J. K. Rowlett1, S. S. Negus2, T. S. Shippenberg3,
N. K. Mello2, S. L. Walsh4, and R. D. Spealman1

1Harvard Medical School, NERPRC, Southborough, MA and 2ADARC/McLean Hospital, Belmont, MA; 3NIH/NIDA, DIR, Baltimore, MD; and 4Behav. Pharmacol. Res. Unit, Johns Hopkins University School of Medicine, Baltimore, MD


Many intravenous polydrug abusers inject cocaine in combination with heroin either by injecting the two drugs serially or by combining the drugs in solution and taking them simultaneously (commonly know as a "speedball"). The increasing use of speedballs has led to their widespread inclusion in recent epidemiological studies of drug abuse in the United States. It has been reported that speedball abusers exhibit a more severe psychopathology compared to other cocaine abusers, are more likely to fail in drug abuse treatment, and are at increased risk of contracting HIV infection. Despite the prevalence of speedball abuse, the interactions between cocaine and opioid drugs are not well understood at either the clinical or preclinical level. This symposium examined current research on the pharmacological interactions between cocaine and abused opioids in order to provide a synthesis of current knowledge and to suggest potential directions for future research.

Neurobiology of cocaine and opioid abuse

The rewarding effects of cocaine have been attributed to an increase in the functional activity of mesolimbic dopamine (DA) neurons. Evidence that endogenous as well as exogenous opioid agonists modulate the activity of this DA system has also been presented. This talk reviewed neurochemical data regarding the effects of cocaine administered alone or in combination with opioids upon mesolimbic DA neurotransmission as well as the role endogenous opioid peptide systems play in regulating the activity of mesolimbic neurons. The influences of cocaine upon opioid peptide gene expression and the dynamics of endogenous opioid peptide systems also were reviewed, along with the relevance of these findings for the development of therapeutic agents to treat speedball abuse.

Discriminative stimulus effects of cocaine/opioid combinations

Drug discrimination procedures have been used extensively to characterize the pharmacological mechanisms of action mediating the abuse-related effects of cocaine and mu opioids, as well as between those drugs. In most of these studies, subjects were trained to discriminate either cocaine or a mu agonist from vehicle and the effects of cocaine alone, mu agonists alone, and cocaine/mu agonist combinations were examined. An important goal of these studies has been to evaluate the hypothesis that cocaine and mu agonists potentiate each other's discriminative stimulus (DS) effects. The results have been variable across studies, and in many instances, across subjects within a study. In subjects trained to discriminate cocaine from vehicle, mu agonists sometimes produce high levels of cocaine-appropriate responding and/or enhance the DS effects of cocaine. Similarly, in subjects trained to discriminate mu agonists, cocaine sometimes generalizes to the training stimulus and enhances the DS effects of the mu agonist. In many subjects, however, cocaine and mu agonists do not cross generalize and either do not alter or attenuate the other's DS effects. Another approach for examining the DS effects of speedballs is to train subjects to discriminate a mixture of cocaine and a mu agonist from vehicle. We recently trained rhesus monkeys to discriminate a 10:1 mixture of cocaine and heroin (0.4 mg/kg cocaine + 0.04 mg/kg heroin). Increasing doses of the 10:1 mixture produced a dose-dependent generalization to the training stimulus, and both cocaine and heroin alone generalized completely to the training stimulus. In addition, many cocaine-like drugs (CFT, amphetamine, bupropion) and heroin-like drugs (alfentanil, fentanyl, morphine) produced high levels of speedball-appropriate responding. The DS effects of the cocaine/heroin mixture could be antagonized by combined treatment with the DA antagonist flupenthixol and the opioid antagonist quadazocine, but either antagonist alone was less effective. These findings suggest that the DS effects of the cocaine/heroin mixture includes aspects of both cocaine and heroin and that the mixture did not produce DS effects that were qualitatively different from either cocaine or heroin alone.

Reinforcing effects of cocaine/opioid combinations

Because both cocaine and mu opioid agonists function as reinforcers, it is possible that the reinforcing effects of cocaine and opioid combinations are enhanced compared to either cocaine or mu agonists alone. To examine this possibility, rhesus monkeys were trained to respond under a progressive-ratio (PR) schedule of intravenous drug injection. Under this schedule, the response requirement for obtaining an injection of drug is increased until responding stops and a measure of reinforcing efficacy can be defined as the maximum response requirement a subject will complete to obtain an injection. Using this procedure, cocaine dose-effect curves for injections/session typically are monophasic, i.e., responding increases with dose until an asymptote or a peak is reached. Heroin dose-effect curves, by contrast, are biphasic functions, i.e., increase to a peak and then decrease. When cocaine was combined with heroin, low doses of cocaine and heroin that did not maintain behavior when tested alone did so when tested in combination. Furthermore, combination with heroin resulted in a leftward shift in the cocaine dose-effect curve, indicating that heroin increased the potency of cocaine as a reinforcer. However, maximum injections/session for cocaine combined with heroin were not different from cocaine alone, suggesting that the relative reinforcing efficacy of combinations of cocaine and heroin were not greater than that of cocaine alone. The enhanced potency of the reinforcing effects of cocaine by combination with heroin may be a contributing factor to the abuse of speedballs.

Medications development for the treatment of combined cocaine-opioid abuse

The combined abuse of pharmacologically dissimilar drugs complicates the search for effective treatment medications. We have found that buprenorphine, an opioid mixed agonist-antagonist, significantly reduced both cocaine and opioid self-administration by rhesus monkeys and human polydrug abusers. We recently studied the effects of buprenorphine on self-administration of cocaine/heroin combinations using a model of speedball abuse developed in our laboratory. In this procedure, drugs and food (1 gm banana pellets) were available in four daily sessions on a second-order FR4 (VR16:S) schedule of reinforcement. Each heroin/cocaine combination was available for 10 consecutive days, and monkeys were treated daily with either saline or buprenorphine. Daily treatment with 0.237 mg/kg/day buprenorphine shifted the speedball dose-effect curves for heroin+0.001 mg/kg/inj cocaine downward and to the right. Buprenorphine also decreased self-administration maintained by speedball combinations of heroin+0.01 mg/kg/inj cocaine and heroin+0.1 mg/kg/inj cocaine. However, buprenorphine was less effective in decreasing self-administration maintained by heroin in combination with 0.1 mg/kg/inj of cocaine than with 0.01 mg/kg/inj of cocaine. Thus, buprenorphine's effectiveness in decreasing speedball self-administration decreased as the dose of cocaine increased. These findings suggest the usefulness of the speedball model for evaluation of the effectiveness of medications for the treatment of speedball polydrug abuse. Our findings in rhesus monkeys are concordant with clinical trials, which found that buprenorphine decreased both cocaine and heroin abuse in polydrug abusers.

Human laboratory studies on cocaine/opioid combinations

Recent studies by Foltin, Fischman and colleagues and by Walsh and colleagues have evaluated the acute subjective and physiological effects of cocaine in combination with opioid mu agonists. The subjective effects of various dose combinations of mu agonists and cocaine were greater than those produced by either drug alone on measures such as "drug liking" and "high". The qualitative profile of subjective effects produced by the opioid/cocaine combination was roughly equivalent to the sum of effects produced by the two drugs singly, rather than producing novel or unique subjective effects. The cardiovascular effects of cocaine, including tachycardia and pressor action, were enhanced when cocaine was administered in the presence of the mu agonist. Other studies have compared the effects of cocaine in opioid-dependent subjects during chronic maintenance on mu opioid agonists. In studies of methodone-maintained patients, the subjective response to cocaine (e.g., "magnitude of drug effect", "good effect") was significantly greater compared to control subjects. Similarly, some studies in buprenorphine-maintained patients reported that the subjective response to cocaine was either enhanced or not appreciably altered compared to a control condition. Enhanced cardiovascular responses to cocaine were also observed under these chronic opioid dosing conditions. These and other findings suggest that administration of mu agonists generally enhances the effects of cocaine in humans, particularly the positive subjective responses to cocaine. This enhancement of subjective effects has been observed following pretreatment with opioids under both acute and chronic dosing procedures. Whether or not enhancement of physiological responses is observed appears to depend upon the specific opioid agonist chosen for testing as well as the range of test doses. The observed enhancement of subjective responses produced by speedball combinations may account for the high prevalence of speedball abuse as well as the persistent abuse of cocaine in patients enrolled in opioid maintenance therapy programs. Summary and Future Directions This symposium has provided a synthesis of recent research aimed at understanding the interactions between cocaine and abused opioids. Dr. Shippenberg described neurobiological evidence that both exogenous and endogenous opioids can modulate the actions of cocaine in the mesolimbic DA system. There is also evidence for a reciprocal influence of cocaine on the endogenous opioid peptide system. Understanding the interplay between these two systems continues to be a major challenge facing research in this area. The escalation of dual cocaine-opioid abuse has prompted renewed interest in the interactions between these drugs in relevant animal models of addiction. As discussed by Drs. Negus, Rowlett, and Mello, an important, though still elusive, goal of this research is to characterize the complex effects that combinations of cocaine and opioids can have in behavioral experiments. Growing evidence indicates that prominent individual differences exist in the DS effects of cocaine-opioid combinations, yet virtually nothing is known about their neurobiological correlates. Along similar lines, studies involving i.v. drug self-administration in monkeys suggest that low doses of cocaine and heroin, which fail to maintain self-administration when tested singly, do so if combined under a progressive-ratio schedule of drug injection. However, this appears not to be the case under other conditions, where self-administration of cocaine may be reduced by the addition of high doses of heroin. Human laboratory studies reviewed by Dr. Walsh have generally shown that opioids enhance the positive subjective effects of cocaine, although the findings do not fit easily within any simple model of drug interaction. Given the prevalence of cocaine use among untreated opioid abusers and methadone-maintained patients, there is a critical need to identify pharmacological mechanisms underlying speedball abuse in order to develop effective therapeutic strategies. By providing a state-of-the-field review of multidisciplinary research, this symposium has helped lay the groundwork for achieving this goal.

Future Opioids

Opium Images
Opium Timeline
Meet The Family
Opioid Receptors
Just For Chemists
The Pleasure and the Pain
Confessions of an English Opium-Eater