Oral application of citalopram (20 mg) and its usefulness for neuroendocrine challenge tests

Transcription

1 International Journal of Neuropsychopharmacology (2002), 5, Copyright 2002 CINP DOI: S X Oral application of citalopram (20 mg) and its usefulness for neuroendocrine challenge tests BRIEF REPORT J. Hennig and P. Netter Institute for Psychobiology and Behavioural Medicine, Department of Psychology, University of Giessen, Germany Abstract The present study was conducted to investigate whether a single oral dosage of 20 mg of the selective serotonin reuptake inhibitor citalopram could be used as a tool to stimulate hormone secretion in neuroendocrine challenge paradigms. A total number of 48 healthy male subjects received either 20 mg citalopram or placebo in a randomized, double-blind cross-over design at an interval of 1 wk between both sessions. Citalopram was well tolerated without side-effects. Growth hormone (GH), prolactin (Prl) and cortisol (Cort) were determined in blood samples obtained at different time-points across the experiment according to drug kinetics. While GH and Prl were not changed after citalopram Cort levels increased as compared to the placebo condition, significantly about 2 h after drug intake. Results will be discussed in respect of the question of whether or not Prl and GH responses after treatment with i.v. applications of SSRIs reflect sertonergic involvement. Received 6 June 2001; Reviewed 21 August 2001; Revised 10 October 2001; Accepted 21 October 2001 Key words: Challenge test, citalopram, cortisol, oral dose. Introduction The neuroendocrine challenge test has been successfully used to investigate responsiveness of neurotransmitter systems in several groups of patients as well as in healthy controls. This paradigm has been intensively applied to investigate serotonergic functioning by the use of precursors, agonists, releasers and, more recently, specific serotonin (5-HT) reuptake inhibitors (SSRI) (for review see Power and Cowen, 1992). However, many of the substances used so far were criticized with respect to their limited specificity for the serotonergic system. For example, prolactin (Prl) responses after treatment with the 5-HT partial agonist buspirone seems to be mediated by the dopaminergic system (Maskall et al., 1995). Although controversial, the same may be true for the 5-HT releaser d-fenfluramine which stimulates Prl as well but, again, this response was seen to involve post-synaptic dopamine receptors (Storlien and Smythe, 1992). In general it seems likely that substances leading to Prl and or growth hormone (GH) responses may include the involvement of dopamine and or norepinephrine. Although clomipramine belongs to the SSRIs, the metabolite dimethyl- Address for correspondence: Dr J. Hennig, Institute for Psychobiology and Behavioral Medicine, Department of Psychology, University of Giessen, Otto-Behaghel-Str. 10, D Giessen, Germany. Tel.: Fax: juergen.hennig psychol.uni-giessen.de clomipramine has been demonstrated to act as a norepinephrine reuptake inhibitor (Laakmann et al., 1984), which, in part, may explain GH increases after i.v. application of clomipramine. Several other examples could be mentioned. The availability of SSRIs with higher specificity is needed to demonstrate that hormone responses are exclusively mediated by the 5-HT system. Citalopram seems to be an interesting substance for this purpose since it is the most specific SSRI available up to now (Hyttel, 1982). It is well tolerated and effective in the treatment of depression and panic disorders (Bezchlibnyk- Butler et al., 2000). The i.v. adminstration (20 mg infused over a time period of 30 min) has recently been demonstrated to influence hormone secretions (Seifritz et al., 1996). In their sample of 8 healthy volunteers cortisol (Cort) and Prl responses were significant, GH concentrations were not increased after treatment with citalopram. The study clearly demonstrates that a single i.v. application of 20 mg citalopram affects Cort and Prl secretions. Consequently, this regimen has been applied to patients with major depression compared to healthy controls (Kapitany et al., 1999). Patients showed significantly lower Prl responses than controls while Cort did not discriminate between the two groups. Side-effects in the study performed by Seifritz et al. (1996) were comparable with those obtained by Kapitany et al. (1999). A correlation between symptoms and peak hormone responses could not be detected. A positive correlation between nausea and Prl increases

2 68 J. Hennig and P. Netter (r 0 34) failed to achieve statistical significance in the study performed by Kapitany et al. (1999). The present study was conducted to investigate whether a single oral dose of 20 mg citalopram induces hormone responses comparable to those obtained after i.v. application. Moreover, using a larger sample (based on pre-calculated estimates of statistical power) the validity of the results should be increased. Finally, in contrast to both of the previously cited studies on hormone responses after citalopram a double-blind cross-over study should exclude possible effects due to expectations of the experimenter. Methods Subjects The total sample consisted of 48 healthy male volunteers aged between 20 and 35 yr. The means and standard deviations for body weight (kg) and height (cm) were and Most of them were students of different disciplines. Prior to the main experiment subjects were thoroughly informed about the objectives of the study, the substance used and the possible sideeffects. Moreover, a careful screening for any hormonal, cardiovascular, or neurological diseases was undertaken to establish physical and mental health. This screening was based on health questionnaires and additional interviews by the medical staff when statements in questionnaires required further information. Subjects were drug free and non-smokers. Moreover, subjects received detailed written instructions of what to do and what to avoid the evening, night and morning prior the experiment. Among others, instructions included were: to go to bed not later than midnight; to avoid any alcoholic or caffeine-containing beverages; to guarantee more than 6 h sleep. On the day of the experiment lunch was taken between 12:00 and 12:30 hours. Afterwards no kind of food intake was allowed. The study was approved by the Ethics Committee of the German Psychological Association. Procedure On the day of the experiments (at 15:00 hours) subjects were seated in a comfortable chair, venepuncture was undertaken and an indwelling catheter was inserted into the left arm antecubital vein. A solution of 0 9% NaCl was infused throughout the whole experiment to avoid clotting and, as a consequence, frequent subsequent venepuncture stress. Immediately after venepuncture drugs were administered in identical capsules to guarantee blindness for subjects and experimenters. Blood samples were taken immediately after venepuncture and 60, 90, 120, 150, 180, and 210 min after drug intake. Blood samples were centrifuged immediately after sampling and each plasma sample was divided into 3 different cups (one for each hormone determination) which were frozen at 30 C until they were assayed. The reason for this procedure was to avoid frequent thawing which may influence hormone determinations. Subsequent hormone analyses were run within one assay charge to reduce interassay variability and were performed by a fully automated analyser for enzyme-immuno-assays (EIA) (Biochem, Labotech, Freiburg, Germany). All chemicals were provided by DRG Diagnostics (Marburg, Germany). Statistics According to the cross-over protocol two-factor ANOVAs for repeated measures (factor drug and factor time) were computed. Results were considered significant if the error probability was lower than 5%. Furthermore, effect sizes (η ) were calculated. Results A single oral dose of 20 mg citalopram was well tolerated in all of our subjects. Figure 1 demonstrates the changes in hormone concentrations after the challenge. Figure 1(a) provides clear evidence that the dosage of 20 mg given orally was not sufficient to stimulate GH. A closer look at the statistics reveals that the main effect for drug as well as the interaction between drug and time-course of hormone concentrations did not reach statistical significance. All F values were less than 1. Consequently, the effect sizes were of negligible amount and not more than 2% of the variance could be explained by the main effects or the interaction terms. Moreover, systematic variations within a single drug condition cannot be obtained. The picture is quite similar for Prl responses. Although inspection of Figure 1(b) may suggest a response after treatment with citalopram, the statistical evaluation does not confirm this impression. In fact the interaction between treatment and time-course does not reveal a significant F value ( 1) and absolutely no important contribution to the variance (less than 0 1%). However, the main effect of drug treatment becomes significant (F 9 07; d.f. 1,46; p 0 01; η 0 16) which is due to the fact that all concentrations of Prl were slightly (but not significantly) higher in the citalopram condition. After controlling for differences obtained in the first measurement by a subsequent ANOVA the F value for the main effect of drug condition is yielding no explanation of the variance (η 0 001). Taken together one can definitely conclude that a single dose of 20 mg citalopram does not affect GH or Prl in healthy volunteers.

3 Citalopram and neuroendocrine challenge tests 69 Growth hormone (ng/ml) Prolactin (ng/ml) Cortisol (ng/ml) (a) (b) (c) Time Figure 1. Means and S.E.M. of changes in (a) growth hormone, (b) prolactin and (c) cortisol. For statistics see text., Placebo;, citalopram. Interestingly, the conclusion of lack of endocrine effects is not valid for Cort concentrations. Figure 1(c) clearly demonstrates increased Cort concentrations about 2 h after drug intake, while with placebo the undisturbed course of the circadian rhythm can be detected. From a statistical point of view the effects for Cort are significant for the interaction between treatment and time-course reflecting the above-mentioned kinetics of responses. This interaction is significant (F 4 47; d.f. 6,270; p 0 001) but of low effect size (η 0 09). This, however, can be explained by the strong circadian decrease of the hormone, which explains more than 60% of the total variance (F 77 1; d.f. 6,270; p 0 001; η 0 61) and, therefore, overshadows the effect size of the drug condition. This, however, does not contradict the fact, that a clear Cort response can be induced by a single oral dose of 20 mg citalopram. Discussion Our study reveals that a single oral dose of 20 mg citalopram significantly increases Cort levels in peripheral blood in a large sample of healthy volunteers as compared to placebo. Citalopram, therefore, can be used as a promising tool in the paradigm of neuroendocrine challenge tests. The lack of GH and Prl responses in our sample does not contradict this first conclusion for several reasons. In respect to GH our results confirm the observation by Seifritz et al. (1996) that citalopram was not associated with an increase in GH. Moreover, the (puzzling) increase of GH observed in their control group is not visible in our sample. The question whether citalopram antagonizes the increase observed in the placebo condition as suggested by Seifritz et al. (1996) cannot be answered without appropriate studies being conducted to investigate this special aspect. It may be more likely that unspecific factors are responsible for the increase of GH after placebo. If this line of argument is correct, one can conclude that citalopram does not influence GH secretion. In respect to Prl the situation is much more complex. Concerning i.v. administration the only two papers published so far (Seifritz et al., 1996; Kapitany et al., 1999) report significant increases while the regimen we used revealed no effect. The lack of a Prl response after oral administration of SSRIs has also been published for paroxetine (Reist et al., 1996). Furthermore, paroxetine has been demonstrated to be more potent than citalopram (Hyttel, 1994). So if specificity is not correlated with potency, one must be cautious in expecting Prl increases after citalopram. More clearly, Cowen and Sargent (1997) pointed out that treatment with SSRIs does not affect Prl levels within 1 wk of administration. Increases were not present before 3 wk of treatment which suggests that SSRIs do not immediately affect Prl concentrations. One of the most convincing findings is that the substance with probably the highest selectivity for the 5-HT-system, the 5-HT-1a partial agonist ipsapirone, does not affect Prl concentrations (Lesch et al., 1989). The question whether or not 5-HT mediates Prl secretion directly or whether Prl secretion is mediated by different mechanisms within the 5-HT system has frequently been addressed. One has to bear in mind that highly significant Prl increases after oral

4 70 J. Hennig and P. Netter doses of d-fenfluramine in patients (e.g. Coccaro et al., 1989), as well as in healthy volunteers (Hennig et al., 2000) may reflect serotonergic involvement, but it does not imply that other substances (like SSRIs) must also increase Prl levels. In fact both seem to independently mediate 5-HT functioning as demonstrated by the lack of predictability of clinical SSRI responses by Prl changes after d-fenfluramine (Park et al., 1995). What is needed to solve the problems is a direct comparison between i.v. and oral administration of SSRIs, e.g. citalopram in the same sample. Two studies compared these two applications (Baumann et al., 1998, Guelfi et al., 2000). However, they used a slow-drop infusion paradigm which may be not comparable with the applications used in the i.v. challenge studies (Seifritz et al., 1996, Kapitany et al., 1999). Unfortunately, data concerning hormone responses are not available for slow-drop infusion studies. Moreover, it would be interesting to investigate i.v. application of SSRIs under anaesthesia without experiences of pain or other stressful side-effects. A direct comparison between the outcome of challenge studies performed in the morning and in the afternoon (as in this study) should be done. Whether or not overnight fasting compared to the protocol we used makes a difference in GH or Prl responses needs to be investigated thoroughly as well. However, again, both possible sources for error variance do not seem to be of relevance for Cort responses. Oral administration of 20 mg citalopram did not cause any side-effects and one can be sure that changes in Cort concentrations are not due to unspecific effects. Whether this is true for changes in Prl after application of citalopram i.v. needs to be investigated more explicitly using larger samples. From a more global point of view further studies are needed to evaluate oral administration of citalopram as a useful tool to investigate functional aspects of the 5-HT system according to the criteria of Van Praag et al. (1987). Given the specificity of citalopram, a very recent report from Attenburrow et al. (2001) reveals that i.v. applications of low dosages (5, 10 mg) infused for 30 min lead to Prl and Cort responses dose dependently. However, interestingly, pretreatment with cyproheptadine (5-HT- 2A 2C antagonist) did not influence the hormone responses induced by citalopram. This is an interesting result indicating that Prl responses after i.v. drug administration may be a result of other than 5-HT-mediated processes. Acknowledgements This study was supported by the German Research Foundation (DFG). References Attenburrow MJ, Mitter PR, Whale R, Terao T, Cowen PJ (2001). Low-dose citalopram as a 5-HT neuroendocrine probe. Psychopharmacology 155, Baumann P, Nil R, Bertschy G, Jecker A, Bra ndli H, Morand J, Kasas A, Vuagniaux O, Ramseier F (1998). A double-blind double-dummy study of citalopram comparing infusion versus oral administration. Journal of Affective Disorders 49, Bezchlibnyk-Butler K, Aleksic I, Kennedy SH (2000). Citalopram a review of pharmacological and clinical effects. Journal of Psychiatry and Neuroscience 25, Coccaro EF, Siever LJ, Klar HM, Maurer G, Cochrane K, Cooper TB, Mohs RC, Davis KL (1989). Serotonergic studies in patients with affective and personality disorders. Correlates with suicidal and impulsive aggressive behavior. Archives of General Psychiatry 46, Cowen PJ, Sargent PA (1997). Changes in plasma prolactin during SSRI treatment: evidence for a delayed increase in 5-HT neurotransmission. Journal of Psychopharmacology 11, Guelfi JD, Strub N, Loft H (2000). Efficacy of intravenous citalopram compared with oral citalopram for severe depression. Safety and efficacy data from a double-blind, double-dummy trial. Journal of Affective Disorders 58, Hennig J, Toll C, Schonlau P, Rohrmann S, Netter P (2000). Endocrine responses after d-fenfluramine and ipsapirone challenge: further support for Cloninger s tridimensional model of personality. Neuropsychobiology 41, Hyttel J (1982). Citalopram pharmacological profile of a specific serotonin uptake inhibitor with antidepressant activity. Progress in Neuropsychopharmacology and Biological Psychiatry 6, Hyttel J (1994). Pharmacological characterization of selective serotonin reuptake inhibitors (SSRIs). International Clinical Psychopharmacology 9, Kapitany T, Schindl M, Schindler SD, Hesselmann B, Fureder T, Barnas C, Sieghart W, Kasper S (1999). The citalopram challenge test in patients with major depression and in healthy controls. Psychiatry Research 88, Laakmann G, Gugath M, Kuss HJ, Zygan K (1984). Comparison of growth hormone and prolactin stimulation induced by chlorimipramine and desimipramine in man in connection with chlorimipramine metabolism. Psychopharmacology 82, Lesch KP, Rupprecht R, Poten B, Muller U, Sohnle K, Fritze J, Schulte HM (1989). Endocrine responses to 5-hydroxytryptamine-1A receptor activation by ipsapirone in humans. Biological Psychiatry 26, Maskall DD, Zis AP, Lam RW, Clark CM, Kuan AJ (1995). Prolactin response to buspirone challenge in the presence of dopaminergic blockade. Biological Psychiatry 38, Park SB, Williamson DJ, Cowen PJ (1995). Do the endocrine and subjective effects of d-fenfluramine predict response to

5 Citalopram and neuroendocrine challenge tests 71 selective serotonin reuptake inhibitors? International Clinical Psychopharmacology 10, Power AC, Cowen PJ (1992). Neuroendocrine challenge tests: assessment of 5-HT function in anxiety and depression. Molecular Aspects in Medicine 13, Reist C, Helmeste D, Albers L, Chhay H, Tang SW (1996). Serotonin indices and impulsivity in normal volunteers. Psychiatry Research 60, Seifritz E, Baumann P, Muller MJ, Annen O, Amey M, Hemmeter U, Hatzinger M, Chardon F, Holsboer-Trachsler E (1996). Neuroendocrine effects of a 20-mg citalopram infusion in healthy males. A placebo-controlled evaluation of citalopram as 5-HT function probe. Neuropsychopharmacology 14, Storlien LH, Smythe GA (1992). d-fenfluramine effects on hypothalamic monoamine activities and their hormonal correlates. Brain Research 597, Van Praag HM, Lemus C, Kahn R (1987). Hormonal probes of central serotonergic activity: do they really exist? Biological Psychiatry 22,