Cocaine abuse continues to be a major public health problem, and the adverse medical consequences of cocaine abuse and dependence contribute to the social and economic costs (Novick et al., 1997; Mendelson and Mello, 2004). As yet, no effective pharmacotherapies have been developed (Mendelson and Mello, 1996; SAMHSA, 2000), but advances in understanding the neurobiological bases of cocaine abuse have led to several pharmacological approaches to treatment. There is considerable evidence that the abuse-related effects of cocaine are mediated by increases in extracellular levels of dopamine in the terminal fields of the mesolimbic dopamine system (Kuhar et al., 1991; Spealman et al., 1992; Carroll et al., 1999; Platt et al., 2002). Accordingly, compounds that modulate dopaminergic activity have been extensively evaluated in preclinical studies as potential treatments for cocaine abuse and dependence (Mello and Negus, 1996; Carroll et al., 1999; Howell and Wilcox, 2001; Platt et al., 2002). Unfortunately, most of these compounds have not been approved for clinical use by the Food and Drug Administration (FDA).

An alternative approach to the medication-based treatment of cocaine abuse has been to study compounds that can modulate dopaminergic activity indirectly by acting on other receptor systems (Carroll et al., 1999; Bergman et al., 2000; Mello and Negus, 2000). Kappa opioid agonists are one candidate medication for cocaine abuse treatment (Archer et al., 1996; Mello and Negus, 2000). Accumulating evidence from in vivo and in vitro studies shows that pharmacological modulation of kappa opioid receptors can affect the activity of dopaminergic neurons that are involved in mediating the neurochemical and behavioral effects of cocaine (Mello and Negus, 2000). For example, high levels of kappa opioid receptors and their endogenous agonist, dynorphin, have been found in the nucleus accumbens (Meng et al., 1993; Mansour et al., 1994; Svingos et al., 1999). Elevated levels of dynorphin and overexpression of kappa opioid receptors have been also found after chronic cocaine treatment (Spangler et al., 1996; Shippenberg et al., 2001; Yuferov et al., 2001). Administration of kappa agonists has been shown to decrease basal and cocaine-stimulated dopamine levels in microdialysis studies (Di Chiara and Imperato, 1988; Devine et al., 1993; Maisonneuve et al., 1994).

Behavioral studies indicate that kappa opioid agonists also attenuate cocaine-induced increases in locomotor activity as well as the development of sensitization to cocaine’s effects on locomotor activity and stereotypy in rats (Crawford et al., 1995; Schenk et al., 2001; Shippenberg et al., 2001). In addition, kappa opioid agonists have been found to decrease cocaine-seeking behavior in rodent models of i.v. self-administration and relapse, as well as place preference (Glick et al., 1995; Kuzmin et al., 1997; Schenk et al., 2001; Shippenberg et al., 2001). Kappa opioid agonists also produced dose-dependent and sustained decreases in cocaine self-administration in rhesus monkeys (Negus et al., 1997; Mello and Negus, 1998b, 2000). Importantly, side effects (e.g., emesis and sedation) sometimes observed at the beginning of chronic treatment, rapidly subsided within two or three days (Negus et al., 1997; Mello and Negus, 1998b, 2000).

Archer and co-workers first postulated that kappa and mu opioid combinations might be useful for treatment of cocaine abuse, because both kappa agonists and mu antagonists reduce dopamine release from the nucleus accumbens (Archer et al., 1996). In support of this hypothesis, compounds with mixed activity at kappa and mu opioid receptors (e.g., ethylketocyclazocine [EKC], Mr2033) decreased cocaine self-administration by rhesus monkeys more consistently, and with fewer side effects than selective kappa opioid agonists (Negus et al., 1997; Mello and Negus, 1998b, 2000). Both EKC and Mr2033 appear to act as low efficacy agonists at mu opioid receptors and as high efficacy agonists at kappa opioid receptors. Subsequent studies in rhesus monkeys confirmed that mixed mu/kappa opioids were more effective in reducing cocaine self-administration than kappa selective opioids, and produced minimal side effects (Bowen et al., 2003). Taken together, these studies suggested that mixed mu/kappa opioids may provide a useful approach to treatment of cocaine abuse.

Translation of the preclinical findings into clinical evaluations of kappa and mixed mu/kappa opioid agonists for cocaine abuse treatment has been hindered by the limited number of kappa opioids approved for clinical use (Walsh et al., 2001a, 2001b). In humans, kappa selective agonists can produce fatigue, sedation, confusion, anxiety and at higher doses, visual and auditory distortions and depersonalization (Pfeiffer et al., 1986; Reece et al., 1994). However, drug use history influences the subjective effects produced by kappa opioids. For example, in men with no history of drug abuse, enadoline, a selective kappa agonist (5–25 μg), produced a number of adverse effects (Reece et al., 1994), whereas in opioid abusers, enadoline in doses up to 80 μg was well-tolerated and produced some dysphoria and sedation (Walsh et al., 2001b). Higher doses of enadoline (160 μg) had psychomimetic effects including visual and perceptual disturbances, dysphoria, and hallucinations (Walsh et al., 2001b).

Nalbuphine (Nubain^®), is classified as a mixed action mu antagonist/kappa agonist, and is clinically available as an analgesic (Jaffe and Martin, 1990; Preston and Jasinski, 1991; PDR, 2003). Nalbuphine is purported to have relatively low abuse liability and is not scheduled by the FDA (PDR, 2003). We selected nalbuphine for clinical evaluation because of its relatively long duration of analgesic action and safety profile (Zacny et al., 1997; PDR, 2003). Moreover, nalbuphine produced significant and sustained reductions in cocaine self-administration by rhesus monkeys without altering food-maintained responding (Negus and Mello, 2002). Nalbuphine also produced dose-dependent downward shifts in the cocaine self-administration dose-effect curve (Negus and Mello, 2002).

The present study was conducted under double-blind, placebo-controlled conditions in men who met DSM-IV criteria for current cocaine abuse. The effects of nalbuphine alone, cocaine alone and nalbuphine + cocaine in combination on cardiovascular and subjective responses were examined. Each subject was studied as his own control across conditions.

Cocaine A dose of 0.2 mg/kg, i.v. cocaine has proven to be safe and to induce significant changes in mood states and physiological responses in our previous clinical studies. This dose of cocaine usually produces peak plasma cocaine levels of approximately 100 ng/mL (Kaufman et al., 1998; Mendelson et al., 1998, 2001; Sholar et al., 1998).

Nalbuphine The analgesic potency of nalbuphine is equivalent to that of morphine and the recommended analgesic dose is 10 to 20 mg/70 kg (PDR, 2003). In normal healthy volunteers, 10 mg/70 kg nalbuphine and morphine produce similar subjective and physiological effects (Zacny et al., 1997). However, in the present study, 10 mg/70 kg nalbuphine produced vomiting and sedation in three men, so a lower dose of 5 mg/70 kg was used. Two of these subjects completed the study and one withdrew. Nalbuphine doses of 5 or 6 mg/70 kg produced some positive subjective effects (e.g., High, Like Drug Effects, Feel Drug Effect) in previous studies (Zacny et al., 1997; Preston and Bigelow, 2000).

After completion of the baseline subjective-effects questionnaires, a physician administered nalbuphine or placebo intravenously over 15 sec into the subject’s antecubital vein. Fifteen minutes later, cocaine or placebo was administered intravenously over 1 min in a volume of 1 ml. Heart rate, pulse oxymetry, and systolic and diastolic blood pressure were monitored throughout the study session with a noninvasive patient monitor model (Scholar II/507E, Criticare Systems, Waukesha, WI). Subjects and research assistants did not know if active drug or placebo was administered. For safety reasons, the hospital pharmacy informed the physician of what each syringe contained. Vital signs were monitored for at least 4 h after completion of the study. A physician certified in cardiopulmonary resuscitation was present during each study, and a cardiac defibrillator and appropriate emergency treatment medications were located in the study room.

Visual Analog Scales On the 13-item VAS scale, subjects were asked to rate the following drug effects: High, Stimulated, Drug Effect, Good Effect, Euphoria, Sick, Dizzy, Alert, Itchy, Anxious, Bad Effect, Craving and Sleepy. Each of the 13 items was rated from 0 (not at all) to 100 (extremely) on a 100 mm line on the computer screen. VAS ratings were completed at 10, 20, 25, 30, 35, 40, 45, 60, 75, 105 and 135 min after nalbuphine or placebo-nalbuphine administration, or at 5, 10, 15, 20, 25, 30, 45, 60, 90, and 120 min after cocaine or placebo-cocaine administration. The VAS scale was usually completed within 1–2 min.

Addiction Research Center Inventory The ARCI consists of 49 true/false questions, and four subscales were analyzed: (1) the morphine-benzedrine scale (MBG), an index of euphoria; (2) the LSD scale, an index of dysphoria; (3) the PCAG scale, an index of sedation; and (4) the amphetamine, benzedrine scales (AMP) an index of amphetamine-like effects. The ARCI was administered at baseline and at 45 min, and 135 min after nalbuphine or placebo-nalbuphine injection, and this corresponds to 30 and 120 min after cocaine or placebo-cocaine injection. As in previous studies, the ARCI was usually completed within 5 min (Foltin and Fischman, 1991).

Subject-Rated Adjective Checklist A Subject-Rated Adjective Checklist consisted of the following 15 items: Excited, Energetic, Fearful, Suspicious, Nervous, Sweaty, Drunk, Itchy, Tingling, Lightheaded, Spaced-Out, Confused, Irritable, Jittery, and Nauseous. Each item was rated on a 5-point scale from 1 (indicating “not at all”) to 5 (indicating “extremely”). Subjects were asked to rate these adjectives at baseline, and at 10, 20, 25, 30, 35, 40, 45, 60, 75, 105 and 135 min after nalbuphine injection, and this corresponds to 5, 10, 15, 20, 25, 30, 45, 60, 90, and 120 min following cocaine/saline injection. This questionnaire was usually completed within 1–2 min.

Cocaine hydrochloride was acquired from the National Institute on Drug Abuse in powder form and was dissolved in sterile water for intravenous injection by the McLean Hospital pharmacy. Sterility was ensured by passing the solution through a 0.22 micron Milipore filter and subjecting it to Limulus Amebocyte Lysate (LAL) test for detection of gram negative bacterial endotoxins. The test kit is manufactured by Whittaker Bioproducts, Walkersville, MD. Placebo cocaine consisted of the same volume of sterile saline for injection.

Plasma cocaine levels were measured in duplicate using a solid phase extraction method described by SPEC Instruction Manual by Ansys with a Hewlett-Packard 5890 Series II gas chromatograph equipped with a capillary column and a Hewlett-Packard 5971 Series Mass selective detector (Abusada et al., 1993). The assay sensitivity is 10 ng/ml and the intra- and interassay C.V.s were 4.1% and 3.4%, respectively.

Area Under the Curve (AUC) was analyzed for Visual Analog Scale (VAS) ratings and Subject-Rated Adjective Checklist ratings using GraphPad Prism (Version 4.0) (GraphPad Software, San Diego, CA). This analysis has the advantage of quantifying both peak levels and duration of the subjective effect changes. An AUC was calculated for each subject for each subjective effects item under each of the three drug conditions. The AUCs for the group average for nalbuphine alone, cocaine alone and nalbuphine + cocaine were compared with ANOVA for repeated measures. Newman-Keuls post-tests were used to compare each drug condition.

The peak levels of cocaine in plasma usually occurred within 4 to 5 min after i.v. administration. The VAS and Subject-Rated Adjective Checklist ratings at 5 min post-cocaine each were compared with paired t tests for nalbuphine alone, cocaine alone and nalbuphine + cocaine in combination.

Estimates of the primary kinetic parameters (i.e. time to peak plasma concentrations (T_MAX), peak plasma concentration (C_MAX), and half-life (t_½) of plasma cocaine were obtained directly from a nonlinear regression-estimation software program based on the Manual of Pharmacologic Calculations with Computer programs using PHARM/PCS Version 4.2 (MicroComputer Specialist MCS, Philadelphia, PA).

Nalbuphine Drug levels in blood after administration of 5 mg/70 kg, i.v. nalbuphine and 0.2 mg/kg, i.v. cocaine are shown in Figure 1. When nalbuphine alone was administered, serum levels at 10 min after i.v. nalbuphine administration averaged 52.5 ± 6.5 ng/ml and gradually decreased to average 16 ± 5.1 ng/ml at 135 min after nalbuphine (Figure 1, top panel). When nalbuphine was given in combination with cocaine, peak serum nalbuphine levels averaged 59.2 ± 12.8 ng/ml within 10 min after nalbuphine (5 min before cocaine administration) and gradually decreased to average 15.7 ± 4.5 ng/ml at 135 min after nalbuphine. There were no statistically significant differences in serum nalbuphine levels when nalbuphine was given alone in comparison to nalbuphine given in combination with cocaine. No nalbuphine was detected in serum after placebo-nalbuphine injection.

Cocaine When cocaine alone was administered, peak cocaine plasma levels of 131 ± 29.46 ng/ml were measured at 4 min (Figure 1, lower panel). Plasma cocaine levels gradually decreased to average 19 ± 4.4 ng/ml at 135 min. The calculated half-life (t_½) of cocaine in plasma was 49.4 ± 7.7 min. Importantly, pretreatment with nalbuphine did not affect cocaine pharmacokinetics. When cocaine was administered 15 min after nalbuphine peak, plasma cocaine levels of 130 ± 28.2 ng/ml were measured within 4 min after cocaine injection. However, at 2 min after cocaine administration, plasma cocaine levels were significantly higher after nalbuphine administration than after cocaine alone (P< 0.005). Plasma cocaine levels gradually decreased to average 29 ± 1.72 ng/ml at 120 min after cocaine. The calculated half-life (t_½) of cocaine in plasma was 51.5 ± 6.1 min. No cocaine was detected in plasma after placebo-cocaine injection.

Figure 1 Blood Levels of Nalbuphine and Cocaine

Cardiovascular Effects of Nalbuphine Alone Baseline values for heart rate, systolic and diastolic blood pressure averaged 69.4 ± 1.7 beat/min, 121.3 ± 2.8 mmHg and 74.4 ± 1.2 mmHg, respectively. Nalbuphine had no significant effect on heart rate and systolic blood pressure. Nalbuphine decreased diastolic blood pressure within 35 min and these decreases were statistically significant at 75 (P = 0.04) and 105 (P = 0.04) min after nalbuphine administration and returned to baseline at the end of study session. Oxygen saturation did not change significantly after nalbuphine administration (data not shown).

Cardiovascular Effects of Cocaine Alone Baseline values for heart rate, systolic and diastolic blood pressure averaged 73.4 ± 3.9 beat/min, 119 ± 4.0 mmHg and 75.1 ± 2.3 mmHg, respectively. Cocaine increased heart rate significantly, and this increase was sustained for 25 min. Peak increases in heart rate were observed at 10 min after cocaine injection and averaged 101 ± 24 beat/min (P = 0.002). Systolic blood pressure increased significantly above baseline within 5 min after cocaine injection (P = 0.02) and peak levels of 142 ± 9 mmHg were measured at 10 min after cocaine injection (P = 0.001). Systolic blood pressure remained significantly elevated for the first 15 min after cocaine injection (P = 0.005) and then returned to baseline levels. Diastolic blood pressure tended to increase after cocaine, but those changes were not statistically significant. Oxygen saturation did not change significantly after cocaine administration (data not shown).

Cardiovascular Effects of Nalbuphine + Cocaine in Combination Baseline values for heart rate, systolic and diastolic blood pressure averaged 68.6 ± 1.5 beat/min, 119.4 ± 2.8 mmHg and 78.6 ± 2.8 mmHg, respectively. Heart rate increased significantly above the baseline within 5–30 min following cocaine injection when nalbuphine (5 mg/70 kg) and cocaine (0.2 mg/kg) were given in combination. Peak increases in heart rate of 97 ± 20 beat/min were observed at 5 min (P = 0.0003) after cocaine injection, and the magnitude of this increase in heart rate was similar to that after cocaine alone. Systolic blood pressure also increased significantly above baseline following administration of nalbuphine and cocaine. A peak increase in systolic blood pressure of 141.9 ± 3.0 mmHg was measured at 10 min (P = 0.02) after cocaine injection and remained significantly elevated for 15 min following cocaine injection. Nalbuphine and cocaine in combination produced significant decreases in diastolic blood pressure as compared to baseline within 45 min after cocaine injection (P = 0.02). Oxygen saturation did not change significantly after nalbuphine + cocaine administration (data not shown).

Figure 2 Cardiovascular Effects of Nalbuphine, Cocaine and Nalbuphine + Cocaine

Subjective Effects of Nalbuphine Alone

Visual Analog Scale The time course of changes in VAS ratings after nalbuphine administration is shown in Figure 3 and overall changes are shown in Table 1. Nalbuphine produced significant increases in VAS ratings of High and Stimulated (Figure 3), Good Effect, Euphoria, Drug Effect, Sleepy, Dizzy and Bad Effect (Table 1) in comparison to baseline ratings (P< 0.0001–0.03). VAS ratings of High, Stimulated, Good Effect, Drug Effect and Euphoria increased significantly within 10–25 min after nalbuphine administration and remained significantly above baseline ratings for 45–135 min (P< 0.05) (Figure 3). Nalbuphine significantly decreased VAS ratings of Alert within 20 min, and the greatest decreases were observed at 60 min after nalbuphine administration (P< 0.0001). Nalbuphine significantly increased VAS ratings of Sleepy, and increases were maximal at 60 min after nalbuphine administration (P< 0.02). Nalbuphine did not have significant effects on VAS ratings of Sick, Itchy, Anxious or Craving (Table 1).

Subject-Rated Adjective Checklist Nalbuphine significantly increased ratings of Drunk, Tingling, Lightheaded and Spaced-Out and decreased ratings of Excited, Energetic, Suspicious, and Nervous in comparison to baseline (P< 0.05). Ratings of Fearful, Sweaty, Itchy, Confused, Irritable, Jittery and Nauseous did not change significantly in comparison to baseline (Table 1).

Addiction Research Center Inventory Ratings on all four ARCI scales (PCAG, MBG, LSD and amphetamine) increased significantly from baseline when measured at 45 min after nalbuphine injection (P< 0.0001–0.01). (Table 2, column 1).

Figure 3 Time Course of Nalbuphine, Cocaine and Nalbuphine + Cocaine Effects on Visual Analog Scale (VAS) Ratings of High and Stimulated

TABLE 1 SUBJECTIVE EFFECTS OF NALBUPHINE, COCAINE AND NALBUPHINE + COCAINE VISUAL ANALOG SCALE (VAS) AND SUBJECT-RATED ADJECTIVE CHECKLIST

TABLE 2 ARCI SCALES: EFFECTS OF NALBUPHINE, COCAINE AND NALBUPHINE + COCAINE, AND COMPARISONS BETWEEN EACH DRUG CONDITION

Visual Analog Scale Visual Analog Scale (VAS) ratings after cocaine alone are shown in Figure 3 and Table 1. In comparison to baseline ratings, cocaine produced significant increases in ratings of High and Stimulated (Figure 3), Euphoria, Good Effect, Drug Effect, Dizzy and Craving for cocaine in comparison to baseline ratings (P< 0.0001–0.04) (Table 1). Ratings of High, Stimulated, Drug Effect, Good Effect and Euphoria were maximal within the first 5–10 min after cocaine injection, and then gradually declined. There were no significant effects of cocaine on Alert, Sleepy, Sick, Bad Effect, Itchy, and Anxious.

Subject-Rated Adjective Checklist Cocaine significantly increased ratings of Energetic, Sweaty, Drunk, Tingling, Lightheaded, Spaced-Out, and Jittery, and decreased ratings of Nervous in comparison to baseline ratings (P< 0.05). Ratings of Excited, Itchy, Confused and Nauseous increased and ratings of Fearful, Suspicious, and Irritable decreased, but these changes were not significantly different from baseline (Table 1).

Addiction Research Center Inventory Ratings on three ARCI scales [PCAG, LSD and amphetamine] increased significantly at 30 min after cocaine administration (45 min after nalbuphine administration) (P = 0.0003–0.05). No significant changes were measured at 120 min after cocaine injection. Cocaine increased MBG scale ratings, but this was not statistically significant (Table 2, column 2).

Visual Analog Scale In comparison to baseline ratings, nalbuphine + cocaine in combination produced significant increases in VAS ratings of High and Stimulated (Figure 3), Euphoria, , Good Effect, Drug Effect, Dizzy, Sick, Bad Effect, and Craving for Cocaine (Table 1) (P< 0.0001 to 0.04). VAS ratings of High, Stimulated, Good Effect, Drug Effect, and Euphoria were maximal at 5 min after cocaine administration and remained elevated for the next 45–120 min. In contrast, ratings of Alert decreased significantly after administration of nalbuphine + cocaine in combination (P< 0.01) and increases in ratings of Sleepy, Itchy, and Anxious were not significantly different from baseline (Table 1).

Addiction Research Center Inventory Ratings on all four ARCI scales (PCAG, MBG, LSD and amphetamine) increased significantly from baseline after nalbuphine + cocaine administration (P< 0.001–0.02) (Table 2, column 3).

Subject-Rated Adjective Checklist Nalbuphine + cocaine in combination significantly increased ratings of Excited, Sweaty, Drunk, Tingling, Lightheaded, Spaced-Out and Jittery (P< 0.05) in comparison to baseline ratings. There was a trend for nalbuphine + cocaine to increase ratings of energetic (P = 0.09) and to decrease ratings of nervous (P = 0.07). Ratings of Fearful decreased and ratings of Suspicious, Itchy, Confused, Irritable and Nauseous increased after nalbuphine + cocaine administration, but these changes were not statistically significant (Table 1).

Visual Analog Scales Figure 4 shows the average area under the curve (AUC) for ratings on VAS items that are usually sensitive to stimulant effects (left column) and/or to opioid effects (right column). Figure 4 also shows average VAS ratings measured at 5 min after cocaine administration when peak plasma cocaine levels were detected [i.e., 20 min after nalbuphine]. On most measures, the effects of nalbuphine and nalbuphine + cocaine were equivalent. The subjective effects of cocaine alone usually were significantly less that those of nalbuphine + cocaine or nalbuphine alone.

Area Under the Curve ANOVA for repeated measures showed a significant main effect for the following stimulant-related items: Euphoria (P = 0.004), High (P = 0.01), Stimulated (P = 0.007), Good Effect (P = 0.01); and opioid-related items: Sleepy (P = 0.03), Dizzy (P = 0.009). The AUC values for ratings of Euphoria, Good Effect (Figure 4), High and Stimulated were all significantly lower after cocaine alone than after either nalbuphine alone or nalbuphine + cocaine (P< 0.05–0.01). However ratings of Euphoria, High, Stimulated, Good Effect, and Alert did not differ significantly after administration of nalbuphine alone and nalbuphine + cocaine. Average AUC values for VAS ratings of Sleepy and Dizzy were significantly higher after both nalbuphine alone and nalbuphine + cocaine than after cocaine alone (P< 0.05–0.01) (Figure 4). The VAS ratings of Sleepy, Dizzy, Sick, and Bad Effect did not differ significantly after nalbuphine alone and nalbuphine + cocaine. The average AUC values for VAS ratings of Sick, Bad Effect, Itchy and Alert did not differ significantly across the three drug conditions.

Peak VAS Ratings Subjects reported significantly greater feelings of Euphoria, High, Stimulated, Alert and Anxious after administration of nalbuphine + cocaine than after nalbuphine alone (P< 0.05–0.01). Ratings of High, Stimulated and Good Effect were higher after administration of nalbuphine + cocaine than after cocaine alone (P< 0.05) (Figure 4, column 1). There were no significant differences between nalbuphine + cocaine and cocaine alone for peak ratings of Sleepy, Dizzy (Figure 4), Sick, Bad Effect, Alert, Itchy, Anxious or Craving for cocaine (data not shown).

Figure 4 Visual Analog Scale (VAS) Ratings and Subject-Rated Adjective Checklist Responses after Nalbuphine, Cocaine and Nalbuphine + Cocaine

Area Under the Curve ANOVA for repeated measures showed a significant main effect for the following items: Energetic (P = 0.026), Drunk (P = 0.006), Spaced-Out (P = 0.0003) and Lightheaded (P = 0.003). Ratings of Excited, Energetic and Jittery were higher after nalbuphine + cocaine than after either cocaine or nalbuphine alone, but these differences were statistically significant only for Energetic (P< 0.05). In comparison to cocaine alone, ratings of Drunk, Spaced-Out and Lightheaded were significantly higher after both nalbuphine alone and nalbuphine + cocaine (P< 0.05–0.001). Ratings of Spaced-Out were significantly higher after nalbuphine alone than in combination with cocaine (P< 0.05).

Peak Adjective Checklist Ratings There were no significant differences between the three drug conditions in peak Adjective Checklist ratings of Energetic (Figure 4), Lightheaded, Fearful, Suspicious, Nervous, Sweaty, Itchy, Irritable, Tingling, Nauseous and Confused (data not shown). In comparison to nalbuphine alone, peak ratings of Excited, Energetic, Jittery, and Drunk tended to be higher after both cocaine alone and after nalbuphine + cocaine. However, peak ratings were significantly higher after cocaine alone than after nalbuphine alone only for Excited (P = 0.003) and Jittery (P< 0.05). Ratings after nalbuphine + cocaine were significantly higher than after nalbuphine alone for Excited (P = 0.003), Jittery (P = 0.01) and Drunk (P = 0.01). Ratings of Spaced-Out were significantly higher after either nalbuphine alone (P = 0.02) or in combination with cocaine (P = 0.02) than after cocaine alone.

Figure 5 Comparison of VAS Ratings After Cocaine Alone and Nalbuphine + Cocaine Minus Nalbuphine Alone

This is the first clinical study to examine the interactions between nalbuphine, a mixed mu/kappa opioid agonist and cocaine in male cocaine abusers. In preclinical studies, kappa opioids attenuated many of the abuse-related and locomotor activating effects of cocaine, and these findings suggested that kappa opioids might be useful for the clinical treatment of cocaine abuse (Archer et al., 1996; see for review Mello and Negus, 2000). Cocaine self-administration studies in non-human primates indicated that kappa opioids with mu receptor activity reduced cocaine self-administration more effectively, and with fewer adverse side effects, than kappa selective opioids (Mello and Negus, 2000; Bowen et al., 2003). Accordingly, we chose nalbuphine from the very limited number of kappa opioids approved for clinical use by the FDA. Nalbuphine is classified as a kappa agonist/partial mu antagonist analgesic in the Physicians Desk Reference (PDR, 2003), and has a complex pharmacological profile (Schmidt et al., 1985; De Souza et al., 1988; Jaffe and Martin, 1990; Preston and Jasinski, 1991). Nalbuphine precipitated withdrawal in both morphine- and methadone-dependent subjects, and did not reduce abstinence signs and symptoms in opioid-dependent subjects during withdrawal from morphine maintenance (Preston and Jasinski, 1991).

One major finding of this study was that a 5 mg/70 kg dose of nalbuphine was safe and well tolerated when administered in combination with cocaine. There was no evidence of synergistic effects and, in general, the cardiovascular effects of the nalbuphine + cocaine combination were less than the effects of cocaine alone. Moreover, plasma/serum levels of cocaine and nalbuphine were equivalent when each drug was given alone and in combination. A second major finding was that the effects of nalbuphine + cocaine were not additive. Ratings of positive subjective effects after nalbuphine alone and nalbuphine + cocaine did not differ significantly, and both were greater than after administration of cocaine alone. The relation of these findings to previous clinical and preclinical studies of opioid-cocaine interactions are described below. The implications of these results for agonist medication approaches to therapy, and some limitations of this study are discussed.

Nalbuphine There were no significant changes in heart rate and systolic blood pressure following nalbuphine alone, and this is consistent with previous clinical studies. For example, when nalbuphine was given to healthy volunteers in a dose range of 2.5–10 mg/70 kg, it did not affect systolic or diastolic blood pressure, or heart rate (Zacny et al., 1997). In the present study, nalbuphine produced significant decreases in diastolic blood pressure at 75 min and 105 min. These decreases in diastolic blood pressure may reflect sedation, because subjects reported feeling Sleepy, Drunk and Spaced-Out after nalbuphine administration.

Cocaine Alone and in Combination with Nalbuphine The cardiovascular effects of cocaine alone were consistent with findings from several clinical laboratory studies (Foltin and Fischman, 1992; Foltin et al., 1995; Abreu et al., 2001; Walsh et al., 2001b). The combined administration of nalbuphine + cocaine did not significantly alter heart rate and systolic blood pressure in comparison to cocaine alone, and this is in agreement with a study of cocaine and hydromorphone combinations (Walsh et al., 1996). When cocaine was given in combination with hydromorphone, increases in heart rate and pressor effects were slightly greater than after cocaine alone, but the observed differences were not statistically significant (Walsh et al., 1996). However, when morphine was given in combination with cocaine, the cardiovascular effects of some dose combinations were significantly greater than the cardiovascular effects of cocaine alone (Foltin and Fischman, 1992). The increases in cardiovascular measures produced by the morphine + cocaine combination were less than predicted by an addition model of drug interaction (Foltin and Fischman, 1992).

Nalbuphine and Cocaine Alone Both cocaine and nalbuphine produced significant increases in the VAS and Subject-Rated Adjective Checklist items and ARCI subscale scores in comparison to baseline. The time course and magnitude of increases the VAS and Adjective Checklist measures of positive subjective effects after cocaine and nalbuphine alone was similar to that reported in other clinical laboratory studies (Foltin and Fischman, 1992; Walsh et al., 1996; Zacny et al., 1997). The sedative effects of nalbuphine were also similar in time course and magnitude to those of the same dose of i.v. nalbuphine administered to normal volunteers (Zacny et al., 1997). Moreover, the VAS scores for Euphoria, High, Stimulated and Good Effect were significantly higher after nalbuphine alone than after cocaine alone. These findings are similar to previous studies in which morphine (10 mg/70 kg) produced greater subjective effects then 16 mg/70 kg cocaine (Foltin and Fischman, 1992). The opioid-related VAS items, Sleepy and Dizzy, also were significantly higher after nalbuphine than after cocaine. However, the profile of responses on the ARCI subscales differed in that nalbuphine alone increased all ARCI scale scores in these cocaine abusers, but nalbuphine significantly decreased scores on the BG scale in drug-naïve volunteers (Zacny et al., 1997).

Nalbuphine + Cocaine in Combination It was surprising to find that the positive subjective effects of nalbuphine alone usually were not significantly different from the effects of nalbuphine + cocaine. This finding is not consistent with reports that cocaine often appears to enhance the effects of opioids. Rather, the subjective responses to nalbuphine + cocaine appeared to be accounted for primarily by nalbuphine, and there was no evidence that the interactions between cocaine and nalbuphine were additive. The contribution of cocaine to the nalbuphine + cocaine combination was estimated by subtracting the average subjective effects of nalbuphine alone from the average subjective effects of nalbuphine + cocaine. Analysis of peak responses, measured 5 min after cocaine, and the AUC indicated that the estimated positive subjective effects of cocaine after administration of the nalbuphine + cocaine combination were less than after cocaine alone. This suggests that nalbuphine may have slightly attenuated the subjective effects of cocaine. This interpretation is consistent with our preclinical findings that chronic administration of mixed mu/kappa opioids selectively reduced cocaine self-administration as described below.

In contrast, in rhesus monkeys, chronic treatment with nalbuphine (0.1 mg/kg/hr or 1.0 or 3.0 mg/kg/day), as well as several other mixed mu/kappa opioids, significantly reduced cocaine self-administration over 7 to 10 days (Mello et al., 1993; Negus et al., 1997; Mello and Negus, 1998b; Negus and Mello, 2002; Bowen et al., 2003). The acute effects of mixed mu/kappa agonists on cocaine’s reinforcing and discriminative stimulus effects have been less consistent. In a drug self-administration procedure where treatment drug effects on multiple unit doses of cocaine could be evaluated in two or three sessions (Caine et al., 2000), the mixed mu/kappa agonists (-)cyclorphan, Mr2034 and MCL-101 dose-dependently decreased cocaine self-administration with minimal side effects (Bowen et al., 2003). However, acute administration of the same mixed mu/kappa agonists that reduced cocaine self-administration did not significantly attenuate cocaine discrimination, and did not produce rightward shifts in the cocaine discrimination dose-effect curve in rhesus monkeys (Negus and Mello, 1999; Bowen et al., 2003).

When an acute dose of morphine or hydromorphone was combined with i.v. cocaine administration, the ratings of positive subjective effects were usually higher than after either cocaine or the opioid alone (Foltin and Fischman, 1992; Walsh et al., 1996). For example, combinations of 20 mg cocaine + 1.5 mg hydromorphone or 40 mg cocaine + 3.0 mg hydromorphone produced higher VAS ratings than either drug alone, but these differences were not statistically significant (Walsh et al., 1996). On an Adjective Rating Scale, a number of subjective effects typical of cocaine (Stimulated, Excited) and opioid effects (Nodding, Itchy, Talkative) were significantly greater after cocaine + hydromorphone in combination, than after either drug alone (Walsh et al., 1996). When morphine (0, 5, and 10 mg/70 kg) was given in combination with cocaine (0, 8, 16 and 32 mg/70 kg), positive subjective effects (High, Liking) were significantly greater than for either drug alone (Foltin and Fischman, 1992). The authors emphasized that the increases in subjective effects ratings after cocaine (32 mg) and morphine (5 mg) in combination were less than would be predicted by an effect-addition model of drug interaction (Foltin and Fischman, 1992).

There is also evidence from preclinical studies that mu opioids may increase the abuse-related effects of cocaine. For example, opioid administration usually increases the discriminative stimulus effects of cocaine in squirrel monkeys (Spealman et al., 1992; Spealman and Bergman, 1994; Rowlett et al., 1998) and in rhesus monkeys (Mello et al., 1995; Negus et al., 1998; Negus and Mello, 2002). Cocaine and heroin also may enhance each other’s reinforcing and discriminative stimulus effects when administered in a speedball (Rowlett and Woolverton, 1997; Negus et al., 1998). Speedball combinations of cocaine and heroin produce robust reinforcing and discriminative stimulus effects in rats and rhesus monkeys (Mello et al., 1995; Hemby et al., 1996; Rowlett and Woolverton, 1997; Mello and Negus, 1998a; Negus et al., 1998; Rowlett et al., 1998; Hemby et al., 1999). Moreover, neurochemical findings suggest that self-administration of cocaine and heroin in combination has synergistic effects on extracellular dopamine release at the nucleus accumbens in rats in comparison to heroin or cocaine alone (Hemby et al., 1999).

Conclusions We conclude that under the conditions of this experiment, a combination of nalbuphine and cocaine was safe and did not produce synergistic cardiovascular effects. Moreover, the addition of cocaine did not significantly increase the subjective effects of nalbuphine. Rather, subjective responses to nalbuphine + cocaine were usually equivalent to responses to nalbuphine alone. These findings suggest that acute administration of nalbuphine may produce a modest attenuation of cocaine’s abuse-related effects. Further studies of the effects of chronic nalbuphine treatment on the physiological and subjective effects of cocaine are warranted.

This research was supported by grants P01-DA14528, U19-DA11007, K05-DA00101 (NKM), K05-DA00064 and T32-DA07252 (JHM) from the National Institute on Drug Abuse, National Institutes of Health. Preliminary data were reported to the College on Problems of Drug Dependence in 2003. We thank Alicja Skupny and Yonghong Cheng for performing the nalbuphine and cocaine assays and Inge Knudson for assistance with data analysis and for preparing the graphic displays.