Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Cocaine is a highly addictive agent that has been associated with myocardial ischemia, infarction and arrhythmias, sudden cardiac death and cardiomyopathies (for review see Minor et al., 1991). Several agents have been proposed as possible treatments for cocaine addiction and/or toxicity (Witkin, 1994). These include other reuptake blockers such as desipramine (Gawin and Kleber, 1984; Tennant and Rawson, 1983) and GBR 12909 (Rothman and Glowa, 1995; Rothman et al., 1989, 1991), dopamine agonists such as bromocriptine (Dackis and Gold, 1985; Hubner and Koob, 1990) and the mixed opiate agonist-antagonist, buprenorphine (Mello et al., 1989). In addition, several agents have been suggested to reduce toxicity to cocaine including the benzodiazepine, diazepam (Catravas and Waters, 1981; Derlet and Albertson, 1990), buprenorphine (Shukla et al., 1991; Witkin et al., 1991) and the NMDA receptor antagonist, MK-801 (Derlet and Albertson, 1990; Rockhold et al., 1991).

The mechanisms by which these agents act on the central nervous system and on behavior have been investigated by many laboratories. In contrast, the effects of these agents alone or in combination with cocaine on cardiovascular function are poorly understood. For example, bromocriptine and buprenorphine have been reported to produce modest decreases in arterial pressure in humans (Preston et al., 1992; Scott et al., 1980) whereas desipramine does not change arterial pressure in rabbits (Dorward et al., 1991). MK-801 has little effect in anesthetized dogs (Hageman and Simor, 1993) but increases arterial pressure and heart rate in conscious rats (Lewis et al., 1989). Because most data available are limited to studies of arterial pressure or heart rate, our study was conducted to better characterize the hemodynamic responses to these agents and their response profiles in combination with cocaine.

Understanding the interactions between cocaine and proposed treatments is important for several reasons. First, treatments for addiction should be examined for possible interactions with cocaine due to the high rate of recidivism among cocaine users. Noncompliant patients may experience additive or synergistic effects because many proposed treatments mimic the neurochemical effects of cocaine to reduce sensitivity to the cocaine-induced euphoria. This may result in enhanced predisposition to cardiovascular toxicity. Second, a better understanding of the actions of these agents on cocaine-induced responses may help to elucidate the mechanisms by which cocaine causes cardiovascular responses and toxicity. This may contribute to better design of treatments for addiction that may also reduce toxicity. Our experiment was designed to examine these interactions using doses of proposed treatments for addiction that were both clinically relevant and had minimal effects alone on hemodynamic variables.

It is known that individuals vary widely in their sensitivity to cocaine-induced coronary vasoconstriction (Lange et al., 1989), myocardial ischemia (Isner et al., 1986; Minor et al., 1991), cardiomyopathies (Minor et al., 1991) and mortality (Mittleman and Wetli, 1987; Smart and Anglin, 1987). These observations suggest that some individuals are at greater risk for severe cocaine-induced cardiovascular complications. We have proposed that the rat may provide a model to determine the causes of differential cardiovascular sensitivity and toxicity to cocaine (Branch and Knuepfer, 1993; Knuepfer et al., 1993a). We reported that in some but not all rats cocaine administration elicited a clear decrease in cardiac output and a substantial (>80%) increase in systemic vascular resistance whereas in the remaining rats cocaine elicited consistently little change or an increase in cardiac output and smaller increases in systemic vascular resistance (Branch and Knuepfer, 1993, 1994a; Knuepfer and Branch, 1993). In our report, we designated the groups vascular and mixed responders (formerly named responders and nonresponders), respectively. The response characteristics of individual rats appear to be consistent at several doses and are not altered by higher doses of cocaine (Branch and Knuepfer, 1993, 1994a). Vascular responders have a greater incidence of cardiomyopathies and sustained hypertension after repeated cocaine administration and both a spike in sympathetic nerve activity and a greater pressor response under chloralose anesthesia after cocaine administration (Branch and Knuepfer, 1994a, 1994b; Knuepfer et al., 1993a). Because this differential sensitivity to cardiomyopathies resembles the varying susceptibility of humans to cocaine-induced cardiotoxicity, we have used this as a possible model for identifying responses in more sensitive individuals although it remains to be proven whether this differential responsiveness in rats is truly related to differential sensitivity to toxicity in humans.

Our study was performed to determine whether several putative treatments for cocaine addiction and toxicity alter hemodynamic responses to cocaine. We examined the effects of desipramine, GBR 12909, bromocriptine, buprenorphine, diazepam and MK-801 on cardiovascular responses to cocaine. Although the mechanism by which these agents act varies widely, all have been used or proposed for use in treating cocaine addiction or toxicity. The results demonstrate that specific treatments alter hemodynamic responses to cocaine. Although the primary goal of these studies was not to examine the mechanism of cocaine's cardiovascular effects in detail, the results suggest possible neurotransmitters that may facilitate or inhibit the unique cardiac output responses to cocaine in individuals.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Animal preparation. Male Sprague-Dawley rats (Harlan, Indianapolis, IN) weighing 300 to 420 g were surgically prepared under pentobarbital sodium (50 mg/kg, i.p.) anesthesia using aseptic technique as previously described (Branch and Knuepfer, 1993, 1994a; Knuepfer and Branch, 1992; 1993). Briefly, a thoracotomy was performed and a pulsed Doppler flow probe (2.4-mm cuff diameter, 20 MHz, Iowa Doppler Products, Iowa City, IA) filled with acoustic gel was sutured snugly on the ascending aorta. The thorax was closed and the lead wires brought subcutaneously to a socket on the skull. Rats were treated with cefazolin (10 mg/kg, i.m., once daily for 3 days) and allowed to recover for a minimum of 10 days. Rats with poor or varying velocity signals or that did not recover normal motor and feeding behavior within 24 hr were euthanized with pentobarbital. After recovery, rats were anesthetized with methoxyflurane for implantation of femoral arterial and venous cannulas filled with 15 mg/ml cefazolin. In a separate group of six rats, arterial and venous cannulas were implanted to examine the effects of higher doses of specific agents on arterial pressure and heart rate, only. One to two days later, each rat was acclimated in a Plexiglas cage for 6 hr. On the next day, rats were placed in the same cage for 2 hr before beginning experimentation.

Experimental procedure. The procedures employed in these experiments have been described in detail (Branch and Knuepfer, 1993, 1994a; Knuepfer and Branch, 1992, 1993). During and after the daily 2-hr acclimation period, arterial pressure, heart rate and blood flows were monitored continuously. Rats were studied for up to 10 days. Cocaine hydrochloride (5 mg/kg, i.v., infused over 45 sec) alone or after pretreatment with another agent was administered twice daily with a minimum cocaine dosing interval of 4 hr. In most cases, cocaine was delivered alone in the morning and was given 10 min after pretreatment in the afternoon. We have not observed significant tachyphylaxis of cardiovascular responses to cocaine when given alone in the morning and afternoon nor when given twice daily for up to 6 days (Branch and Knuepfer, 1994a). All experiments were conducted between the hours of 9 A.M. and 4 P.M. in a quiet room.

Materials. Materials used included the methanesulfonate salt of 2-bromo--ergocryptine (bromocriptine) and desipramine hydrochloride from Sigma Chemical Company (St. Louis, MO). Cocaine hydrochloride was obtained from the National Institute on Drug Abuse. GBR 12909, provided by NOVO-Nordisk Pharmaceuticals, Malov, Denmark through the Medications Development Division of the National Institute on Drug Abuse (NIDA), was prepared in 3 mg/ml tartaric acid solution (Fischer Scientific Co., Fair Lawn, NJ). Buprenorphine was obtained in solution from Reckitt & Colman Pharmaceuticals, Inc. (Richmond, VA). MK-801 ((+)-5-methyl-10,11-dihydro-5H-dibenzo[a, d]cyclohepten-5,10-imine hydrogen maleate) was purchased from Research Biochemicals. Inc. Diazepam was supplied by Hoffman-La Roche, Inc. (Nutley, NJ) in ampules containing a solution of 5 mg/ml. Drugs were dissolved in 0.9% sterile saline and were administered i.v. in a final volume of 1 ml/kg over a period of approximately 45 sec. Drug concentrations were calculated as the salt form. Cefazolin (Geneva Pharmaceuticals/Marsam Pharmaceuticals, Cherry Hill, NJ) was used postoperatively to reduce the risk of sepsis.

Data analysis. Data were analyzed at several time points. First, the peak arterial pressure response to cocaine, invariably occurring within the first minute, was recorded. A second set of values was obtained at the time of the peak change in cardiac output if it was not coincident with the peak change in arterial pressure. Using the data at the time of the maximum change in cardiac output, rats were classified as mixed or vascular responders. In addition, data were obtained during the sustained modest pressor response defined at 1, 3 and 5 min after initiating cocaine injection. Peak data points and sustained responses were examined separately using analysis of variance to avoid the occurrence of significant interactions. Two-way analysis of variance (for studies of vascular and mixed responders) included a post hoc simple main effects test to determine which groups were different. With one exception described below, these procedures have been used in previous reports (Branch and Knuepfer, 1994a; Knuepfer and Branch, 1992, 1993). In our study, instead of comparing hemodynamic responses to cocaine after drug pretreatment to the precocaine (postpretreatment) levels, cocaine-induced changes were determined from baseline levels before administration of the pretreatment. This change allows for consideration of the effects of altered baselines due to drug pretreatment on cocaine-induced responses. The figures of drug time courses reflect any differences in baselines that occurred with pretreatments.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Conscious rats instrumented for cardiac output determination (n = 33) had a mean arterial pressure of 117.1 ± 1.7 mmHg, a heart rate of 388 ± 4 bpm and an ascending aortic velocity signal of 9.7 ± 0.3 kHz shift. Cocaine administration (5 mg/kg, i.v.) elicited pressor responses and variable changes in cardiac output and systemic vascular resistance. Each rat received cocaine alone several times (3-12 trials, mean = 7.8 ± 0.5 trials) to determine hemodynamic responsivity. Individual rats (n = 26) were designated vascular responders if the mean maximum decrease in cardiac output was more than 8% (mean = -15.4 ± 1%). The remaining rats, classified as mixed responders, had smaller decreases or increases in cardiac output (mean = 8.2 ± 3.8%). The resting arterial pressures and heart rates were not different between groups. In contrast, the mean ascending aortic flow signals were significantly different in vascular and mixed responders (10.0 ± 0.3 and 8.45 ± 0.5 kHz shift, respectively). Figure 1 depicts the mean responses to cocaine alone in all rats. Five rats were tested with two pretreatment drugs (on different days), five were tested with three drugs and two were examined in four different experimental protocols. In all cases, control responses to cocaine were repeated before each pretreatment regimen to verify consistency of responses.

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Mean hemodynamic responses to intravenous cocaine (5 mg/kg, i.v., injected over 45 sec as depicted by the bar labeled COC) in vascular responders (open squares, n = 24) and mixed responders (filled squares, n = 7). Specific responses shown include mean arterial pressure (MAP, mmHg), heart rate (HR, bpm), cardiac output (CO, % change) and systemic vascular resistance (SysVR, % change). Control values are given in the first sentence of "Results." Data were analyzed with an unpaired Students' t test at the time of the peak pressor response and with a two-way ANOVA and simple main effects test during the sustained response (1, 3 and 5 min after cocaine administration). Asterisks denote significant (P < .05) differences between groups (vascular and mixed responders) at specific time points.

In a separate group of rats instrumented for arterial pressure and heart rate determination only (n = 7), mean arterial pressure was 116.8 ± 2.2 mmHg and heart rate was 404 ± 9 b/min. These rats were used to determine appropriate doses of some pretreatment drugs.

Effects of bromocriptine. Resting hemodynamic values between vascular and mixed responders were similar (table 1). The D₂ receptor agonist, bromocriptine (0.1 mg/kg, i.v.), elicited a biphasic arterial pressure response; a brief pressor response within 30 to 90 sec after injection (fig. 2) followed by a small depressor response (table 2). The pressor response was caused by an increase in cardiac output in mixed responders (P = .02) and by an increase in systemic vascular resistance in vascular responders (P = .045, table 2; fig. 2). Five minutes later, arterial pressure was lower in vascular responders only due to a decrease in systemic vascular resistance while cardiac output returned to pre-drug levels (table 2, time 0 in fig. 3). Heart rate and stroke volume were not significantly affected by this dose of bromocriptine. A larger dose of bromocriptine (1 mg/kg) elicited variable changes in arterial pressure and heart rate in six rats instrumented with arterial and venous cannulas only (table 3). An analysis of variance revealed a significant increase in arterial pressure for both doses.

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Cardiovascular effects of 0.1 mg/kg bromocriptine (Brc), 1 mg/kg desipramine (Des), 0.1 to 0.5 mg/kg diazepam (Dzp), 0.3 mg/kg buprenorphine (Bup) and 50 µg/kg MK-801 (MK8) are shown at the time of the peak pressor response. Asterisks denote significant changes (P < .05) from control values as determined by a two-way ANOVA (bromocriptine, desipramine and diazepam) and by a Students' paired t test (buprenorphine and MK-801). Other abbreviations are described in figure 1.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Responses to cocaine (5 mg/kg, i.v.) before and after pretreatment (5 min before) with bromocriptine (Brc) in vascular (n = 6) and mixed responders (n = 6). Control values are shown in table 1. As denoted by asterisks at time zero, arterial pressure was reduced in vascular responders and systemic vascular resistance was reduced in all rats five minutes after bromocriptine administration (table 2). Data were analyzed with a two-way ANOVA at the time of the peak pressor response and with a three-way ANOVA during the sustained response (1, 3 and 5 min after cocaine administration). Asterisks denote differences due to the drug (bromocriptine) pretreatment. In addition, vascular responders had a smaller decrease in cardiac output at one minute and smaller increases in systemic vascular resistance at 1, 3 and 5 min whereas mixed responders had a smaller increase in systemic vascular resistance only at 5 min after cocaine administration. Abbreviations are described in figure 1.

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Arterial pressure and cardiac output responses to cocaine at the time of the peak change in cardiac output (0.5-3 min after cocaine). Responses to cocaine alone (5 mg/kg, i.v., Coc) are depicted for all control values combined. In addition, responses to cocaine after pretreatment with bromocriptine (0.1 mg/kg, Brc), desipramine (1 mg/kg, Des), GBR 12909 (1 mg/kg, GBR), diazepam (0.1 and 0.5 mg/kg, Dzp), buprenorphine (0.3 mg/kg, Bup) and MK-801 (0.05 mg/kg, MK) are shown. Data were analyzed by one- (buprenorphine and MK-801) or two-way (all other drugs) ANOVA. There was a significant interaction for the maximum cardiac output response noted after GBR 12909 pretreatment because the mixed responders had a negative change and the vascular responders had a positive change. Asterisks denote a significant change (P < .05) in comparison to cocaine alone for control values obtained in each experiment. The control values for cocaine administration are a mean obtained from each animal used and are not necessarily represented by the combined control value shown for cocaine alone. Significant differences between vascular and mixed responders are designated with a #. Other abbreviations are described in figure 1.

Effects of desipramine. Desipramine was studied in 14 rats (table 1). The monoamine uptake inhibitor, desipramine (1 mg/kg, i.v.), increased arterial pressure in all rats (fig. 2) within 1 to 2 min after administration. As seen with bromocriptine, the pressor response was caused by an increase in cardiac output in mixed responders and with an increase in systemic vascular resistance in vascular responders (fig. 2). Furthermore, heart rate fell in vascular responders only. Stroke volume was not altered in either group. A larger dose of desipramine (10 mg/kg, i.v.) elicited an equivalent peak increase in arterial pressure in five conscious rats without cardiac output instrumentation (table 3). Ten minutes later, arterial pressure remained elevated in all rats but vascular responders had a significantly higher resting arterial pressure than mixed responders. The increase in arterial pressure was due to an increase in systemic vascular resistance (table 2, time 0 in fig. 5) because heart rate and cardiac output remained depressed in vascular responders only (table 2).

View larger version (28K): [in this window] [in a new window]   Fig. 5.   Responses to cocaine (5 mg/kg, i.v.) before and after pretreatment with desipramine (Des, 1 mg/kg, i.v.) in vascular (n = 6) and mixed responders (n = 7). Control values are shown in table 1. As denoted by asterisks at time zero, crease in cardiac output and increase in systemic vascular resistance.

Effects of GBR 12909. The effects of administration of the vehicle for GBR 12909 (3 mg/ml tartaric acid) were examined in 10 conscious rats. Vehicle injections elicited increases in arterial pressure and heart rate that were not unlike those elicited by low doses of GBR 12909 (table 3; fig. 6). There were no differences in hemodynamic parameters 5 min after vehicle injection (fig. 6).

View larger version (26K): [in this window] [in a new window]   Fig. 6.   Responses to GBR 12909 (0.5-5 mg/kg, i.v.) compared to vehicle (Veh) responses in vascular and mixed responders. A dose-response relationship was noted for all parameters shown using a three-way ANOVA for data at 1, 3 and 5 min after GBR 12909 administration. Peak arterial pressure responses were related to dose also (two-way ANOVA). No differences were noted between the effects of GBR 12909 on vascular responders compared to mixed responders so the combined data are presented. Other abbreviations are described in figure 1.

View larger version (27K): [in this window] [in a new window]   Fig. 7.   Responses at several time points to administration of cocaine (5 mg/kg, i.v.) alone (squares) and 5 min after administration of GBR 12909 (1 mg/kg, i.v., circles) in conscious, freely-moving rats (n = 15). Control values are shown in table 1. Asterisks at time zero denote the increase in arterial pressure and cardiac output elicited by GBR 12909 alone. At 1-min after cocaine administration, cardiac output and stroke volume were significantly reduced and systemic vascular resistance was elevated in mixed responders only. Data were analyzed and displayed as described in figure 3.

Effects of diazepam. There was a difference in pretreatment baseline values for heart rate between mixed and vascular responders in rats studied using diazepam (table 1). Two doses of diazepam (0.1 and 0.5 mg/kg) were administered to rats instrumented for cardiac output determination (n = 12 and 6, respectively). Dose-related differences in arterial pressure, heart rate and cardiac output baseline values or cocaine-induced responses were not observed using analysis of variance. Therefore, these data were combined in figures 2, 4 and 8. Diazepam pretreatment elicited an initial increase in arterial pressure in all rats within 60 to 90 sec due to an increase in cardiac output (fig. 2). Mixed responders, but not vascular responders, demonstrated an increase in heart rate also (fig. 2). After 10 min, arterial pressure was significantly lower compared to baseline values in all rats although the differences were only significant in vascular responders (table 2). No apparent sedative effects were noted but rats were relatively quiescent before and after all drugs administered except cocaine. In contrast, a higher dose of diazepam (1 mg/kg, i.v.) elicited an initial behavioral excitation in some rats (as noted by increased motor activity) followed by an apparent lethargy (lying on cage bottom for several minutes) in all six rats tested. These responses were associated with an initial increase in arterial pressure (table 3) that was no longer apparent 10 min later. Heart rate was elevated for the entire 10-min period before cocaine administration.

View larger version (28K): [in this window] [in a new window]   Fig. 8.   Responses to cocaine (5 mg/kg, i.v.) before and after pretreatment with diazepam (Dzp, 0.1 or 0.5 mg/kg, i.v.) in vascular (n = 13) and mixed responders (n = 5). Control values are shown in table 1. An ANOVA on the dose-response relationship demonstrated no differences in cocaine-induced effects so responses to the two doses were combined. Arterial pressure was lower 10 min after treatment with diazepam (designated by asterisks at time zero). Data are presented and analyzed as described in figure 3.

Effects of buprenorphine. Buprenorphine (0.3 mg/kg) alone produced a delayed (6-7 min) increase in arterial pressure due to increases in heart rate, cardiac output and systemic vascular resistance in seven vascular responders (fig. 2, mixed responders were not tested) without affecting stroke volume. Ten minutes after administration of buprenorphine, arterial pressure, heart rate and cardiac output were still significantly elevated (table 2 and time 0 in fig. 9). The effects of cocaine alone and of buprenorphine plus cocaine were similar except that the increase in systemic vascular resistance was enhanced by buprenorphine pretreatment 1 minute after cocaine (fig. 9). Responses at the time of the maximum cocaine-induced decrease in cardiac output were not changed significantly by buprenorphine pretreatment (fig. 4).

View larger version (22K): [in this window] [in a new window]   Fig. 9.   Responses to cocaine (5 mg/kg, i.v.) before and after pretreatment with buprenorphine (0.1 mg/kg, i.v.) in vascular responders (n = 7). Control values are shown in table 1. Arterial pressure, heart rate and cardiac output were significantly elevated 10 min after administration of buprenorphine (denoted by asterisks at time zero). Data were analyzed by a one-way ANOVA (peak values) and by two-way ANOVA (1, 3 and 5 min). Abbreviations are described in figure 1.

Effects of MK-801 (dizocilpine). Administration of MK-801 (50 µg/kg) evoked an increase in arterial pressure mediated by increases in systemic vascular resistance, heart rate, and cardiac output in nine vascular responders (mixed responders were not tested) that reached peak values approximately 7 to 9 min after administration (fig. 2). These values remained elevated 10 min after administration when cocaine was injected. The arterial pressure and heart rate responses to cocaine administration were greater (fig. 10) due to higher baseline values (table 2). The pressor response appeared to be due to an increase in baseline cardiac output possibly due to inhibition of the bradycardia. At the time of the peak cardiac output response, the cardiovascular responses were unaltered (fig. 4). Stroke volume responses were not altered (data not shown).

View larger version (23K): [in this window] [in a new window]   Fig. 10.   Responses to cocaine (5 mg/kg, i.v.) before and after pretreatment with MK-801 (dizocilpine, 0.05 mg/kg, i.v.) in vascular responders (n = 9). Control values are shown in table 1. As denoted by asterisks at time zero, arterial pressure, heart rate, cardiac output and systemic vascular resistance were elevated 10 minutes after administration of MK-801 (table 2). Data were analyzed as described in figure 9 and are presented as described in figure 1.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

These data provide the first detailed description of cardiac output and systemic vascular resistance responses to a variety of agents proposed for the treatment of cocaine addiction and/or toxicity. The responses to proposed treatments and to the combination of the treatments and cocaine were described in two subsets of a population. Our laboratory has reported that the cardiac output responsiveness to cocaine is highly variable and is correlated with their predisposition to the development of cocaine-induced hypertension and cardiomyopathies (Branch and Knuepfer, 1994a; Knuepfer et al., 1993a). We divided rats into vascular responders (with a decrease in cardiac output) and mixed responders (no change or an increase in cardiac output) to facilitate analysis of the differences in responsivity. In our study, resting ascending aortic flow values were significantly higher in vascular responders compared to mixed responders. It may be argued that this difference may predispose these rats to a decrease in cardiac output in response to cocaine. This is unlikely because we have published several reports using relatively large numbers of rats classified in this manner (Branch and Knuepfer, 1993, 1994a; Knuepfer et al., 1993a, b). We have not noted such differences in these studies. Therefore, this difference may contribute to the differential responsiveness but is not likely to be the sole cause.

Amphetamine administration (1 mg/kg), ethanol administration (0.475 or 0.95 mg/kg) or a brief air jet stress evoke differential cardiovascular responsiveness that is directly related to the differential hemodynamic responses elicited by cocaine in vascular and mixed responders (Branch and Knuepfer, 1994a; Gan and Knuepfer, 1994; Knuepfer et al., 1993b). In our study, bromocriptine (0.1 mg/kg, i.v.) or desipramine (1 mg/kg, i.v.) evoked an acute increase in arterial pressure in all rats. The pressor response was a result of increasing cardiac output in mixed responders and of increasing systemic vascular resistance in vascular responders. These data suggest that rats exhibit differential hemodynamic responses to dopamine (D₂ receptor) agonists or reuptake blockers in addition to acute stress, ethanol and amphetamine. These data provide further support for a differential sensitivity to a range of drug treatments whether these are agents that mimic cocaine's pharmacologic effects (e.g., desipramine, bromocriptine) or not. We propose that individual rats are predisposed to specific hemodynamic response patterns evoked by behavioral stress that, in most examples cited here, occurs after administration of psychoactive agents.

Bromocriptine. Dopamine agonists, such as bromocriptine, are the most common class of agents used to treat cocaine addiction and toxicity (Halikas et al., 1993). Bromocriptine has been shown to reduce cocaine craving in humans (Dackis and Gold, 1985), and self-administration behavior and motor responses in rats (Campbell et al., 1989; Hubner and Koob, 1990) presumably by desensitizing dopamine receptors responsible for cocaine-induced euphoria. With regard to the autonomic nervous system, bromocriptine alone increased heart rate and pupillary diameter but lowered arterial pressure in human subjects (Preston et al., 1992). We noted a biphasic response (increase followed by a decrease) in arterial pressure after bromocriptine administration (0.1 and 1 mg/kg, i.v.) in conscious rats (fig. 2; table 2).

Desipramine. Tricyclic antidepressants have been shown to be effective in treating craving for cocaine (Tennant and Rawson, 1983) and cocaine toxicity (Antelman et al., 1981). For example, desipramine is widely used for treatment of addiction to cocaine (Halikas et al., 1993). Desipramine may reduce the stimulant properties of cocaine in some patients and enhance it in others (Fischman et al., 1990) suggesting that individual differences may alter the effectiveness of desipramine in treating cocaine addiction. The autonomic responses to both agents may also vary because both cocaine and desipramine produce an initial brief excitation of sympathetic activity in some conscious animals followed by a sustained inhibition of sympathetic activity in all subjects (Branch and Knuepfer, 1994b; Dorward et al., 1991; Knuepfer and Branch, 1992). As noted with cocaine, arterial pressure was elevated for at least 10 min after desipramine administration (1 or 10 mg/kg) despite the reported sympathoinhibition caused by both agents. Others have not observed a change in arterial pressure with desipramine administration in conscious rats (Tella et al., 1993), rabbits (Dorward et al., 1991) and humans (Kosten et al., 1992) although Fischman et al. (1990) reported an increase in arterial pressure in human subjects after chronic desipramine maintenance therapy. Therefore, the decrease in central sympathetic drive may not offset the enhanced catecholamine levels due to reuptake blockade.

GBR 12909. GBR 12909 is more selective and more potent in its ability to bind to the dopamine transporter compared to cocaine (Andersen, 1989; Izenwasser et al., 1990; Rothman et al., 1989) and inhibits cocaine-induced increases in extracellular dopamine (Rothman et al., 1991). It has been shown to produce similar behavioral responses to those elicited by cocaine in animal studies (Cunningham and Callahan, 1991; Heikkila and Manzino, 1984; Howell and Byrd, 1991). GBR 12909 substitutes for cocaine at least in some animals in drug discrimination studies (Johanson and Barrett, 1993). Because of its selectivity for the dopamine transporter and the known involvement of dopamine in eliciting cocaine-induced euphoria, GBR 12909 has been proposed as a treatment for cocaine addiction (Rothman and Glowa, 1995; Rothman et al., 1989, 1991).

Dopamine hypothesis. The actions of bromocriptine, desipramine and GBR 12909 give insight into the causes of hemodynamic responses to cocaine. Both bromocriptine and desipramine will enhance dopamine receptor activation by different mechanisms. Because these agents selectively reduce the decrease in cardiac output elicited by cocaine, it is possible that dopamine receptor activation alleviates the cardiodepression noted in some rats. Interestingly, GBR 12909 would be expected to have similar results but did not. It is possible that the dose of GBR 12909 was insufficient to attenuate the responses to cocaine, but, due to the prolonged effects of GBR 12909 higher doses were not used. Schindler et al. (1991) reported that the cocaine-induced pressor response was not antagonized by D1 or D2 receptor antagonists or mimicked by a D2 agonist in conscious squirrel monkeys. In our study, the agonists or uptake inhibitors could mimic the effects of cocaine to some extent (figs. 2 and 6) but only bromocriptine was effective in reducing the pressor response to cocaine. In contrast, all three agents were capable of preventing differential cardiac output and systemic vascular resistance responses in the two groups of animals. These data suggest that dopamine receptors may be responsible for the differences in hemodynamic response patterns noted in vascular and mixed responders.

Diazepam. Diazepam is effective in ameliorating stress-induced hormonal and neurochemical responses (Lahti and Barsuhn, 1975) that are similar to effects observed after cocaine administration (Levy et al., 1992; Moldow and Fischman, 1987; Rivier and Vale, 1987). Larger doses of diazepam are reported to reduce cocaine toxicity (Catravas and Waters, 1981; Derlet and Albertson, 1989; Guinn et al., 1980) although others have reported no significant protection (Trouvé and Nahas, 1990). The effectiveness of diazepam in reducing toxicity may only be manifest at greater doses (>1 mg/kg) when anticonvulsant effects are manifest whereas lower doses, such as those used in our study, may not be effective in protecting rats from lethal doses of cocaine (Smith et al., 1991). Diazepam is effective for treating toxicity in patients experiencing cocaine-induced seizures (Jonsson et al., 1983; Resnick and Resnick, 1984). In humans, sedative and antianxiety effects are noted at doses of 30 to 300 µg/kg (Rall, 1990). Although doses in humans cannot be directly compared to those in rats, the substantial difference (100-fold) between doses in rats and those in humans suggests that greater sedation is necessary to prevent toxicity in rats. Diazepam (0.1 or 0.5 mg/kg) produced biphasic arterial pressure responses (fig. 2; table 2). A small depressor response remaining noted only in vascular responders was not likely to substantially alter responses to cocaine administration. Therefore, these data suggest that diazepam may not alter cocaine-induced cardiovascular toxicity.

Buprenorphine. Buprenorphine is a mixed agonist/antagonist at the mu opioid receptor with potent analgesic effects and little or no potential for dependence (Cowan et al., 1977; Mello et al., 1989). Buprenorphine has been suggested as a treatment for cocaine addiction because it suppresses cocaine-induced responding in self-administration studies (Mello et al., 1989; Mendelson et al., 1990; Winger et al., 1992) and blocks cocaine-induced place preference (Suzuki et al., 1992). Little is known concerning the possible cardiovascular interactions between buprenorphine and cocaine. It was reported that arterial pressure and heart rate responses to cocaine and plasma cocaine levels were not altered by buprenorphine maintenance therapy in human subjects (Teoh et al., 1993). Buprenorphine, even at a low dose (0.3 mg/kg), reduced toxicity to lethal injections of cocaine in mice (Shukla et al., 1991; Witkin et al., 1991). At this dose, we noted net increases in arterial pressure, heart rate and cardiac output in vascular responders that were still present after 10 min (fig. 2; table 2). Despite these changes in hemodynamic variables, buprenorphine had little effect on the cardiovascular responses to cocaine with the exception of a possible increase in the systemic vascular resistance response. If the shift in baseline is not considered, buprenorphine pretreatment would blunt the initial pressor response and enhance both the decrease in cardiac output and heart rate (data not shown). Therefore, if buprenorphine-induced cardiovascular effects were allowed to resolve, it is possible that cocaine-induced cardiodepression might be greater.

MK-801. It has been reported that MK-801 (dizocilpine) and other NMDA receptor antagonists reduce toxicity to lethal doses of cocaine (Derlet and Albertson, 1990; Rockhold et al., 1991; Witkin and Tortella, 1991) although the precise mechanism by which this occurs remains to be determined. MK-801 also reduces the untoward proarrhythmic effects of cocaine on the myocardium (Hageman and Simor, 1993). Interestingly, MK-801 did not alter heart rate, arterial pressure or sympathetic nerve activity substantially in pentobarbital-anesthetized dogs (Hageman and Simor, 1993). In contrast, we and others (Lewis et al., 1989) noted that this dose of MK-801 elicited increases in arterial pressure and heart rate in conscious rats (fig. 2; table 2). Because higher doses produce more profound changes in cardiovascular parameters in conscious animals making interpretation more difficult, these were not used in our study. Although the changes evoked by MK-801 were reduced somewhat 10 min after pretreatment (table 2), it is possible that the shift in baseline heart rate may have contributed to the enhanced tachycardia followed by a reduced bradycardia after cocaine administration. In addition, an increase in aortic flow and in systemic vascular resistance was noted 10 min after MK-801 administration. Taking into account the higher arterial pressure, heart rate and cardiac output, the cocaine-induced responses were shifted upward in vascular responders. We suggest that the difference in baseline may be largely responsible for changes in hemodynamic responses to cocaine because no differences in the pressor or cardiac output responses are noted if the baseline shift is not taken into consideration (data not shown). It appears as though the increase in the cocaine-induced pressor response is due entirely to an increase in the cardiac output response with the likely contribution of heart rate to this response. These data suggest that this dose of MK-801 may exacerbate pressor and heart rate responses to cocaine possibly due to the change in baseline. These data also demonstrate that the pressor responses can be dissociated from the cardiac output responses. Interestingly, propranolol has the opposite effect; ameliorating the pressor response and enhancing the decrease in cardiac output (Branch and Knuepfer, 1994a).