): Intravenous Woolverton WL, Balster RL (1981): Effects of antipsychotic self-administration of dopamine receptor agonists by compounds in rhesus monkeys given a choice between rhesus monkeys. JPharmacol Exp Ther 230:678-683 cocaine and food. Drug Alcohol Depend 8:69-78 Wurster RM, Griffiths RR, Findley JD, Brady JG (1977): Woolverton WL, Kleven MS (1988): Evidence for cocaine Reduction of heroin self-administration in baboons by dependence in monkeys following a prolonged period manipulation of behavioral and pharmacological condi of exposure. Psychopharmacology (Berl) 94:288-291 tions. Pharmacol Biochem Behav 7:519-528 Woolverton WL, Johnson KM (1992): Neurobiology of Yanagita T, Deneau GA, Seevers MH (1965): A technique for cocaine abuse. Trends Pharmacol Sci 13:193-200 chronic intravenous programmed or self-administration Woolverton WL, Kleven MS (1992): Assessment of new med of drugs to monkeys. Proc Int Union of Physiol Sci ications for stimulant abuse treatment. In Harris LS (ed), 87:453-457 Problems of Drug Dependence 1991, NIDA Research Young AM, Swain HH, Woods JH (1981): Comparison of opi Monograph No. 119, DHHS (ADM) 92-1888. Washing oid agonists in maintaining responding and in sup ton, DC, U.S. Government Printing Office, pp 155-159 pressing morphine withdrawal in rhesus monkeys. Woolverton WL, Virus RM (1989): The effects of a Di and a Psychopharmacology (Berl) 74:329-335