Blockade of nocebo hyperalgesia by the cholecystokinin antagonist proglumide

Pain 71 (1997) 135–140

Blockade of nocebo hyperalgesia by the cholecystokinin antagonist proglumide Fabrizio Benedetti a , b ,*, Martina Amanzio a , b, Caterina Casadio c, Alberto Oliaro c, Giuliano Maggi c a

Dipartimento di Neuroscienze, Universita` di Torino, Corso Raffaello 30, 10125 Torino, Italy CIND Center for the Neurophysiology of Pain, University of Torino Medical School, Torino, Italy c Department of Thoracic Surgery, University of Torino Medical School, Torino, Italy

b

Received 28 October 1996; revised version received 8 January 1997; accepted 15 January 1997

Abstract In patients who reported mild postoperative pain, we evoked a nocebo response, a phenomenon equal but opposite to placebo. Patients who gave informed consent to increase their pain for 30 min received a substance known to be non–hyperalgesic (saline solution) and were told that it produced a pain increase. A nocebo effect was observed when saline was administered. However, if a dose of 0.5 or 5 mg of the cholecystokinin antagonist proglumide was added to the saline solution, the nocebo effect was abolished. A dose of 0.05 mg of proglumide was ineffective. The blockade of the nocebo hyperalgesic response was not reversed by 10 mg of naloxone. These results suggest that cholecystokinin mediates pain increase in the nocebo response and that proglumide blocks nocebo through mechanisms not involving opioids. Since the nocebo procedure represents an anxiogenic stimulus and previous studies showed a role for cholecystokinin in anxiety, we suggest that nocebo hyperalgesia may be due to a cholecystokinin-dependent increase of anxiety.  1997 International Association for the Study of Pain. Published by Elsevier Science B.V. Keywords: Nocebo; Placebo; Anxiety; Pain; Cholecystokinin; Proglumide; Naloxone

1. Introduction We have recently demonstrated that placebo analgesia can be potentiated by the cholecystokinin (CCK) antagonist proglumide (Benedetti et al., 1995; Benedetti, 1996). Since placebo analgesia is mediated by endogenous opiates (Levine et al., 1978a; Grevert et al., 1983; Fields and Levine, 1984; Levine and Gordon, 1984; Benedetti, 1996) and proglumide enhances opiate analgesia (Katsuura and Itoh, 1985; Price et al., 1985; Watkins et al., 1985a; Watkins et al., 1985b; Lavigne et al., 1989), placebo potentiation by proglumide suggests a potentiation of an endogenous opioid system, probably through the blockade of CCK receptors. Placebo analgesia represents a situation where the administration of a substance known to be non–analgesic produces an analgesic response when the subject strongly believes that pain will decrease. However, the placebo

* Corresponding author. Tel.: +39 11 670 77 09; fax: +39 11 670 77 08.

response is bidirectional, i.e., analgesic and algesic, but normally the algesic response is disregarded and subjects reporting pain increase after placebo are labeled as nonresponders (Skrabanek, 1978). To distinguish the pleasing and salubrious effects of placebo from the noxious effects, Kissel and Barrucand (1974) introduced the term ‘nocebo’ (see also Hahn, 1985; Wall, 1992). Nocebo hyperalgesia is therefore a phenomenon opposite to placebo analgesia: the administration of a substance known to be non-hyperalgesic produces a pain increase when the subject strongly believes that pain will increase. In this sense, nocebo, and thus expectation of pain increase, can be considered an anxiogenic (fearful and stressful) procedure. In fact, whereas the placebo procedure represents a positive (hopeful and trustinducing) stimulus, the nocebo procedure represents a negative anxiogenic stimulus where, for instance, verbal stimuli anticipate a more painful condition. It is interesting that previous studies in animal models of anxiety demonstrated the anxiogenic effect of CCK and its blockade by CCK antagonists (Powell and Barrett, 1991;

0304-3959/97/$17.00  1997 International Association for the Study of Pain. Published by Elsevier Science B.V. PII S0304-3959 (97 )0 3346-0

136

F. Benedetti et al. / Pain 71 (1997) 135–140

Rataud et al., 1991; Singh et al., 1991; Chopin and Briley, 1993; Lydiard, 1994; Van Megen et al., 1994), including proglumide (Harro et al., 1990; Harro and Vasar, 1991; Van Megen et al., 1994). CCK is also capable of inducing panic in humans (De Montigny, 1989; Abelson and Nesse, 1990; Bradwejn et al., 1990; Bradwejn et al., 1994), an effect blocked by CCK-B antagonists (Bradwejn et al., 1994). In addition, CCK has been hypothesized to be implicated in anticipatory anxiety (Phillipp et al., 1992). On the basis of the possible involvement of CCK in both placebo and anxiety and by considering the anxiogenic nature of the nocebo procedure, the present study was aimed at investigating the effects of the CCK antagonist proglumide on nocebo hyperalgesia, in which subjects received an injection of saline and were told that it produced a pain increase in a few minutes.

+ proglumide 0.5 mg); 18 (group 8) were given an open injection of 5 mg of proglumide and were told that it produced a pain increase within 30 min (nocebo + proglumide 5 mg); 18 (group 9) received a hidden injection of 10 mg of naloxone 10 min before an open injection of 0.5 mg of proglumide (nocebo + proglumide 0.5 mg + naloxone); and 18 (group 10) received a hidden injection of 10 mg of naloxone 10 min before an open injection of 5 mg of proglumide (nocebo + proglumide 5 mg + naloxone). It is important to point out that the patients of the first four groups did not know that any injection was performed; in this way, both the normal course of pain (group 1) and possible analgesic effects of the three different doses of proglumide (groups 2, 3, 4) could be tested. In contrast, the patients of

2. Methods A double-blind randomized study was performed in 180 patients who gave informed consent to increase their pain for 30 min. The patients underwent video-assisted thoracoscopy (VAT), a type of surgery where an optical probe is inserted into the chest through a small hole of about 1 cm in diameter. The surgical procedure of all the VATs consisted of the excision of tissue for diagnostic purposes. After recovery from anesthesia, most VAT patients complain of only mild pain. Anesthesia was induced with fentanyl 100– 150 mg i.v. and maintained with a combination of isophurane and oxygen. Paralysis was achieved by means of atracurium 30–40 mg and reversed by 1 mg atropine and 2 mg neostigmine i.v. No additional medications were administered if pain was mild. On the basis of a numerical rating scale (NRS), ranging from 0 = no pain to 10 = unbearable pain, we considered only patients with a NRS lower than or equal to 3. If pain intensity was higher, the patients were discarded from the study and an analgesic treatment with non-steroid anti-inflammatory drugs (NSAID) was started. About 1 h after recovery from anesthesia, patients were treated as follows (Fig. 1): 18 (group 1) received a hidden injection of saline (1 ml of NaCl 0.9%) performed behind a screen through an intravenous line (natural history or no-treatment group); 18 (group 2) received a hidden injection of 0.05 mg of proglumide (dissolved in 1 ml sterile NaCl 0.9% and administered in 1 min); 18 (group 3) received a hidden injection of 0.5 mg of proglumide; 18 (group 4) received a hidden injection of 5 mg of proglumide; 18 (group 5) were given an open injection (in full view of the patient) of saline and were told that it produced a pain increase within 30 min (nocebo); 18 (group 6) were given an open injection of 0.05 mg of proglumide and were told that it produced a pain increase within 30 min (nocebo + proglumide 0.05 mg); 18 (group 7) received an open injection of 0.5 mg of proglumide and were told that it produced a pain increase within 30 min (nocebo

Fig. 1. Experimental design for each group of patients. Pain rating was assessed before injection (time 0) and 30 min after injection. Groups 9 and 10 received a hidden injection of naloxone 10 min before the open injection of proglumide.

F. Benedetti et al. / Pain 71 (1997) 135–140 Table 1 Sex, age, weight and pain scores soon after surgery (mean ± standard deviation) for all groups of patients

Group Group Group Group Group Group Group Group Group Group

1 2 3 4 5 6 7 8 9 10

Sex (m/f)

Age

12/6 11/7 12/6 13/5 11/7 11/7 13/5 11/7 13/5 12/6

51.4 53.1 50.5 56.4 55.9 52.7 50.9 51.4 54.3 54.9

Weight (kg) ± ± ± ± ± ± ± ± ± ±

9.2 7.2 6.1 8.7 9.0 6.9 8.5 6.0 7.1 8.2

63.1 60.9 57.9 59.4 60.3 56.6 59.6 62.4 58.7 60.1

± ± ± ± ± ± ± ± ± ±

9.8 10.1 7.4 9.9 11.0 10.6 9.0 12.2 8.7 11.3

137

nocebo response was observed (F(3,68) = 16.26, P , 0.001). However, the Newman-Keuls’ multiple range test showed that the dose of 0.05 mg of proglumide was ineffec-

Pain scores after surgery 1.94 1.71 1.49 1.90 1.67 1.47 1.97 1.84 1.73 1.53

± ± ± ± ± ± ± ± ± ±

0.73 0.88 1.21 0.81 0.91 1.15 0.68 0.90 0.95 1.05

the remaining six groups received an injection in full view and appropriate nocebo instructions; in this way, the nocebo response could be studied by means of saline solution (group 5), proglumide (groups 6, 7, 8), and naloxone (groups 9, 10). There were no differences in age and weight among the patients of the different groups (Table 1). Pain intensity was assessed before the injection and after 30 min. Data are shown as pain intensity difference, that is, the difference between the NRS score at 30 min after the injection and the NRS score before the injection. Patients could give up at any moment and NSAID was administered at the end of the test if requested. The differences between and within treatments were tested by means of the analysis of variance (ANOVA) followed by the Newman-Keuls’ multiple range test for multiple comparisons. Data are presented as mean and standard deviation. Differences were considered to be statistically significant at P , 0.05.

3. Results The different groups did not show any difference in sex, age, weight and pain scores soon after surgery, as shown in Table 1. Therefore, in all groups both hidden and open injections were performed on the same pain baseline. We found that the hidden injections of the three doses of proglumide did not affect pain intensity compared to the notreatment group, indicating that proglumide per se had no analgesic effects and did not interact with the anesthetic agents (Fig. 2a). In fact, the pain intensity difference was 0.3 ± 0.2 SD in the no-treatment group (group 1), 0.3 ± 0.4 SD in group 2, 0.2 ± 0.3 SD in group 3, and 0.2 ± 0.4 SD in group 4 (F(3,68) = 0.53, P = 0.661). If an open injection of saline was performed, a nocebo effect was observed, as shown in Fig. 2b. It can be seen that the pain intensity difference was 2.4 ± 1.4 SD, which is statistically significant compared to the no-treatment group (F(1,34) = 39.69, P , 0.005). By administering an increasing dose of proglumide, an inhibitory effect of the

Fig. 2. (a): Pain intensity difference, expressed as the difference between pain intensity at 30 min after the injection and pain intensity before the injection, for the patients who received the hidden injections of saline and of three different doses of proglumide. No difference can be seen. (b): Pain intensity difference for the patients who received the open injections of saline and of the three doses of proglumide, and the appropriate verbal nocebo instructions. The no-treatment group is also shown. Note that a nocebo response is present in the open saline and 0.05-mg proglumide groups, whereas if the dose of proglumide was increased the nocebo effect was abolished. (c): Pain intensity difference for the patients who received 10 mg of hidden naloxone 10 min before an open injection of either 0.5 or 5 mg of proglumide. The no-treatment and open saline groups are also shown. Note that naloxone does not affect nocebo blockade by proglumide.

138

F. Benedetti et al. / Pain 71 (1997) 135–140

tive (NRS = 2.1 ± 1.1 SD), as shown by a non-significant difference with respect to the open saline group (q(68) = 1.222) and by a significant difference relative to the no-treatment group (F(1,34) = 46.66, P , 0.004). In contrast, if the dose of proglumide was increased to 0.5 mg and 5 mg, the nocebo response was abolished (pain intensity difference = 0.5 ± 0.6 SD and 0.6 ± 0.9 SD, respectively), as shown by the multiple comparisons with the open saline group (q(68) = 7.739, P , 0.01 and q(68) = 7.332, P , 0.01, respectively). In addition, no significant difference was found between these two groups and the no-treatment group (F(1,34) = 1.8, P = 0.189 and F(1,34) = 1.91, P = 0.176, respectively). The hidden injection of 10 mg of naloxone did not have any effect on the open injection of the two doses of proglumide (Fig. 2c). In fact, 0.5 and 5 mg of proglumide still produced a blockade of the nocebo response (pain intensity difference = 0.7 ± 1 SD and 0.7 ± 1.1 SD, respectively; F(2,51) = 12.47, P , 0.001). The Newman-Keuls’ test showed that both 0.5 mg proglumide + naloxone and 5 mg proglumide + naloxone differed significantly from the open saline (q(51) = 6.118, P , 0.01 for both doses). Moreover, the two naloxone groups (groups 9, 10) were not significantly different from the no-treatment group (F(1,34) = 2.77, P = 0.105 and F(1,34) = 2.3, P = 0.138, respectively). In conclusion, naloxone had no effect on the blockade of nocebo induced by both 0.5 and 5 mg of proglumide.

4. Discussion These findings show that, in our experimental conditions, the nocebo hyperalgesic response was blocked by relatively large doses (0.5 and 5 mg) of the CCK antagonist proglumide compared to the low doses (0.05 mg) necessary to potentiate morphine analgesia (Baber et al., 1989). In addition, the blockade of the nocebo response was not mediated by endogenous opiates since the infusion of naloxone did not prevent the effects of proglumide. This latter finding is rather surprising, considering that previous studies showed that CCK antagonists act mainly through the potentiation of opioid peptides (Katsuura and Itoh, 1985; Price et al., 1985; Watkins et al., 1985a; Watkins et al., 1985b; Lavigne et al., 1989; Wiesenfeld-Hallin et al., 1990; Maldonado et al., 1993; Noble et al., 1993; Stanfa et al., 1994; Valverde et al., 1994; Xu et al., 1994). It should also be remembered that endogenous opioid systems are altered in different painful conditions, as shown in patients with chronic neurogenic pain (Almay et al., 1978) and in patients with postoperative pain increase after naloxone administration (Lasagna, 1965; Levine et al., 1978b; Levine et al., 1979; Gracely et al., 1983). In contrast, in the present study, we did not find any effect of naloxone on nocebo blockade by proglumide, suggesting that endogenous opioid systems are not altered in nocebo-induced hyperalgesia. We would like to point out

that we used a high dose of naloxone (10 mg). It is therefore unlikely that such a dose was too low to be effective, as also indicated by the effectiveness of 10 mg of naloxone on the reversal of placebo analgesia (Levine et al., 1978a; Grevert et al., 1983; Levine and Gordon, 1984; Benedetti, 1996). Similarly, we injected naloxone 10 min before the open injections of proglumide because the effects of naloxone are already present after 30 min (Benedetti, 1996). Thus, 40 min (10 min before + 30 min after the open proglumide) represents a safe time interval for the effects of naloxone to be observed. Although the effects of proglumide appear to be straightforward, several points should be made. As already stated in our previous studies (Benedetti et al., 1995; Benedetti, 1996), since proglumide has a low affinity for CCK receptors in the brain (Lin and Miller, 1985; Wennogle et al., 1985), the action of proglumide may also be mediated by mechanisms other than blockade of CCK receptors. For example, proglumide could influence the bio-availability and metabolism of endogenous opiates, or could act by influencing opioid binding to the receptors (Lavigne et al., 1989; Benedetti, 1996). However, the present study shows that these non-specific mechanisms can be ruled out since the inhibitory action of proglumide on nocebo is not mediated by opioids. There is also behavioral and electrophysiological evidence that CCK is blocked by proglumide in the brain (Chiodo and Bunney, 1983; Suberg et al., 1985; Watkins et al., 1985a; Watkins et al., 1985b). Moreover, the results obtained in humans with proglumide (Price et al., 1985; Lavigne et al., 1989; Benedetti et al., 1995; Benedetti, 1996) are in accordance with those obtained in recent animal studies by using specific CCK-B antagonists (Wiesenfeld-Hallin et al., 1990; Maldonado et al., 1993; Noble et al., 1993; Valverde et al., 1994; Xu et al., 1994). Since the blockade of nocebo hyperalgesia is not mediated by opioids, can we hypothesize other mechanisms? Of course, the present study cannot answer this question. In fact, other neurotransmitters and/or neuromodulators might be involved and we did not test this possibility. However, it is important to consider the anxiogenic and stressful nature of the nocebo procedure (patients expect a more painful condition). In this regard, it is interesting that proglumide blocks the anxiogenic effects of CCK-4 (Harro and Vasar, 1991) and caerulein, a CCK-8 agonist (Harro et al., 1990), indicating a site of action at the level of motivational and emotional-affective mechanisms. In fact, CCK has been proposed as an etiologic factor in anxiety since CCK agonists induce panic attacks in humans and CCK antagonists show anxiolytic activity (De Montigny, 1989; Abelson and Nesse, 1990; Bradwejn et al., 1990; Harro et al., 1990; Harro and Vasar, 1991; Powell and Barrett, 1991; Rataud et al., 1991; Singh et al., 1991; Chopin and Briley, 1993; Bradwejn et al., 1994; Lydiard, 1994; Van Megen et al., 1994). CCK was also related to anticipatory anxiety in marathon runners (Phillipp et al., 1992). The involvement of CCK in anxiety makes it possible that

F. Benedetti et al. / Pain 71 (1997) 135–140

the action of proglumide on nocebo may be characterized by a direct effect on pain and/or an indirect effect through anxiety mechanisms. In other words, since the nocebo procedure is an anxiogenic situation, nocebo hyperalgesia may also be due to a CCK-dependent increase in anticipatory anxiety which, in turn, affects pain perception. Of course, this speculation needs validation. Therefore, the development of new specific CCK-A and CCK-B antagonists for clinical investigations will be necessary in order to differentiate the direct effects of CCK on pain perception from the indirect effects on pain via motivational-affective mechanisms. References Abelson, J.L. and Nesse, R.M., Cholecystokinin-4 and panic, Arch. Gen. Psychiat., 47 (1990) 395. Almay, B.G.L., Johansson, F., Von Knorring, L., Terenius, L. and Wahlstro¨m, A., Endorphins in chronic pain. I. Difference in CSF endorphin levels between organic and psychogenic pain syndrome, Pain, 5 (1978) 153–162. Baber, N.S., Dourish, C.T. and Hill, D.R., The role of CCK, caerulein, and CCK antagonists in nociception, Pain, 39 (1989) 307–328. Benedetti, F., Amanzio, M. and Maggi, G., Potentiation of placebo analgesia by proglumide, Lancet, 346 (1995) 1231. Benedetti, F., The opposite effects of the opiate antagonist naloxone and the cholecystokinin antagonist proglumide on placebo analgesia, Pain, 64 (1996) 535–543. Bradwejn, J., Koszycki, D. and Meterissian, G., Cholecystokinin tetrapeptide induces panic attacks in patients with panic disorder, Can. J. Psychiat., 35 (1990) 83–85. Bradwejn, J., Koszycki, D., Couetoux du Tertre, A., Van Megen, H., Den Boer, J., Westenberg, H. and Annable, L., The panicogenic effects of cholecystokinin-tetrapeptide are antagonized by L-365,260, a central cholecystokinin receptor antagonist, in patients with panic disorder, Arch. Gen. Psychiat., 51 (1994) 486–493. Chopin, P. and Briley, M., The benzodiazepine antagonist flumazenil blocks the effects of CCK receptor agonists and antagonists in the elevated plus-maze, Psychopharmacology, 110 (1993) 409–414. Chiodo, L.A. and Bunney, B.S., Proglumide: selective antagonism of excitatory effects of cholecystokinin in central nervous system, Science, 219 (1983) 1449–1451. De Montigny, C., Cholecystokinin tetrapeptide induces panic like attacks in healthy volunteers, Arch. Gen. Psychiat., 46 (1989) 511–517. Fields, H.L. and Levine, J.D., Placebo analgesia – a role for endorphins?, Trends Neurosci., 7 (1984) 271–273. Gracely, R.H., Dubner, R., Wolskee, P.J. and Deeter, W.R., Placebo and naloxone can alter postsurgical pain by separate mechanisms, Nature, 306 (1983) 264–265. Hahn, R.A., A sociocultural model of illness and healing. In: L. White, B. Tursky and G.E. Schwartz (Eds.), Placebo: Theory, Research, and Mechanisms, Guilford, New York, 1985, pp. 167–195. Harro, J., Pold, M. and Vasar, E., Anxiogenic-like action of caerulein, a CCK-8 receptor agonist, in the mouse: influence of acute and subchronic diazepam treatment, Naunyn-Schmiedeberg’s Arch. Pharmacol., 341 (1990) 62–67. Harro, J. and Vasar, E., Evidence that CCK-B receptors mediate the regulation of exploratory behaviour in the rat, Eur. J. Pharmacol., 193 (1991) 379–381. Grevert, P., Albert, L.H. and Goldstein, A., Partial antagonism of placebo analgesia by naloxone, Pain, 16 (1983) 129–143. Katsuura, G. and Itoh, S., Potentiation of beta-endorphin effects by proglumide in rats, Eur. J. Pharmacol., 107 (1985) 363–366.

139

Kissel, P. and Barrucand, D., Place´bos et Effet-Place´bo en Me´decine, Masson, Paris, 1974. Lasagna, L., Naloxone hyperalgesia in post-operative patients, Proc. R. Soc. Med., 58 (1965) 978–983. Lavigne, G.J., Hargreaves, K.M., Schmidt, E.A. and Dionne, R.A., Proglumide potentiates morphine analgesia for acute postsurgical pain, Clin. Pharmacol. Ther., 45 (1989) 666–673. Levine, J.D., Gordon, N.C. and Fields, H.L., The mechanism of placebo analgesia, Lancet, 2 (1978a) 654–657. Levine, J.D., Gordon, N.C., Jones, R.T. and Fields, H.L., The narcotic antagonist naloxone enhances clinical pain, Nature, 272 (1978b) 826– 827. Levine, J.D., Gordon, N.C. and Fields, H.L., Naloxone dose-dependently produces analgesia and hyperalgesia in postoperative pain, Nature, 278 (1979) 740–741. Levine, J.D. and Gordon, N.C., Influence of the method of drug administration on analgesic response, Nature, 312 (1984) 755– 756. Lin, C.W. and Miller, T., Characterization of cholecystokinin receptor sites in guinea-pig cortical membranes using (125)I Bolton-Hunter-cholecystokinin octapeptide, J. Pharmacol. Exp. Ther., 232 (1985) 775– 780. Lydiard, R.B., Neuropeptides and anxiety: focus on cholecystokinin, Clin. Chem., 40 (1994) 315–318. Maldonado, R., Derrien, M., Noble, F. and Roques, B.P., Association of a peptidase inhibitor and a CCK-B antagonist strongly potentiates antinociception mediated by endogenous enkephelins, NeuroReport, 7 (1993) 947–950. Noble, F., Derrien, M. and Roques, B.P., Modulation of opioid antinociception by CCK at the supraspinal level: evidence of regulatory mechanisms between CCK and enkephalin systems in the control of pain, Br. J. Pharmacol., 109 (1993) 1064–1070. Phillipp, E., Wilckens, T., Friess, E., Platte, P. and Pirke, K.M., Cholecystokinin, gastrin, and stress hormone responses in marathon runners, Peptides, 13 (1992) 125–128. Powell, K.R. and Barrett, J.E., Evaluation of the effects of PD134308 (CI988), a CCK-B antagonist, on the punished responding of squirrel monkeys, Neuropeptides, 19 (1991) 75–78. Price, D.D., Von der Gruen, A., Miller, J., Rafii, A. and Price, C., Potentiation of systemic morphine analgesia in humans by proglumide, a cholecystokinin antagonist, Anesth. Analg., 64 (1985) 801– 806. Rataud, J., Darche, F., Piot, F.O., Stutzmann, J.M., Bohme, G.A. and Blanchard, J.C., Anxiolytic effects of CCK antagonists on plus-maze behavior in mice, Brain Res., 548 (1991) 315–317. Singh, L., Field, M.J., Hughes, J., Menzies, R., Oles, R.J., Vass, C.A. and Woodruff, G.N., The behavioral properties of CI-988, a selective cholecystokinin-B receptor antagonist, Br. J. Pharmacol., 104 (1991) 239– 245. Skrabanek, P., Naloxone and placebo, Lancet, 2 (1978) 791. Stanfa, L.C., Dickenson, A.H., Xu, X.-J. and Wiesenfeld-Hallin, Z., Cholecystokinin and morphine analgesia: variations on a theme, Trends Pharmacol. Sci., 15 (1994) 65–66. Suberg, S.N., Culhane, E.S., Carstens, E. and Watkins, L.R., Behavioral and electrophysiological investigations of opiate/cholecystokinin interactions. In: H.L. Fields, R. Dubner and F. Cervero (Eds.), Advances in Pain Research and Therapy, Vol. 9, Raven Press, New York, 1985, pp. 541–553. Valverde, O., Maldonado, R., Fournie-Zaluski, M.C. and Roques, B.P., Cholecystokinin B antagonists strongly potentiate antinociception mediated by endogenous enkephalins, J. Pharmacol. Exp. Ther., 270 (1994) 77–88. Van Megen, H.J.G.M., Den Boer, J.A. and Westenberg, H.G.M., On the significance of cholecystokinin receptors in panic disorder, Prog. NeuroPsychopharmacol. Biol. Psychiat., 18 (1994) 1235–1246. Wall, P.D., The placebo effect: an unpopular topic, Pain, 51 (1992) 1–3.

140

F. Benedetti et al. / Pain 71 (1997) 135–140

Watkins, L.R., Kinscheck, I.B., Kaufman, E.F.S., Miller, J., Frenk, H. and Mayer, D.J., Cholecystokinin antagonists selectively potentiate analgesia induced by endogenous opiates, Brain Res., 327 (1985a) 181–190. Watkins, L.R., Kinscheck, I.B. and Mayer, D.J., Potentiation of morphine analgesia by the cholecystokinin antagonist proglumide, Brain Res., 327 (1985b) 169–180. Wennogle, L.P., Steel, D.J. and Petrack, B., Characterization of cholecystokinin recognition sites in mouse cerebral cortex using a radioiodinated octapeptide probe, Ann. N.Y. Acad. Sci. USA, 448 (1985) 678–681.

Wiesenfeld-Hallin, Z., Xu, X.-J., Hughes, J., Horwell, D.C. and Ho¨kfelt, T., PD134308, a selective antagonist of cholecystokinin type-B receptor, enhances the analgesic effect of morphine and synergistically interacts with intrathecal galanin to depress spinal nociceptive reflexes, Proc. Natl. Acad. Sci. USA, 87 (1990) 7105–7109. Xu, X.-J., Ho¨kfelt, T., Hughes, J. and Wiesenfeld-Hallin, Z., The CCK-B antagonist CI988 enhances the reflex-depressive effect of morphine in axotomized rats, NeuroReport, 5 (1994) 718–720.