Hormone responses to fenfluramine and placebo challenge in endogenous depression

137

Psychiatry Research, 43: 137-146 Elsevier

Hormone Responses to Fenfluramine and Placebo Challenge in Endogenous Depression Pesach Lichtenberg, Baruch Shapira, Dan Gillon, Seth Kindler, Thomas B. Cooper, Michael E. Newman, and Bernard Lerer Received August 28, 1991; revised version received February 26, 1992; accepted June 6. 1992. Abstract. Plasma prolactin and cortisol levels after oral administration of d-l fenfluramine hydrochloride (60 mg) and placebo were examined in 24 endogenously depressed patients and 21 age- and sex-matched normal control subjects in a randomized, double-blind study. Prolactin levels were significantly increased by fenfluramine in both groups, but the response was significantly blunted in the depressed patients compared with the controls. This effect was partially dependent upon elevated baseline cortisol levels in the depressed group and was also influenced by a history of weight loss. Plasma cortisol levels were not increased by fenfluramine in either group. These findings confirm previous reports and suggest that patients with endogenous major depression are characterized by central serotonergic hyporesponsivity. The need to account for baseline effects on hormonal responses to putative serotonergic agents is supported by the findings; however, these effects appear to be less striking when endogenicity is a prominent clinical feature of the depressive syndrome. The apparently complex influence of weight loss on prolactin response to serotonergic challenge remains to be clarified as well as the role played by the bioavailability of the challenge drug and its metabolite. Key Words. Affective

disorder,

melancholia,

prolactin,

cortisol,

serotonin.

The role of brain serotonergic mechanisms in the pathogenesis of affective illness, particularly depression, and in the therapeutic action of antidepressant treatments remains a focus of considerable interest (Heninger and Charney, 1987; Meltzer and Lowy, 1987). Hormonal responses to pharmacologic challenges that are used as putative probes of central serotonergic function are an important avenue for evaluating these issues (Price et al., 1990). Plasma hormone release after oral administration of the anorectic drug d-l fenfluramine hydrochloride is a widely used strategy of this type. Available evidence suggests that fenfluramine acts pre-

Pesach Lichtenberg, M.D., is Senior Psychiatrist, Herzog Hospital-Ezrath Nashim. Baruch Shapira, M.D., is Director, Depression Treatment Unit, Herzog Hospital-Ezrath Nashim, and Lecturer, Hebrew University-Hadassah Medical School. Dan Gillon, M.D., is Resident in Cardiology, Hadassah-Hebrew University Medical Center. Seth Kindler, M.D., is Senior Psychiatrist, Tel Hashomer Hospital. Thomas B. Cooper, M.A., is Director, Analytical Psychopharmacology Laboratory, Orangeburg, NY. Michael E. Newman, Ph.D., is Senior Neurochemist, Biological Psychiatry Laboratory, Hadassah-Hebrew University Medical Center. Bernard Lerer, M.D., is Director, Biological Psychiatry Laboratory, Hadassah-Hebrew University Medical Center, and Associate Professor of Psychiatry, Hebrew UniversityHadassah Medical School, Jerusalem, Israel. (Reprint requests to Dr. B. Lerer, Dept. of Psychiatry, Hadassah-Hebrew University Medical Center, P.O. Box 12000, Jerusalem 91120, Israel.) 01651781/92/$05.00

@ 1992 Elsevier Scientific

Publishers

Ireland

Ltd.

138 synaptically to induce serotonin release, although both reuptake blockade and direct action on serotonergic receptors have also been reported (Costa et al., 1971; Fuxe et al., 1975). Plasma prolactin is significantly increased by fenfluramine in normal subjects (Quattrone et al., 1978, 1983; Lewis and Sherman, 1985) as well as in depressed patients (Siever et al., 1984; Shapira et al., 1989; Mitchell and Smythe, 1990) and patients with other psychiatric disorders (Lerer et al., 1988; McBride et al., 1989). This effect is dose related and is attenuated by prior administration of metergoline, which, although not specific, has potent serotonin receptor blocking activity (Quattrone et al., 1978, 1983). An increase in plasma cortisol after fenfluramine challenge has also been reported (Lewis and Sherman, 1984; Weizman et al., 1988) but has not been consistently replicated (Lerer et al., 1988). Siever et al. (1984) reported blunting of the plasma prolactin response to fenfluramine challenge in depressed patients compared with age- and sex-matched normal control subjects. Mitchell and Smythe (1990) also observed blunting of the prolactin response to fenfluramine but found the effect to be limited to patients with endogenous depression. The effect was also partially dependent upon baseline cortisol and prolactin levels which were, respectively, elevated and reduced in the depressed subjects compared with controls. Deakin et al. (1990) also found that the degree of blunting of the plasma prolactin response to serotonergic challenge (Ztryptophan) was inversely proportional to baseline cortisol levels. When depressed subjects were subdivided by Mitchell and Smythe (1990) according to the presence or absence of significant weight loss (> 5 kg), prolactin response to fenfluramine tended to be more blunted in the subjects with greater weight loss. The latter finding contrasts with the observations of Cowen and Charig (1987) and Deakin et al. (1990), who found that significant weight loss was not associated with blunting of the prolactin response to l-tryptophan challenge, while depressed patients with little or no weight loss had significantly blunted responses compared with controls. Two studies found no difference in prolactin release after fenfluramine administration between patients and controls (Asnis et al., 1988; Weizman et al., 1988). Weizman et al. (1988) reported a reduced cortisol response to the challenge agent in depressed patients, but this effect was not observed by Asnis et al. (1988) or by Mitchell and Smythe (1990) in their larger sample. Prolactin response to fenfluramine challenge was also examined by Coccaro et al. (1989) in acutely depressed patients and in remitted subjects with a history of affective disorder and was blunted in both groups compared with levels in controls. However, prolactin response to fenfluramine was also reduced in patients with a primary diagnosis of personality disorder, irrespective of whether depression was present as a current or past diagnosis. Blunted prolactin responses were associated with a history of suicide attempts in all subjects and with impulsive aggression in patients with personality disorder. Neither this study nor those of Siever et al. (1984) Weizman et al. (1988), and Asnis et al. (1988) related prolactin responses to fenfluramine to baseline cortisol levels. Thus, while three studies have reported blunted prolactin responses to fenfluramine in major depression (Siever et al., 1984; Coccaro et al., 1989; Mitchell and Smythe, 1990) the replicability of this finding (Asnis et al., 1988; Weizman et al., 1988) and its specificity for depression (Coccaro et al., 1989) are questionable.

139

Lack of specificity is also suggested by reduced prolactin release after fenfluramine challenge in medication-free patients with chronic schizophrenia (Lerer et al., 1988) and in adult autistic subjects (McBride et al., 1989) neither of whom manifested a prominent affective component. On the other hand, within the major depression category, endogenous (melancholic) features appear to be more closely associated with blunting of the prolactin response to fenfluramine (Lopez Ibor et al., 1988; Mitchell et al., 1990). In the light of these intriguing but complex findings, a further evaluation of hormonal responses to fenfluramine challenge in a relatively large group of depressed subjects, reasonably homogeneous in endogenicity and severity of depressive illness, is of considerable importance. The present study reports data of this type and also encompasses an evaluation of hormonal response to placebo challenge in the index group and to both challenges in a cohort of age- and sex-matched control subjects.

Methods Depressed patients hospitalized in the Depression Treatment Unit of the Herzog Hospital-Ezrath Nashim, Jerusalem, who met Research Diagnostic Criteria (Spitzer et al., 1978) for major depressive disorder: endogenous subtype (probable level) and had 21-item Hamilton Rating Scale for Depression (HRSD; Guy, 1976) scores above 18 were invited to

Subjects.

participate in the research protocol. All were free of psychotropic medication (besides chloral hydrate for nocturnal sedation) for at least 2 weeks before they underwent fenfluramine challenge. Thirty-seven subjects met these inclusion criteria. Of these, one woman (52 years old) withdrew consent on the day of testing and two women (65 and 70 years old, respectively) were excluded from the data analysis because of abnormally elevated baseline plasma prolactin levels (> 3 SD above the group mean). Of the remaining group of 33 subjects, 24 were included in the present data analysis on the basis of age (within 5 years) and sex matching with the concurrently recruited normal control subjects described below. The sample of depressed subjects did not overlap with that reported in a previous publication (Shapira et al., 1989) but included the 18 subjects whose fenfluramine responses were reevaluated after a course of electroconvulsive therapy (Shapira et al., 1992). Ten of the depressed patients were men and 14 were women. Their mean age was 53.4 (SD = 13.47) years (range = 26-72); 18 were unipolar and 6 bipolar; 5 were psychotic. The mean HRSD score of the group was 28.4 (SD = 5.74; range = 18-38). The duration of their depressive episodes was 14.5 (SD = 17.53 months; range = I-72 months). The control group consisted of 21 subjects (IO men and I I women; mean age = 51.4 years, SD = 18.14, range = 25-75) who were recruited from the hospital staff and by notices in the local press. They were physically healthy, were not taking medication of any type, and were determined to lack a present or past history of psychiatric disorder as well as any family history among their first degree relatives. Fenfluramine Challenge. A dose of d-l fenfluramine hydrochloride (60 mg orally) or a placebo challenge was administered after an 8-hour fast in a randomized, double-blind design. The active and placebo doses were separated by at least 48 hours. As a result of randomization, I3 of the depressed subjects and 10 of the controls were challenged initially with placebo and thereafter with fenfluramine. The rest of the subjects were challenged in the reverse order. The procedure for the challenge was as follows: At 8 a.m. on the morning of testing, an intravenous catheter was inserted in a forearm vein and kept patent with a heparin lock. Forty-five minutes later, two baseline blood samples (separated by I5 min) were drawn, and the challenge medication was then administered. Blood sampling continued hourly for the next 6 hours into EDTA-treated tubes. Light, nontryptophan-containing, caffeine-free snacks

140 were permitted at 1 hour and 3 hours after administration of the challenge drugs (after blood samples had been drawn). After completion of the test, blood samples were centrifuged and the plasma extracted and kept frozen at -80 “C until hormone assays were performed. Laboratory Assays. These were performed on coded samples in balanced fashion with fenfluramine and placebo challenge samples from depressed and control subjects assayed on the same day. Prolactin levels were determined by radioimmunoassay using kits supplied by CIS (France) (interassay variation 6.3%, intra-assay variation 5.4%) and cortisol levels were determined using kits supplied by Diagnostic Products Corporation (USA) (interassay variation 4.570, intra-assay variation 3.0%). In addition, aliquots of plasma from 20 of the depressed patients and 5 of the control subjects were frozen separately at -80 ‘C for determination of plasma levels of fenfluramine and its metabolite, norfenfluramine. These were dispatched by air freight, on dry ice, to the Analytical Psychopharmacology Laboratory, Nathan Kline Institute, Orangeburg, NY, and arrived in a fully frozen condition. Fenfluramine and norfenfluramine were quantified by gas chromatography with nitrogen detection as previously described (Krebs et al., 1984; Shapira et al., in press). Statistical Analysis. Since the two baseline prolactin and cortisol levels did not differ significantly, these were combined into a single mean value. Because of considerable variance in the plasma hormone levels, log,, transformation of the data for each subject was performed so as to permit parametric analyses by multivariate analysis of variance (MANOVA) with repeated measures and t tests on individual time points when overall significance was demonstrated. Baseline values of cortisol, which differed significantly between the groups, were entered as covariates, and differences between the groups at each hour were computed taking this into account. For analyses in which the number of subjects was too small to permit parametric analysis, Mann-Whitney tests were used. Peak minus baseline (A) values were calculated by subtracting the baseline hormone level for each subject from the highest level during the subsequent 6 hours. Pearson correlation coefficients were used to determine the relationship between log,,-transformed variables. All values are given as mean + standard deviation (SD).

Results Hormonal Response in Depressed and Control Subjects. Fig. 1 shows the plasma prolactin response to fenfluramine and placebo for depressed and control subjects (untransformed data shown but log,,-transformed data analyzed). Baseline prolactin levels were similar in the two groups, including when baseline cortisol levels were covaried for. MANOVA with repeated measures showed a significant plasma prolactin response to fenfluramine challenge as compared with the effect of placebo (F = 128.9; df= 1,43; p = 0.000). The prolactin response to fenfluramine of the depressed group was significantly blunted compared with that of the control subjects (F = 4.0; df= 6, 258; p = 0.007 for group X challenge X time interaction), the difference being most prominent at the 4th (t = 2.24, p = 0.03), 5th (t = 2.99, p = O.OOS), and 6th hours (t = 2.62, p = 0.01). Covarying for baseline cortisol levels (at each hour) did not reduce significance levels except for the 6th hour (F = 3.7 1; df= 1, 42; p = 0.06). Fig. 2 shows peak minus baseline prolactin levels for the placebo and fenfluramine challenges. The response to fenfluramine was significantly blunted in the depressed group compared with the response in the control group (t = 2.39, df= 43,~ = 0.03), but this level of significance was reduced when baseline cortisol levels were covaried

141 Fig. 1. Plasma prolactin response to fenfluramine and placebo challenge in depressed (n = 24) and control (n = 21) subjects 15

FENFWRAMINE

,5,

CHALLENQE

_I

PLACEBO

CHALLENQE

1

I

0 CONTROL . DEPRESSED

0 CONTROL . DEPRESSED

P l2 . E

O-

E t 5 & a

o-

3-

I 0

1

2

3

4

5

6

0

’ p < 0.05, **

p < 0.01

1

2

3

4

5

6

TIME Ihours)

TIME [hours) (Student t test, two-tailed).

for (F = 2.84; df = 1, 42; p = 0. IO). In the depressed group, baseline cortisol levels were negatively correlated with peak minus baseline prolactin levels at a marginal level of significance (r = -0.28, p = 0.05). Baseline cortisol levels were not significantly related to baseline prolactin levels (r = 0.2 1, p = 0.5 for fenfluramine challenge day; r = 0.10, p = 0.48 for placebo challenge day).

Fig. 2. Peak minus baseline (A) prolactin responses to fenfluramine and placebo in depressed and control subjects P

E

15

5

q CONTROL

0

z

DEPRESSED

12

5

0 % t.! = z

9 6

2 WI 3 2 ?

0

5

*p<

PLACEBO

FENFLURAUINE

L

0.05 (Student t test, two-tailed).

Fig. 3 shows cortisol levels in the depressed and control subjects in response to fenfluramine and placebo challenge. Baseline cortisol levels were significantly higher in the depressed group on both the fenfluramine (depressed: 19.7 Z!C5.87; control: 14.0 + 5.49 ng/ml; t = 3.54,~ = 0.001) and placebo (depressed: 20.3 f 6.02; control: 15.2 + 6.36 ng/ml; t = 3.06, p = 0.003) challenge days. Analysis of variance with repeated measures showed a nonsignificant effect of fenfluramine versus placebo in the control group (F= 3.33; df = 1,41;p = 0.07) and none in the depressed subjects (F = 0.13; df = 1, 46; p = 0.7). After the significant baseline difference in cortisol

142 Fig. 3. Plasma cortisol responses to fenfluramine depressed patients and normal controls. FENFLURAMINE

and placebo challenge

in

CHALLENGE

-*--p__p-__-p__ o-

P

OT

01 0

1

2

3

4

TIME Ihours)

5

6

0

1

2

3

4

5

6

TIME (hours1

levels was covaried for, there was no differential response to the challenge agent between the groups (F= 2.1; df = 5, 258; p = 0.16). There were no significant differences in baseline prolactin and cortisol levels or in A prolactin response to fenfluramine between male and female subjects in either group nor between unipolar and bipolar or psychotic and nonpsychotic depressed subjects. Age correlated weakly with baseline cortisol levels (r = 0.3 1, p = 0.08) but not with baseline prolactin levels nor with A prolactin response to fenfluramine. Effect of Weight Loss. Depressed subjects were divided into subgroups with (n = 12) or without (n = 12) a history of weight loss exceeding 5 kg (on the basis of self-report). The depressed subjects with weight loss had higher baseline prolactin levels than the depressed subjects who reported no weight loss (t = 2.51, p = 0.01). This difference remained after covariation for levels of baseline cortisol (which did not differ between the depressed patients with and without weight loss). The groups did not differ in A prolactin response to fenfluramine. Compared with values in the control subjects, A prolactin response to fenfluramine emerged as significant only for the depressed patients who did not report weight loss (t = 2.1, p = 0.03). This significance was lost after covariation for baseline cortisol levels. Fenfluramine and Norfenfluramine Levels. Levels were available from subgroups of subjects only (controls: n = 5; depressed patients: n = 20). When the five controls for whom data were available were compared with five age- and sexmatched depressed subjects, there was no difference between the groups for peak plasma levels of fenfluramine (controls: 70.0 + 19.11 ng/ml; depressed subjects: 63.6 + 24.66 ng/ ml) or norfenfluramine (controls: 12.0 + 3.68 ng/ ml; depressed subjects: 16.4 f 6.08 ng/ml). The difference between peak minus baseline (A) values of prolactin in response to fenfluramine in the two subgroups approached statistical significance (Mann-Whitney U = 3.0, p = 0.06). No trend was observed toward a difference for the administration of fenfluramine versus norfenfluramine. Because of the small number of control subjects, this comparison is of limited value.

143 Within the groups of 20 depressed patients and 5 control subjects, there was neither a significant relationship between fenfluramine or norfenfluramine and prolactin levels at each of the 6 hours nor between peak prolactin levels and peak levels of the challenge drug and its metabolite. There was, however, a trend for peak minus baseline prolactin levels to be correlated with peak fenfluramine levels in the depressed group (r = 0.38, p = 0.05). Discussion The results of this study strongly support previous reports that found a significant blunting of prolactin response to fenfluramine challenge in unmedicated patients with major depressive disorder (Siever et al., 1984; Coccaro et al., 1989; Mitchell and Smythe, 1990). The clear differentiation observed between the depressed and control groups in the present sample is most likely related to the homogeneity of the subject group in terms of endogenicity. Mitchell and Smythe (1990) found significant blunting of the prolactin response to fenfluramine only in the subgroup of depressed patients with endogenous features whom they studied. Lopez Ibor et al. (1988) did not include a control group in their study but within their depressed subjects observed a significant relationship between endogenicity and prolactin response to fenfluramine. The present results contrast with those of Asnis et al. (1988) and Weizman et al. (1988) who, in considerably smaller samples, found no difference in the prolactin response to fenfluramine between depressed subjects and controls. As reported by Mitchell and Smythe (1990), baseline cortisol levels influenced prolactin response to fenfluramine challenge. This influence was, however, partial; even when baseline cortisol levels were taken into account, the degree of blunting in the depressed subjects remained significant in a comparison of the full 6-hour time course. This is consistent with the observation by Mitchell et al. (1990) that in patients with endogenous depression (in a mixed sample), reduced prolactin response to fenfluramine is not dependent on baseline cortisol levels and supports the suggestion by Lopez Ibor et al. (1988) that this measure is a strong reflection of endogenicity in depressed subjects. The influence of weight loss on prolactin response to serotonergic challenge appears to be complex. Precise degree of weight loss is difficult to establish definitively on the basis of history. In the present study, depressed patients who reported significant weight loss had higher baseline prolactin levels than those who reported little or no weight loss, but there was no significant difference between groups in A prolactin response to fenfluramine. However, compared with values for the control subjects, blunting of the prolactin response to fenfluramine was more significant in the group without a history of weight loss (although this difference was dependent on baseline cortisol levels). These findings partially support the observations of Cowen and Charig (1987) and Deakin et al. (1990) who found that depressed patients with significant weight loss did not manifest blunting of the prolactin response to I-tryptophan challenge. Moreover, Goodwin et al. (1987) showed that weight loss in normal subjects increased plasma prolactin response to I-tryptophan challenge. In contrast, Mitchell and Smythe (1990) found a tendency toward greater blunting of the prolactin response to fenfluramine in depressed

144 patients with greater weight loss, but this was dependent on elevated baseline cortisol levels. There is clearly a complex interrelationship between prolactin response to serotonergic challenge in depressed subjects and weight loss-a relationship that is further complicated by the influence of baseline cortisol levels. Further studies are needed to tease apart these interrelated factors. As we have previously observed in normal and schizophrenic subjects (Lerer et al., 1988) fenfluramine administration did not induce a significant increase in plasma cortisol levels during the 6-hour challenge period as compared with the effect of placebo. This contrasts with previous reports in normal subjects (Lewis and Sherman, 1984) and with the finding of Weizman et al. (1988) that cortisol responses to fenfluramine were blunted in depressed patients compared with controls. Neither Asnis et al. (1988) nor Mitchell and Smythe (1990) (after they covaried for baseline cortisol levels) found a significant difference between depressed and control subjects in cortisol response to fenfluramine. Unlike Mitchell and Smythe (1990) we did not find a difference between the depressed and control groups in baseline prolactin levels. The role of plasma levels of fenfluramine and its metabolite, norfenfluramine, is another methodological issue to be considered. In the present study, levels were available for only a portion of the subjects. The limited analysis permitted by the sample size showed no difference in hourly or peak levels of either the challenge drug or its metabolite between matched subgroups of the depressed patients and controls (in whom the difference in A prolactin response to fenfluramine approached statistical significance). However, a relationship between A prolactin response to fenfluramine and peak fenfluramine plasma levels was observed in the depressed group. Coccaro et al. (1989) interpreted their plasma drug level data as not influencing differences in prolactin response to fenfluramine among the groups they studied, while Siever et al. (1984) and Mitchell and Smythe (1990) did not report plasma fenfluramine or norfenfluramine levels. The extent to which bioavailability of the challenge agent influences hormone response is clearly a crucial factor to be addressed in future studies. Notwithstanding the methodological issues that have been discussed, the finding of blunted prolactin response to fenfluramine challenge in endogenously depressed patients reported here is of considerable theoretical importance, particularly in the context of our previously reported observations that both imipramine treatment (Shapira et al., 1989) and electroconvulsive therapy (Shapira et al., 1992) significantly enhanced the response. Two points must be addressed, however, in any interpretation of the present findings. The first is the degree to which prolactin response to fenfluramine challenge indeed reflects central serotonergic function. The evidence has been considered in detail by Coccaro et al. (1988) and McBride et al. (1989), with both groups of authors concluding that a serotonergic mechanism was strongly supported. However, as these authors noted, by virtue of the predominantly presynaptic mechanism of action of the challenge drug, fenfluramine studies permit limited conclusions about the synaptic localization of a putative defect in central serotonergic function uncovered by the procedure. A further and possibly related concern is the degree of specificity of the present findings to depression. As noted earlier, blunted prolactin responses to fenfluramine

145

have also been reported in medication-free schizophrenic patients (Lerer et al., 1988) in adult autistic subjects (McBride et al., 1989) and in patients with personality disorders with a prominent effective component (Coccaro et al., 1989). Tentative explanations may be advanced on two levels. From a neurochemical standpoint, the challenge may be uncovering functional abnormalities at different levels. Thus, in schizophrenic subjects who are characterized by elevated plasma serotonin levels (Garelis et al., 1975; DeLisi et al., 1981), agonist-induced downward regulation of postsynaptic serotonergic receptors in the hypothalamus may explain blunted prolactin responses to fenfluramine. In depressed patients, a presynaptic deficit in serotonin availability may be operative. The finding of blunted responsiveness in subjects with personality disorders (Coccaro et al., 1989) is less easily explained on this basis and lends itself to the hypothesis advanced by Coccaro et al. (1989) that outwardly or inwardly directed aggression rather than depression per se may be associated with reduced central serotonergic function. Resolution of these questions may well be made possible by the use of receptor-specific probes of central serotonergic function that allow separate evaluation of the different elements that contribute to the response. Acknowledgments. The research reported Mental Health grants #40734 and #43873.

was supported

in part by National

Institute

of

References Asnis, G.M.; Eisenberg, J.; van Praag, H.M.; Lemus, C.Z.; Friedman, J.M.H.; and Miller, A.H. The neuroendocrine response to fenfluramine in depressives and normal controls. Biological

Psychiatry,

24:117-120,

1988.

Coccaro, E.F.; Siever, L.J.; Klar, H.M.; Maurer, G.; Cochrane, K.; Cooper, T.B.; Mohs, R.C.; and Davis, K.L. Serotonergic studies in patients with affective and personality disorders: Correlates with suicidal and impulsive aggressive behavior. Archives of General Psychiatry,

46:587-599,

1989.

Costa, E.A.; Groppetti, A.; and Refuelta, A. Action of fenfluramine on monoamine stores of rat tissue. British Journal of Pharmacology, 41:57-64, 1971. Cowen, P.J., and Charig, E.M. Neuroendocrine responses to intravenous tryptophan in major depression. Archives of General Psychiatry, 44:958-966, 1987. Deakin, J.F.W.; Pennel, I.; Upadhyaya, A.J.; and Lofthouse, R. A neuroendocrine study of 5HT function in depression: Evidence for biological mechanisms of endogenous and psychosocial causation. Psychopharmacology, 101:85-92, 1990. DeLisi, L.; Neckers, L.; Weinberger, D.; and Wyatt, R.J. Increased whole blood serotonin concentration in chronic schizophrenic patients. Archives of General Psychiatry, 38:647-650, 1981.

Fuxe, K.; Farnebo, L.O.; Hamberger, B.; and Ogren, S.O. On the in vivo and in vitro actions of fenfluramine and its derivatives on central monoamine neurons, especially 5hydroxytryptamine neurons, and their relation to the anorectic activity of fenfluramine. Postgraduate

Medical

Journal,

55l(Suppl.

1):35-45,

1975.

Garelis, E.; Gillin. J.; Wyatt, R.J.; and Neff, N. Elevated blood serotonin concentration in unmedicated chronic schizophrenic patients. American Journal of Psychiatry, 132: 184-186, 1975. Goodwin, G.M.; Fairburn, C.G.; and Cowen, P.J. The effects of dieting and weight loss on neuroendocrine responses to tryptophan, clonidine and apomorphine in normal volunteers: Important implications for neuroendocrine investigations in depression. Archives of General Psychiatry,

441952-957,

1987.

146

Guy, W. ECDEU Assessment Rockville, MD: National Institute Heninger, G.R., and Charney, Implications for the etiology and

for Psychopharmacology. (DHEW Publication) of Mental Health, 1976. D.S. Mechanism of action of antidepressant treatments: treatment of affective disorders. In: Meltzer, H.Y., ed. Psychopharmacology: The Third Generation of Progress. New York: Raven Press, 1987. Krebs, H.A.; Chens, L.K.; and Wright, G.J. Determination of fenfluramine and norfenfluramine in plasma using a nitrogen-sensitive detector. Journal of Chromatography, 8:103-107,

Manual

1984.

Lerer, B.; Ran, A.; Blacker, M.; Silver, H.; Weller, M.P.I.; Drummer, D.; Ebstein, B.; and Calev, A. Neuroendocrine responses in chronic schizophrenia: Evidence for serotonergic dysfunction. Schizophrenia Research, I IO:405410, 1988. Lewis, D.A., and Sherman, B.M. Serotonergic stimulation of adrenocorticotropin secretion in man. Journal of Clinical Endocrinology and Metabolism, 58:458-462, 1984. Lewis, D.A., and Sherman, B.M. Serotonergic regulation of prolactin and growth hormone secretion in man. Acta Endocrinologica, 110:152-157, 1985. Lopez-Ibor, J.J.; Saiz-Ruiz, J.; and Iglesias, L.M.M. The fenfluramine challenge test in the affective spectrum: A possible marker of endogeneity and severity. Pharmacopsychiatry. 2 1:914, 1988. McBride, F.A.; Anderson, G.M.; Hertzig, M.E.; Sweeney, J.A.; Kream, J.; Cohen, D.J.; and Mann, J.J. Serotonergic responsivity in male young adults with autistic disorder. Archives

of General

Meltzer,

Psychiatry,

46:213-221,

1989.

H.Y., and Lowy, M.T. The serotonin

hypothesis of depression. In: Meltzer, H.Y., ed. Psychopharmacology: The Third Generation of Progress. New York: Raven Press, 1987. Mitchell, P., and Smythe, G. Hormonal responses to fenfluramine in depressed and control subjects. Journal of Affective Disorders, 19:43-51, 1990. Mitchell, P.; Smythe, G.; Parker, G.; Wilhelm, K.; Hickie, I.; Brodaty, H.; and Boyce, P. Hormonal responses to fenfluramine in depressive subtypes. British Journal of Psychiatry, 157:551-557,

1990.

Price, L.S.; Charney, D.S.; Delgado, P.L.; Goodman, W.F.; Krystal, J.H.; Woods, S.W.; and Heninger, G.R. Clinical data on the role of serotonin in the mechanism of action of antidepressant drugs. Journal of Clinical Psychiatry, 51 (Suppl. 4):44-50, 1990. Quattrone, A.; Direnzo, G.; Schettini, G.; Tedeschi, G.; and Scopacaso, F. Increased plasma prolactin levels induced by d-fenfluramine: Relation to central serotonergic stimulation. European Journal of Pharmacology, 49:163-167, 1978. Quattrone, A.; Tedesci, G.; Aguglia, F.; Scopacasa, F.; Direnzo, G.F.; and Annunziato, L. Prolactin secretion in man: A useful tool to evaluate the activity of drugs on central 5hydroxytryptaminergic neurones-Studies with fenfluramine. British Journal of Clinical Pharmacology,

16:471-475,

1983.

Shapira, B.; Lerer, B.; Kindler, S.; Lichtenberg, P.; Gropp, C.; Ebstein, B.; Cooper, T.B.; and Calev, A. Enhanced serotonergic responsivity following electroconvulsive therapy in patients with major depression. British Journal of Psychiatry, 160:223-229, 1972. Shapira, B.; Reiss, A.; Kaiser, N.; Kindler, S.; and Lerer, B. Effect of imipramine treatment on the prolactin response to fenfluramine and placebo challenge in depressed patients. Journal of Affective

Disorders,

16: l-4, 1989.

Siever, L.J.; Murphy, D.L.; Slater, S.; De La Vega, E.; and Lipper, S. Plasma prolactin changes following fenfluramine in depressed patients compared to controls: An evaluation of central serotonergic responsivity in depression. Life Sciences, 34:1029-1039, 1984. Spitzer, R.L.; Endicott, J.; and Robins, E. Research Diagnostic Criteria: Rationale and reliability. Archives of General Psychiatry, 35:773-782, 1978. Weizman, A.; Mark, M.; Gil-Ad, 1.; Tyano, S.; and Laron, Z. Plasma cortisol, protactin, growth hormone and immunoreactive P-endorphin response to fenfluramine challenge in depressed patients. Clinical Neuropharmacology, 11:250-256,1988.