Drug and Alcohol Dependence, 26 (1990) 9Elsevier Scientific Publishers Ireland Ltd.
Substitution
9
17
of psychoactive drugs in pentobarbitaldependent rats
G.J. Yutrzenkaa, G.A. Patrickb and W. Rosenbergerb “Department of Physiology and Pharmacology, University of South Dakota School Vermillion, SD and bDepartment of Pharmacology and Toxicology, Medical College University, Richmond VA IlJ.S.A.I
of Medicine, of Virginia,
414 East Clark Street. Virginia Commonwealth
(Received July 13th, 19891
The substitution of either bromazepam, diazepam, methaqualone, mazindol, nortriptyline or bupropion for pentobarbital, in dependent rats, was assessed using a continuous drug infusion method. Male, Sprague - Dawley rats were made dependent on pentobarbital during 12 days of continuous, intraperitoneal, pentobarbital infusion. On Day 13, pentobarbital was replaced with either saline, vehicle, or one of the drugs of interest and rats were infused for 24 h. On Day 14, all rats were infused, for 24 h. with saline. Changes in both body weight and behavioral indices of withdrawal were assessed during Day 13 and 14. It was observed that bromazepam and methaqualone substituted for pentobarbital in a dose-dependent fashion. Diazepam also substituted in pentobarbital dependent rats but, inexplicably, the low dose of diazepam provided better substitution than did the higher dose. On the other hand, neither mazindol or nortriptyline substituted for pentobarbital and there was a tendency for exacerbation of the’withdrawal signs. Finally, it was noted that the low dose of bupropion appeared to decrease the severity of the withdrawal symptoms. The data supports the view that the substitution of compounds for pentobarbital, in dependent rats, is limited to those compounds which, presumably, possess similar mechanisms of action in the CNS. Key words: substitution; dependence; rats; barbiturates;
depressants;
Introduction Cross-dependence studies have been shown to be one means by which to investigate the abuse liability of a drug [1,2]. The cross-dependence paradigm arises from the investigations of Himmelsbac [3], in which it was suggested that drugs which could suppress the withdrawal syndrome when substituted for another drug, were also liable to produce a dependence similar to that of the drug for which they were substituted. Substitution studies were first carried out using opiate drugs [3] and have subsequently been extended to include a variety of CNS depressant agents [4-g]. It has been demonstrated that those compounds which have a similar mechanism of action are most likely to Correspondence
to: G.J. Yutrzenka.
stimulants; antidepressants
substitute for each other [7,10]. However, it has been determined that not all compounds within a class are equally capable of substituting and maintaining the physical dependence state [7,10,11]. In addition, it has been determined that simply producing sedation, as may occur with a variety of neuroleptic agents, is not sufficient to totally suppress the typical withdrawal syndrome [4,7,10]. The current study extends previous investigations of the cross-dependence phenomena associated with the production of physical dependence on pentobarbital (PB) [9]. This study investigated the ability of two benzodiazepines (diazepam and bromazepam), a nonbarbiturate, sedative-hypnotic (methaqualone), a CNS stimulant (mazindol) and two antidepressant compounds (bupropion and nortriptyline) to substitute for PB in the dependent rat.
0376.8716/90/$03.50 0 1990 Elsevier Scientific Publishers Ireland Ltd. Printed and Published in Ireland
10
Methods Animals Male, Sprague - Dawley rats (Dominion Labs, Dublin, VA), weighing 175-200 g, were used in all drug infusion studies while male CF1 mice (Dominion Labs, Dublin, VA) weighing 25-30 g, were used for potency estimation experiments. Rats were housed, individually, in stainless steel cages while mice were housed, in groups of 10, in plastic cages fitted with wire tops. Food and water were available ad libitum and all animals were acclimated to the facility for several days prior to use in the studies. Drugs Pentobarbital sodium was dissolved in 0.9% saline. Bupropion, nortriptyline and mazindol water. distilled were dissolved in Methaqualone was dissolved in a vehicle consisting of distilled water/propylene glycoll ethanol (70:20:101. Diazepam and bromazepam were dissolved in a vehicle comprised of distilled water/propylene glycol/ethanol (50:40:101 at pH 2.0. The amount of propylene glycol and ethanol used was the minimum which would insure that the compounds would stay in solution for the required 24 h infusion period. In preliminary studies it was noted that there were no readily apparent toxic effects nor was there evidence of gross anatomical abnormalities of either abdominal organs or the peritoneal wall in rats receiving 8 ml of vehicle, per day, for 5 consecutive days. Compounds to be tested were obtained from the Committee on Problems of Drug Dependence (CPDD) and were provided in coded vials. Investigators were blind to the identity of the compound until after completion of the study. Procedure Rats were prepared with an intraperitoneal cannula (PE 90) while under methoxyflurane anesthesia [12]. Rats were allowed several days for recovery and were then placed into an infusion harness and were continuously infused with 0.9% saline during a 3-day acclimation period. For the next 12 days, rats received a continuous infusion of either 0.9% saline (con-
trol) or escalating doses of pentobarbital sodium reaching a final dose of 950 mglkg per 24 h. Eight milliliters of PB solution was infused during each 24-h period. Rats were infused with PB at an initial dose 100 mg/kg per 24 h. The infused dose was adjusted, daily, according to the degree of CNS depression exhibited by the individual rats. Rats were evaluated for severity of CNS depression using a 12 point rating scale previously described [12]. In general, during Day 1 to Day 7 PB doses were increased by 100 mg/kg for rats exhibiting a depression score of 4 or less, by 50 mg/kg for rats with a score of 5- 8 and no change from the previous dose for rats with a score of 9 or more. During Day 8 to Day 12 daily doses of PB were increased by 50 mgl kg for rats with a score of 8 or less and no increase in dose for rats with a score of 9 or more. By Day 12 almost all rats were receiving PB at a dose of 950 mglkg per 24 h. Use of this dosing procedure allows for adequate dosing to overcome tolerance while keeping druginduced mortality to a minimum. Body weight and water consumption were monitored daily. On Day 13 there began a 24-h drug substitution period during which either saline, vehicle, or the drug of interest was substitued for PB (substitution phase). This was followed, on Day 14, by a 24 h infusion of 0.9% saline in all rats (withdrawal phase). During both Day 13 and Day 14 withdrawal signs were assessed. The intensity of the withdrawal symptoms were scored by two observers, who were blind to drug treatment, using a withdrawal symptomatology rating scale [9]. Body weight and water consumption were determined at 0, 8 and 24 h of both days. Potency estimation studies Potency estimation studies were carried out in mice so as to estimate the appropriate dose of the drug to be substituted for PB. The potency estimation was conducted using both the inverted screen test [13] as well as by monitoring alteration of spontaneous locomotor were curves response activity. Dose determined using at least three doses of each
11
bisected a plastic cage containing two mice. Movement of the mice disrupted the beam and a “count” of activity was recorded. Following drug administration, spontaneous locomotor activity was recorded at the following time intervals: 5- 15 min, 35- 50 min, 65-95 min and 125- 185 min. The ED, dose was determined to be that dose which reduced spontaneous locomotor activity to one-half that recorded for concurrently tested vehicletreated, control mice. Potency ratios of substi-
drug with six mice per dose. Vehicle treated mice were assessed concurrently with drug treated mice. The inverted screen test was conducted at 15, 30, 60 and 120 min following drug administration. The ED, dose, which was determined to be the dose at which one-half of the treated mice failed to right themselves within the 60-s test period, was computed for each time period. Spontaneous locomotor activity was determined using a single beam photocell which
Table I. Estimation Treatment
of ED,
and potency
Inverted
ratio of compounds
screen
ImU
30
15
17.8 (14.1-
22.4)
24.5 (20.7-
29.0)
ED, PRb Bromazepam
ED, PR Methaqualone ED50 PR Treatment
14.9 (9.9 - 22.6) 1.19 and 1.64 Spontaneous
to pentobarbital.
test
Time of test after treatment
Pentobarbital ED; Diazepam
relative
locomotor
16.3 (10.7 - 24.8) 1.09 and 1.5
120
-
-
0.82 (0.31-2.2) 21.6 and 29.8
1.3 (0.63-2.7) 13.6 and 18.8
0.52 (0.22- 1.2) 34.2 and 47.1
0.79 (0.28- 2.2) 22.4 and 30.9
20.9 (10.9 - 39.9) 0.85 and 1.17
activity
Time of test after treatment 5-15
60
Imid
35-50
125-185
65-95
Pentobarbital ED, Diazepam
24.7 (20.6-
29.7)
-
ED, PR’ Bromazepam
2.5 (l.l10
ED, PR Methaqualone
0.55 (0.02845.2
ED, PR Nortriptyline
4.97 (1.395.0
ED, PR
18.3 (2.91.35
17.8)
11.6)
a ED,, in mg/kg, and 95% confidence intervals. b PR = Potency ratio relative to pentobarbital c PR = Potency ratio relative to pentobarbital
9.64 (4.552.6
20.4)
4.2 (0.62-27.9) 5.9
at 15 and 30 min. at 5- 15 min.
5.4)
10.7)
1.5 (0.4816.6
4.6)
12
tuted drugs, relative to PB, were determined at time of peak activity and when, in addition, the vehicle effect was no longer evident. Statistical analysis Withdrawal scores were analyzed for significant difference using the Mann-Whitney U-test [14]. ED, values and 95% confidence intervals were determined by the method of Litchfield and Wilcoxon [15]. Changes in body weight and 24-h water consumption were analyzed by Student’s t-test [14]. Results Potency estimation The potency ratios of the substituted compounds are shown in Table I. Both of the benzoA Mean withdrawal scores of control rats and PB Fig. 2. dependent rats during (A) substitution with either saline, vehicle or bromazepam in doses of 28 mg/hg per 24 h (BRZ (28)) or 14 mg/kg per 24 h (BRZ (14)) and (B) substitution with saline. Each point represents the mean of four-six rats. Filled symbols indicate statistically significant differences (P 6 0.05) as compared to vehicle substitution in PB-dependent rats.
Time (ho”rS)
Mean withdrawal scores of control rats and PB Fig. 1. dependent rats during (A) substitution with either saline, vehicle or PB, in a dose of 950 mg/kg per 24 h (PB (950)) and (B) substitution with saline. Each point represents the mean of four-six rats. Filled symbols indicate statistically significant differences (P d 0.05) as compared to saline substitution in PB-dependent rats.
compounds exhibited markedly diazepine increased potency and duration of effect relative to PB. Bromazepam appeared to be more potent and longer acting as compared to diazepam. Methaqualone was determined to be as potent as PB on the inverted screen test and about five times more potent than PB on suppression of locomotor activity. Estimation of potency of the benzodiazepines and methaqualone was complicated at the early time points (i.e., 15 and 30 min) by the effects of the alcoholic vehicle. Therefore, potency estimations were determined at times of peak activity of the drug and when, in addition, the vehicle effect was no longer evident. On the basis of these potency studies the substituted drugs were infused in two doses with the higher dose being that dose which was
13
B
3
1
f----
Substitution studies In all experiments, control and PB-infused rats did not differ significantly with respect to either body weight or water consumption during the 12-day infusion period (data not shown). As shown in Fig. lA, control rats exhibited few signs of withdrawal on Day 13. Likewise those PB-dependent rats which continued to receive PB on Day 13 also showed few behavioral signs of withdrawal although there was elevation in withdrawal scores at 10, 12 and 24 h of this period. As compared to rats which continued to receive PB, substitution of saline into PBdependent rats resulted in significantly increased withdrawal scores by 4 h following the start of the substitution period. In addition, substitution of an ethanol/propylene glycol vehicle for PB resulted in withdrawal scores which were similar to those seen following saline substitution.
Mean withdrawal scores of control rats and PBFig. 3. dependent rats during (A) substitution with either saline, vehicle or diazepam in doses of 40 mg/kg per 24 h (DZ (40)) or 20 mg/kg per 24 h (DZ (20)) and (B) substitution with saline. Each point represents the mean of three-six rats. Filled symbols indicate statistically significant differences (P G 0.05) as compared to vehicle substitution in PB-dependent rats.
estimated to be equipotent to 950 mg/kg of PB and the second dose being one-half of the estimated equipotent dose. This was done to take into account the possibility that accumulation of the equipotent dose of the substituted drug, over a 24-h infusion period, might lead to toxicity. Neither nortriptyline, mazindol, nor bupropion produced any significant effect on the inverted screen test at doses that were below lethal doses (data not shown). Mice treated with nortriptyline exhibited a slight reduction of spontaneous locomotor activity. Bupropion and mazindol both produced a stimulation of locomotor activity (data not shown). Thus it was not possible to determine potency ratio for these compounds relative to PB.
Fig. 4. Mean withdrawal scores of control rats and PBdependent rats during (A) substitution with either saline, vehicle or methaqualone in doses of either 200 mg/kg per 24 h (MTQ (200)) or 100 mg/kg per 24 h (MTQ (100)) and (B) substitution with saline. Each point represents the mean of four or five rats. Filled symbols indicate statistically significant difference Cp < 0.05) as compared to vehicle substitution in PB-dependent rats.
14
On day 14, all rats received saline infusion and withdrawal scores were recorded during the succeeding 24-h period (Fig. 1Bl. The withdrawal scores of PB-dependent rats were observed to be elevated relative to scores recorded for saline-infused control rats. The greatest degree of withdrawal was noted for those rats which had been maintained on PB on Day 13. It was also observed that those rats which had begun to exhibit withdrawal on day 13 continued to show signs of withdrawal on Day 14. Bromazepam provided dose-dependent substitution for PB in dependent rats (Fig. 2Al. It was noted that the higher dose of bromazepam substituted completely for PB while the lower dose provided no substitution until the tenth hour of this period. This may be due to accumu-
Mean withdrawal scores of control rats and PBFig. 5. dependent rats during (A) substitution with either saline or bupropion in doses of 300 mg/kg per 24 h (BUP (300)) or 150 mg/kg per 24 h (BUP (150)) and (B) during saline substitution. Each point represents the mean of three-five rats. Filled symbols indicate statistically significant differences (P < 0.05) as compared to vehicle substitution in PB-dependent rats.
lation of bromazepam during this infusion period. On Day 14 there was noted a significant increase in withdrawal scores for rats which had received bromazepam on Day 13 as compared to PB-dependent rats receiving vehicle on Day 13 (Fig. 2B). Diazepam was also noted to substitute for PB (Fig. 3Al. Inexplicably, the lower dose of diazepam substituted to a greater degree than did the higher dose. Again, saline substitution on Day 14 produced elevated withdrawal scores in rats previously infused with diazepam (Fig. 3Bl. In PB-dependent rats, methaqualone at a dose of 200 mg/kg per 24 h provided complete substitution (Fig. 4Al. Interestingly, this effect was carried over into the next day when saline was substituted for methaqualone (Fig. 4Bl. It is hypothesized that the accumulation of methaqualone may have been sufficient to prevent emergence of withdrawal signs on Day 14. Dependent rats infused with the lower dose of methaqualone exhibited withdrawal symptoms which were similar in degree to that observed for dependent rats infused with vehicle. Not unexpectedly, neither nortriptyline nor mazindol were found to substitute for PB. In fact, mazindol produced a substantial exacerbation of the withdrawal signs. Interestingly, the lower dose of bupropion appeared to partially ameliorate the behavioral signs of withdrawal (Fig. 5A,5Bl. Loss of body weight is a consistent finding during withdrawal from barbiturates 19,121.PBdependent rats which were subsequently infused with either saline or vehicle were noted to exhibit an approximate 10% loss of body weight by 24 h of the substitution period (Fig. 6A). The body weight was noted to be returning toward control values by the end of Day 14. Rats continuing to receive PB on Day 13 demonstrated no significant alteration in body weight on Day 13 but did demonstrate substantial weight loss following saline substitution on Day 14. A fluctuation in the body weight of control rats was noted over the course of the 2 days of observation and is presumed to be reflective of a circadian pattern of body weight regulation.
15
Percent change in body weight of control and PBFig. 7. dependent rats during both the drug substitution period (Day 13) and the subsequent saline substitution period (Day 14) (A). Either saline, vehicle or methaqualone in doses of 200 mg/kg per 24 h (MTQ (200)) or 100 mg/kg per 24 h (MTQ (100)) was substituted into dependent rats. (B) Either saline or bupropion in doses of 300 mgkg per 24 h (BUP (300)) or 150 mg/hg per 24 h (BUP 050)) was substituted into dependent rats. Each point represents the mean 2 S.E.M. of the percent change in body weight as compared to time zero. Filled symbols indicate statistically significant differences (P < 0.05) as compared to vehicle substitution in PB-dependent rats. The arrow indicates the start of saline infusion on Day 14. Percent change in body weight of control and PBFig. 6. dependent rats during both the substitution phase (Day 13) and the withdrawal phase (Day 14). (A) PB-dependent rats were infused, for 24 h, with either saline, alcoholic vehicle, or PB, 950 mg/lcg per 24 h (PB (950)) followed by saline infusion for 24 h. (B) PB-dependent rats were infused, for 24 h, with bromazepam in doses of either 28 mg/kg per 24 h (BRZ (28)) or 14 mg/hg per 24 h (BRZ (14)) followed by saline infusion for 24 h. (Cl PB-dependent rats were infused for 24 h, with diazepam in doses of either 40 mg/kg per 24 h (DZ (40)) or 20 mg/hg per 24 h (DZ (20)) followed by saline infusion for 24 h. Control rats received saline during both the substitution and withdrawal phases. Each point is the mean + S.E.M. of the percent change in body weight, as compared to time zero, for three-six rats. Filled symbols indicate statistically signfficant differences (P 6 0.05) as compared to vehicle substitution in PB-dependent rats. The arrow indicates the start of saline infusion on Day 14.
Substitution of bromazepam or diazepam into PB-dependent rats effectively prevented a significant loss of body weight (Fig. 6B, 6C). Substitution of saline into these rats on Day 14 resulted in significant declines in body weight at 24 h of this period. Methaqualone substitution also provided dose-dependent attenuation of the typical decrease in body weight observed during PB withdrawal (Fig. 7A). Substitution with either bupropion (Fig. 7B) or mazindol (data not shown) on Day 13 resulted in exacerbation of weight loss observed in PBdependent rats. The loss of body weight was
16
sustained through at least the end of Day 14. When nortriptyline was substituted for PB it was observed that the higher dose of this drug ameliorated the weight loss of dependent rats while the lower dose of nortriptyline exacerbated the weight loss (data not shown). Again, those effects on body weight were carried through into Day 14. Discussion Cross-dependence studies have been shown to be advantageous when used to help predict the dependence liability of compounds as well as to explore similarities among compounds with respect to mechanisms which may underlie the development of physical dependence [1,2,3,5]. It is generally hypothesized that those compounds which substitute for another compound may possess both a mechanism of action and a propensity for development of physical dependence which is similar to that of the first compound. On the other hand, those compounds which fail to substitute are not as likely to produce the same spectrum of dependence [1,3,7]. Previous investigations have demonstrated the utility of the continuous infusion method in the development of a rodent model of dependence on opiates [16] as well as on CNS depressants [12,17]. In addition, a previous investigation has demonstrated the use of this model in the substitution of the benzodiazepines, temazepam and midazolam, in PB-dependent rats [9]. The current investigation extends these studies to include two additional benzodiazepines (diazepam, bromazepaml, a non-barbiturate, sedative-hypnotic (methaqualonel, a CNS stimulant (mazindoll and two antidepressant agents (nortriptyline, bupropion). It has been demonstrated that the administration of bromazepam, diazepam or methaqalone were all effective in substituting for PB and acted to maintain the dependent state. This was evidenced by both the amelioration of behavioral signs of withdrawal (Figs. 2, 3, 41 as well by the reduction in the characteristic loss of body weight observed during the withdrawal
period (Figs. 6BA, 6C, 7Al. The fact that the dependent state was being maintained was demonstrated by the substitution of saline for the drug on Day 14. Those rats in which the physical dependence had been maintained by the substituted drug were, following saline substitution, observed to exhibit both an increased withdrawal scores as well as a substantial decrease in body weight. Taken together, this provides evidence that the substitution of these drugs did indeed mantain the physical dependence state. Neither mazindol or nortriptyline were effective in substituting for PB in the dependent rats. This was evidenced by both increased withdrawal scores and a substantial loss of body weight. The situation with bupropion substitution was not as clear. In this case, the lower dose of bupropion appeared to substitute somewhat as evidenced by lower withdrawal scores (Fig. 51. However this dose of bupropion was seemingly ineffective in ameliorating the loss of body weight (Fig. 7Bl. The higher dose of bupropion, on the other hand, produced both increased withdrawal scores and a substantial loss of body weight. It was concluded that bupropion did not substitute for PB. The somewhat differing results obtained with bupropion as compared to nortriptyline might be explained, at least in part, in that bupropion action seems to be directed at dopaminergic neurons whereas nortriptyline action is primarily directed towards inhibition of reuptake of serotonin and/or norepinephrine [18,19]. The results obtained in this study are not unexpected and are consistent with the basic premise which underlies the cross-dependence paradigm. The benzodiazepines, methaqualone and PB have all been postulated to act, at least in part, at a GABA/benzodiazepine/barbiturate receptor complex [20,21,22]. The typical pharmacological effects of these agents are similar in many respects and include sedation, hypnosis and anticonvulsant activity [23]. On the other hand, mazindol, bupropion and nortriptyline are presumed to possess actions directed towards alteration of reuptake and/or
17
storage of biogenic amines in the neuron [24,25]. The current investigation serves to demonstrate the selectivity of the cross-dependence paradigm by underscoring the inability of CNS stimulants and antidepressants to maintain the physical dependence established towards a CNS depressant agent. In addition the study extends the results of previous investigations which describe the utility of the cross-dependence paradigm in studies serving to elucidate the phenomena which underlie physical dependence. Acknowledgements This study was supported, in part, by USPHS grants DA-00490, DA-07027 and DA04244. Bromazepam, bupropion, diazepam, mazindol. methaqualone and nortriptyline were provided by Dr. Arthur Jacobson, Drug Testing Committee, Committee on Problems of Drug Dependence. We express our appreciation to Ms. Shellie Delaney and Ms. Raelene Gilbert for their help in preparation of this manuscript. References 1
J.L. Katz and S.R. Goldberg, Agents Actions, 23 (19881 18.
2
J.H. Woods, J.L Katz and G. Winger., Pharmacol. Rev., 39 (1987) 251. C.K. Himmelsbach, J. Pharmacol. Exp. Ther., 71 (1941) 41. J.K. Belknap, Clin. Toxicol., 12 (1978) 427. B.E. Jones, J.A. Prada and W.R. Martin, Psychopharmacology, 47 (197617.
3 4 5
6 7 8 9 10
11 12 13 14 15 16 17 18 19 20 21 22 23
24
25
P.R.E. Norton, J. Pharm. Pharmacol., 22 (1970) 763, E. Tagashira, T. Urano, T. Hiramori and S. Yanaura, Jpn. J. Pharmacol., 33 (1983) 659. T. Yanagita and S. Takahashi, J. Pharmacol. Exp. Ther., 185 (1973) 303. G.J. Yutrzenka, G.A. Patrick and W. Rosenberger, Physiol. Behav., 46 (1989) 55. T. Yanagita, Handbook of Experimental Pharmacology, Vol. 55. In: F. Hoffmeister (Ed.), Springer-Verlag. New York, 1981, pp. 395. W.R. Martin, L.F. MeNicholas and S. Cherian, Life Sci., 3lf1982) 721. G.J. Yutrzenka, G.A. Patrick and W. Rosenberger, J. Phamacol. Exp. Ther., 232 (1985) 111. L.L. Coughenour, J.R. McLean and R.B. Parker, Pharmacol. Biochem. Behav., 6 (1977) 351. R.H. Kolstoe, Introduction to Statistics for the Behavioral Sciences, Dorsey Press, Homewood, IL, 1973. J.T. Litchfield and F. Wilcoxon, J. Pharmacol. Exp. Ther., 96 (1949) 99. D.G. Teiger, J. Pharmacol. Exp. Ther., 190 (1974) 408. G.J. Yutrzenka, Pharmacol. Biochem. Behav., 34 (1989) 49. L.E. Hollister, Rational Drug Ther., 16 (1982) 1. M.V. Rudorfer, R.N. Golden and W.Z. Potter, Psychiat. Clin. N. Am., 7 (1984) 519. R.W. Olsen, J. Neurochem., 37 (1981) 1. S.R. Naik, P.R. Naik and U.K. Sheth, Psychopharmacology, 57 (1978) 103. T. Suzuki, Y. Koike, Y. Chida and M. Misawa, Jpn. J. Pharmacol., 46 (1988) 403. S.C. Harvey, The Pharmacological Basis of Therapeutics, 7th edn. A.G. Gilman, L.S. Goodman, T.W. Rail and F. Murad (Eds.), MacMillan Publishing Co., New York, 1985, p. 339. R.J. Baldessarini, The Pharmacological Basis of Therapeutics. 7th edn. A.G. Gilman, L.S. Goodman, T.W. Rall and F. Murad (Eds.), MacMillan Publishing Co., New York, 1985, p. 412. A. Goth, Medical Pharmacology: Principles and Concepts. 11th edn., C.V. Mosby Co., St. Louis, 1984, p. 175.
Substitution
9
17
of psychoactive drugs in pentobarbitaldependent rats
G.J. Yutrzenkaa, G.A. Patrickb and W. Rosenbergerb “Department of Physiology and Pharmacology, University of South Dakota School Vermillion, SD and bDepartment of Pharmacology and Toxicology, Medical College University, Richmond VA IlJ.S.A.I
of Medicine, of Virginia,
414 East Clark Street. Virginia Commonwealth
(Received July 13th, 19891
The substitution of either bromazepam, diazepam, methaqualone, mazindol, nortriptyline or bupropion for pentobarbital, in dependent rats, was assessed using a continuous drug infusion method. Male, Sprague - Dawley rats were made dependent on pentobarbital during 12 days of continuous, intraperitoneal, pentobarbital infusion. On Day 13, pentobarbital was replaced with either saline, vehicle, or one of the drugs of interest and rats were infused for 24 h. On Day 14, all rats were infused, for 24 h. with saline. Changes in both body weight and behavioral indices of withdrawal were assessed during Day 13 and 14. It was observed that bromazepam and methaqualone substituted for pentobarbital in a dose-dependent fashion. Diazepam also substituted in pentobarbital dependent rats but, inexplicably, the low dose of diazepam provided better substitution than did the higher dose. On the other hand, neither mazindol or nortriptyline substituted for pentobarbital and there was a tendency for exacerbation of the’withdrawal signs. Finally, it was noted that the low dose of bupropion appeared to decrease the severity of the withdrawal symptoms. The data supports the view that the substitution of compounds for pentobarbital, in dependent rats, is limited to those compounds which, presumably, possess similar mechanisms of action in the CNS. Key words: substitution; dependence; rats; barbiturates;
depressants;
Introduction Cross-dependence studies have been shown to be one means by which to investigate the abuse liability of a drug [1,2]. The cross-dependence paradigm arises from the investigations of Himmelsbac [3], in which it was suggested that drugs which could suppress the withdrawal syndrome when substituted for another drug, were also liable to produce a dependence similar to that of the drug for which they were substituted. Substitution studies were first carried out using opiate drugs [3] and have subsequently been extended to include a variety of CNS depressant agents [4-g]. It has been demonstrated that those compounds which have a similar mechanism of action are most likely to Correspondence
to: G.J. Yutrzenka.
stimulants; antidepressants
substitute for each other [7,10]. However, it has been determined that not all compounds within a class are equally capable of substituting and maintaining the physical dependence state [7,10,11]. In addition, it has been determined that simply producing sedation, as may occur with a variety of neuroleptic agents, is not sufficient to totally suppress the typical withdrawal syndrome [4,7,10]. The current study extends previous investigations of the cross-dependence phenomena associated with the production of physical dependence on pentobarbital (PB) [9]. This study investigated the ability of two benzodiazepines (diazepam and bromazepam), a nonbarbiturate, sedative-hypnotic (methaqualone), a CNS stimulant (mazindol) and two antidepressant compounds (bupropion and nortriptyline) to substitute for PB in the dependent rat.
0376.8716/90/$03.50 0 1990 Elsevier Scientific Publishers Ireland Ltd. Printed and Published in Ireland
10
Methods Animals Male, Sprague - Dawley rats (Dominion Labs, Dublin, VA), weighing 175-200 g, were used in all drug infusion studies while male CF1 mice (Dominion Labs, Dublin, VA) weighing 25-30 g, were used for potency estimation experiments. Rats were housed, individually, in stainless steel cages while mice were housed, in groups of 10, in plastic cages fitted with wire tops. Food and water were available ad libitum and all animals were acclimated to the facility for several days prior to use in the studies. Drugs Pentobarbital sodium was dissolved in 0.9% saline. Bupropion, nortriptyline and mazindol water. distilled were dissolved in Methaqualone was dissolved in a vehicle consisting of distilled water/propylene glycoll ethanol (70:20:101. Diazepam and bromazepam were dissolved in a vehicle comprised of distilled water/propylene glycol/ethanol (50:40:101 at pH 2.0. The amount of propylene glycol and ethanol used was the minimum which would insure that the compounds would stay in solution for the required 24 h infusion period. In preliminary studies it was noted that there were no readily apparent toxic effects nor was there evidence of gross anatomical abnormalities of either abdominal organs or the peritoneal wall in rats receiving 8 ml of vehicle, per day, for 5 consecutive days. Compounds to be tested were obtained from the Committee on Problems of Drug Dependence (CPDD) and were provided in coded vials. Investigators were blind to the identity of the compound until after completion of the study. Procedure Rats were prepared with an intraperitoneal cannula (PE 90) while under methoxyflurane anesthesia [12]. Rats were allowed several days for recovery and were then placed into an infusion harness and were continuously infused with 0.9% saline during a 3-day acclimation period. For the next 12 days, rats received a continuous infusion of either 0.9% saline (con-
trol) or escalating doses of pentobarbital sodium reaching a final dose of 950 mglkg per 24 h. Eight milliliters of PB solution was infused during each 24-h period. Rats were infused with PB at an initial dose 100 mg/kg per 24 h. The infused dose was adjusted, daily, according to the degree of CNS depression exhibited by the individual rats. Rats were evaluated for severity of CNS depression using a 12 point rating scale previously described [12]. In general, during Day 1 to Day 7 PB doses were increased by 100 mg/kg for rats exhibiting a depression score of 4 or less, by 50 mg/kg for rats with a score of 5- 8 and no change from the previous dose for rats with a score of 9 or more. During Day 8 to Day 12 daily doses of PB were increased by 50 mgl kg for rats with a score of 8 or less and no increase in dose for rats with a score of 9 or more. By Day 12 almost all rats were receiving PB at a dose of 950 mglkg per 24 h. Use of this dosing procedure allows for adequate dosing to overcome tolerance while keeping druginduced mortality to a minimum. Body weight and water consumption were monitored daily. On Day 13 there began a 24-h drug substitution period during which either saline, vehicle, or the drug of interest was substitued for PB (substitution phase). This was followed, on Day 14, by a 24 h infusion of 0.9% saline in all rats (withdrawal phase). During both Day 13 and Day 14 withdrawal signs were assessed. The intensity of the withdrawal symptoms were scored by two observers, who were blind to drug treatment, using a withdrawal symptomatology rating scale [9]. Body weight and water consumption were determined at 0, 8 and 24 h of both days. Potency estimation studies Potency estimation studies were carried out in mice so as to estimate the appropriate dose of the drug to be substituted for PB. The potency estimation was conducted using both the inverted screen test [13] as well as by monitoring alteration of spontaneous locomotor were curves response activity. Dose determined using at least three doses of each
11
bisected a plastic cage containing two mice. Movement of the mice disrupted the beam and a “count” of activity was recorded. Following drug administration, spontaneous locomotor activity was recorded at the following time intervals: 5- 15 min, 35- 50 min, 65-95 min and 125- 185 min. The ED, dose was determined to be that dose which reduced spontaneous locomotor activity to one-half that recorded for concurrently tested vehicletreated, control mice. Potency ratios of substi-
drug with six mice per dose. Vehicle treated mice were assessed concurrently with drug treated mice. The inverted screen test was conducted at 15, 30, 60 and 120 min following drug administration. The ED, dose, which was determined to be the dose at which one-half of the treated mice failed to right themselves within the 60-s test period, was computed for each time period. Spontaneous locomotor activity was determined using a single beam photocell which
Table I. Estimation Treatment
of ED,
and potency
Inverted
ratio of compounds
screen
ImU
30
15
17.8 (14.1-
22.4)
24.5 (20.7-
29.0)
ED, PRb Bromazepam
ED, PR Methaqualone ED50 PR Treatment
14.9 (9.9 - 22.6) 1.19 and 1.64 Spontaneous
to pentobarbital.
test
Time of test after treatment
Pentobarbital ED; Diazepam
relative
locomotor
16.3 (10.7 - 24.8) 1.09 and 1.5
120
-
-
0.82 (0.31-2.2) 21.6 and 29.8
1.3 (0.63-2.7) 13.6 and 18.8
0.52 (0.22- 1.2) 34.2 and 47.1
0.79 (0.28- 2.2) 22.4 and 30.9
20.9 (10.9 - 39.9) 0.85 and 1.17
activity
Time of test after treatment 5-15
60
Imid
35-50
125-185
65-95
Pentobarbital ED, Diazepam
24.7 (20.6-
29.7)
-
ED, PR’ Bromazepam
2.5 (l.l10
ED, PR Methaqualone
0.55 (0.02845.2
ED, PR Nortriptyline
4.97 (1.395.0
ED, PR
18.3 (2.91.35
17.8)
11.6)
a ED,, in mg/kg, and 95% confidence intervals. b PR = Potency ratio relative to pentobarbital c PR = Potency ratio relative to pentobarbital
9.64 (4.552.6
20.4)
4.2 (0.62-27.9) 5.9
at 15 and 30 min. at 5- 15 min.
5.4)
10.7)
1.5 (0.4816.6
4.6)
12
tuted drugs, relative to PB, were determined at time of peak activity and when, in addition, the vehicle effect was no longer evident. Statistical analysis Withdrawal scores were analyzed for significant difference using the Mann-Whitney U-test [14]. ED, values and 95% confidence intervals were determined by the method of Litchfield and Wilcoxon [15]. Changes in body weight and 24-h water consumption were analyzed by Student’s t-test [14]. Results Potency estimation The potency ratios of the substituted compounds are shown in Table I. Both of the benzoA Mean withdrawal scores of control rats and PB Fig. 2. dependent rats during (A) substitution with either saline, vehicle or bromazepam in doses of 28 mg/hg per 24 h (BRZ (28)) or 14 mg/kg per 24 h (BRZ (14)) and (B) substitution with saline. Each point represents the mean of four-six rats. Filled symbols indicate statistically significant differences (P 6 0.05) as compared to vehicle substitution in PB-dependent rats.
Time (ho”rS)
Mean withdrawal scores of control rats and PB Fig. 1. dependent rats during (A) substitution with either saline, vehicle or PB, in a dose of 950 mg/kg per 24 h (PB (950)) and (B) substitution with saline. Each point represents the mean of four-six rats. Filled symbols indicate statistically significant differences (P d 0.05) as compared to saline substitution in PB-dependent rats.
compounds exhibited markedly diazepine increased potency and duration of effect relative to PB. Bromazepam appeared to be more potent and longer acting as compared to diazepam. Methaqualone was determined to be as potent as PB on the inverted screen test and about five times more potent than PB on suppression of locomotor activity. Estimation of potency of the benzodiazepines and methaqualone was complicated at the early time points (i.e., 15 and 30 min) by the effects of the alcoholic vehicle. Therefore, potency estimations were determined at times of peak activity of the drug and when, in addition, the vehicle effect was no longer evident. On the basis of these potency studies the substituted drugs were infused in two doses with the higher dose being that dose which was
13
B
3
1
f----
Substitution studies In all experiments, control and PB-infused rats did not differ significantly with respect to either body weight or water consumption during the 12-day infusion period (data not shown). As shown in Fig. lA, control rats exhibited few signs of withdrawal on Day 13. Likewise those PB-dependent rats which continued to receive PB on Day 13 also showed few behavioral signs of withdrawal although there was elevation in withdrawal scores at 10, 12 and 24 h of this period. As compared to rats which continued to receive PB, substitution of saline into PBdependent rats resulted in significantly increased withdrawal scores by 4 h following the start of the substitution period. In addition, substitution of an ethanol/propylene glycol vehicle for PB resulted in withdrawal scores which were similar to those seen following saline substitution.
Mean withdrawal scores of control rats and PBFig. 3. dependent rats during (A) substitution with either saline, vehicle or diazepam in doses of 40 mg/kg per 24 h (DZ (40)) or 20 mg/kg per 24 h (DZ (20)) and (B) substitution with saline. Each point represents the mean of three-six rats. Filled symbols indicate statistically significant differences (P G 0.05) as compared to vehicle substitution in PB-dependent rats.
estimated to be equipotent to 950 mg/kg of PB and the second dose being one-half of the estimated equipotent dose. This was done to take into account the possibility that accumulation of the equipotent dose of the substituted drug, over a 24-h infusion period, might lead to toxicity. Neither nortriptyline, mazindol, nor bupropion produced any significant effect on the inverted screen test at doses that were below lethal doses (data not shown). Mice treated with nortriptyline exhibited a slight reduction of spontaneous locomotor activity. Bupropion and mazindol both produced a stimulation of locomotor activity (data not shown). Thus it was not possible to determine potency ratio for these compounds relative to PB.
Fig. 4. Mean withdrawal scores of control rats and PBdependent rats during (A) substitution with either saline, vehicle or methaqualone in doses of either 200 mg/kg per 24 h (MTQ (200)) or 100 mg/kg per 24 h (MTQ (100)) and (B) substitution with saline. Each point represents the mean of four or five rats. Filled symbols indicate statistically significant difference Cp < 0.05) as compared to vehicle substitution in PB-dependent rats.
14
On day 14, all rats received saline infusion and withdrawal scores were recorded during the succeeding 24-h period (Fig. 1Bl. The withdrawal scores of PB-dependent rats were observed to be elevated relative to scores recorded for saline-infused control rats. The greatest degree of withdrawal was noted for those rats which had been maintained on PB on Day 13. It was also observed that those rats which had begun to exhibit withdrawal on day 13 continued to show signs of withdrawal on Day 14. Bromazepam provided dose-dependent substitution for PB in dependent rats (Fig. 2Al. It was noted that the higher dose of bromazepam substituted completely for PB while the lower dose provided no substitution until the tenth hour of this period. This may be due to accumu-
Mean withdrawal scores of control rats and PBFig. 5. dependent rats during (A) substitution with either saline or bupropion in doses of 300 mg/kg per 24 h (BUP (300)) or 150 mg/kg per 24 h (BUP (150)) and (B) during saline substitution. Each point represents the mean of three-five rats. Filled symbols indicate statistically significant differences (P < 0.05) as compared to vehicle substitution in PB-dependent rats.
lation of bromazepam during this infusion period. On Day 14 there was noted a significant increase in withdrawal scores for rats which had received bromazepam on Day 13 as compared to PB-dependent rats receiving vehicle on Day 13 (Fig. 2B). Diazepam was also noted to substitute for PB (Fig. 3Al. Inexplicably, the lower dose of diazepam substituted to a greater degree than did the higher dose. Again, saline substitution on Day 14 produced elevated withdrawal scores in rats previously infused with diazepam (Fig. 3Bl. In PB-dependent rats, methaqualone at a dose of 200 mg/kg per 24 h provided complete substitution (Fig. 4Al. Interestingly, this effect was carried over into the next day when saline was substituted for methaqualone (Fig. 4Bl. It is hypothesized that the accumulation of methaqualone may have been sufficient to prevent emergence of withdrawal signs on Day 14. Dependent rats infused with the lower dose of methaqualone exhibited withdrawal symptoms which were similar in degree to that observed for dependent rats infused with vehicle. Not unexpectedly, neither nortriptyline nor mazindol were found to substitute for PB. In fact, mazindol produced a substantial exacerbation of the withdrawal signs. Interestingly, the lower dose of bupropion appeared to partially ameliorate the behavioral signs of withdrawal (Fig. 5A,5Bl. Loss of body weight is a consistent finding during withdrawal from barbiturates 19,121.PBdependent rats which were subsequently infused with either saline or vehicle were noted to exhibit an approximate 10% loss of body weight by 24 h of the substitution period (Fig. 6A). The body weight was noted to be returning toward control values by the end of Day 14. Rats continuing to receive PB on Day 13 demonstrated no significant alteration in body weight on Day 13 but did demonstrate substantial weight loss following saline substitution on Day 14. A fluctuation in the body weight of control rats was noted over the course of the 2 days of observation and is presumed to be reflective of a circadian pattern of body weight regulation.
15
Percent change in body weight of control and PBFig. 7. dependent rats during both the drug substitution period (Day 13) and the subsequent saline substitution period (Day 14) (A). Either saline, vehicle or methaqualone in doses of 200 mg/kg per 24 h (MTQ (200)) or 100 mg/kg per 24 h (MTQ (100)) was substituted into dependent rats. (B) Either saline or bupropion in doses of 300 mgkg per 24 h (BUP (300)) or 150 mg/hg per 24 h (BUP 050)) was substituted into dependent rats. Each point represents the mean 2 S.E.M. of the percent change in body weight as compared to time zero. Filled symbols indicate statistically significant differences (P < 0.05) as compared to vehicle substitution in PB-dependent rats. The arrow indicates the start of saline infusion on Day 14. Percent change in body weight of control and PBFig. 6. dependent rats during both the substitution phase (Day 13) and the withdrawal phase (Day 14). (A) PB-dependent rats were infused, for 24 h, with either saline, alcoholic vehicle, or PB, 950 mg/lcg per 24 h (PB (950)) followed by saline infusion for 24 h. (B) PB-dependent rats were infused, for 24 h, with bromazepam in doses of either 28 mg/kg per 24 h (BRZ (28)) or 14 mg/hg per 24 h (BRZ (14)) followed by saline infusion for 24 h. (Cl PB-dependent rats were infused for 24 h, with diazepam in doses of either 40 mg/kg per 24 h (DZ (40)) or 20 mg/hg per 24 h (DZ (20)) followed by saline infusion for 24 h. Control rats received saline during both the substitution and withdrawal phases. Each point is the mean + S.E.M. of the percent change in body weight, as compared to time zero, for three-six rats. Filled symbols indicate statistically signfficant differences (P 6 0.05) as compared to vehicle substitution in PB-dependent rats. The arrow indicates the start of saline infusion on Day 14.
Substitution of bromazepam or diazepam into PB-dependent rats effectively prevented a significant loss of body weight (Fig. 6B, 6C). Substitution of saline into these rats on Day 14 resulted in significant declines in body weight at 24 h of this period. Methaqualone substitution also provided dose-dependent attenuation of the typical decrease in body weight observed during PB withdrawal (Fig. 7A). Substitution with either bupropion (Fig. 7B) or mazindol (data not shown) on Day 13 resulted in exacerbation of weight loss observed in PBdependent rats. The loss of body weight was
16
sustained through at least the end of Day 14. When nortriptyline was substituted for PB it was observed that the higher dose of this drug ameliorated the weight loss of dependent rats while the lower dose of nortriptyline exacerbated the weight loss (data not shown). Again, those effects on body weight were carried through into Day 14. Discussion Cross-dependence studies have been shown to be advantageous when used to help predict the dependence liability of compounds as well as to explore similarities among compounds with respect to mechanisms which may underlie the development of physical dependence [1,2,3,5]. It is generally hypothesized that those compounds which substitute for another compound may possess both a mechanism of action and a propensity for development of physical dependence which is similar to that of the first compound. On the other hand, those compounds which fail to substitute are not as likely to produce the same spectrum of dependence [1,3,7]. Previous investigations have demonstrated the utility of the continuous infusion method in the development of a rodent model of dependence on opiates [16] as well as on CNS depressants [12,17]. In addition, a previous investigation has demonstrated the use of this model in the substitution of the benzodiazepines, temazepam and midazolam, in PB-dependent rats [9]. The current investigation extends these studies to include two additional benzodiazepines (diazepam, bromazepaml, a non-barbiturate, sedative-hypnotic (methaqualonel, a CNS stimulant (mazindoll and two antidepressant agents (nortriptyline, bupropion). It has been demonstrated that the administration of bromazepam, diazepam or methaqalone were all effective in substituting for PB and acted to maintain the dependent state. This was evidenced by both the amelioration of behavioral signs of withdrawal (Figs. 2, 3, 41 as well by the reduction in the characteristic loss of body weight observed during the withdrawal
period (Figs. 6BA, 6C, 7Al. The fact that the dependent state was being maintained was demonstrated by the substitution of saline for the drug on Day 14. Those rats in which the physical dependence had been maintained by the substituted drug were, following saline substitution, observed to exhibit both an increased withdrawal scores as well as a substantial decrease in body weight. Taken together, this provides evidence that the substitution of these drugs did indeed mantain the physical dependence state. Neither mazindol or nortriptyline were effective in substituting for PB in the dependent rats. This was evidenced by both increased withdrawal scores and a substantial loss of body weight. The situation with bupropion substitution was not as clear. In this case, the lower dose of bupropion appeared to substitute somewhat as evidenced by lower withdrawal scores (Fig. 51. However this dose of bupropion was seemingly ineffective in ameliorating the loss of body weight (Fig. 7Bl. The higher dose of bupropion, on the other hand, produced both increased withdrawal scores and a substantial loss of body weight. It was concluded that bupropion did not substitute for PB. The somewhat differing results obtained with bupropion as compared to nortriptyline might be explained, at least in part, in that bupropion action seems to be directed at dopaminergic neurons whereas nortriptyline action is primarily directed towards inhibition of reuptake of serotonin and/or norepinephrine [18,19]. The results obtained in this study are not unexpected and are consistent with the basic premise which underlies the cross-dependence paradigm. The benzodiazepines, methaqualone and PB have all been postulated to act, at least in part, at a GABA/benzodiazepine/barbiturate receptor complex [20,21,22]. The typical pharmacological effects of these agents are similar in many respects and include sedation, hypnosis and anticonvulsant activity [23]. On the other hand, mazindol, bupropion and nortriptyline are presumed to possess actions directed towards alteration of reuptake and/or
17
storage of biogenic amines in the neuron [24,25]. The current investigation serves to demonstrate the selectivity of the cross-dependence paradigm by underscoring the inability of CNS stimulants and antidepressants to maintain the physical dependence established towards a CNS depressant agent. In addition the study extends the results of previous investigations which describe the utility of the cross-dependence paradigm in studies serving to elucidate the phenomena which underlie physical dependence. Acknowledgements This study was supported, in part, by USPHS grants DA-00490, DA-07027 and DA04244. Bromazepam, bupropion, diazepam, mazindol. methaqualone and nortriptyline were provided by Dr. Arthur Jacobson, Drug Testing Committee, Committee on Problems of Drug Dependence. We express our appreciation to Ms. Shellie Delaney and Ms. Raelene Gilbert for their help in preparation of this manuscript. References 1
J.L. Katz and S.R. Goldberg, Agents Actions, 23 (19881 18.
2
J.H. Woods, J.L Katz and G. Winger., Pharmacol. Rev., 39 (1987) 251. C.K. Himmelsbach, J. Pharmacol. Exp. Ther., 71 (1941) 41. J.K. Belknap, Clin. Toxicol., 12 (1978) 427. B.E. Jones, J.A. Prada and W.R. Martin, Psychopharmacology, 47 (197617.
3 4 5
6 7 8 9 10
11 12 13 14 15 16 17 18 19 20 21 22 23
24
25
P.R.E. Norton, J. Pharm. Pharmacol., 22 (1970) 763, E. Tagashira, T. Urano, T. Hiramori and S. Yanaura, Jpn. J. Pharmacol., 33 (1983) 659. T. Yanagita and S. Takahashi, J. Pharmacol. Exp. Ther., 185 (1973) 303. G.J. Yutrzenka, G.A. Patrick and W. Rosenberger, Physiol. Behav., 46 (1989) 55. T. Yanagita, Handbook of Experimental Pharmacology, Vol. 55. In: F. Hoffmeister (Ed.), Springer-Verlag. New York, 1981, pp. 395. W.R. Martin, L.F. MeNicholas and S. Cherian, Life Sci., 3lf1982) 721. G.J. Yutrzenka, G.A. Patrick and W. Rosenberger, J. Phamacol. Exp. Ther., 232 (1985) 111. L.L. Coughenour, J.R. McLean and R.B. Parker, Pharmacol. Biochem. Behav., 6 (1977) 351. R.H. Kolstoe, Introduction to Statistics for the Behavioral Sciences, Dorsey Press, Homewood, IL, 1973. J.T. Litchfield and F. Wilcoxon, J. Pharmacol. Exp. Ther., 96 (1949) 99. D.G. Teiger, J. Pharmacol. Exp. Ther., 190 (1974) 408. G.J. Yutrzenka, Pharmacol. Biochem. Behav., 34 (1989) 49. L.E. Hollister, Rational Drug Ther., 16 (1982) 1. M.V. Rudorfer, R.N. Golden and W.Z. Potter, Psychiat. Clin. N. Am., 7 (1984) 519. R.W. Olsen, J. Neurochem., 37 (1981) 1. S.R. Naik, P.R. Naik and U.K. Sheth, Psychopharmacology, 57 (1978) 103. T. Suzuki, Y. Koike, Y. Chida and M. Misawa, Jpn. J. Pharmacol., 46 (1988) 403. S.C. Harvey, The Pharmacological Basis of Therapeutics, 7th edn. A.G. Gilman, L.S. Goodman, T.W. Rail and F. Murad (Eds.), MacMillan Publishing Co., New York, 1985, p. 339. R.J. Baldessarini, The Pharmacological Basis of Therapeutics. 7th edn. A.G. Gilman, L.S. Goodman, T.W. Rall and F. Murad (Eds.), MacMillan Publishing Co., New York, 1985, p. 412. A. Goth, Medical Pharmacology: Principles and Concepts. 11th edn., C.V. Mosby Co., St. Louis, 1984, p. 175.
