Platelet Serotonin Transporter in Children and Adolescents with Obsessive-Compulsive Disorder or Tourette's Syndrome

Platelet Serotonin Transporter in Children and Adolescents with Obsessive-Compulsive Disorder or Tourette's Syndrome FLOYD R. SALLEE, M.D., PH.D., HARVEY RICHMAN, PH.D., KEVIN BEACH, B.S., GOPALAN SETHURAMAN, M.S., AND LORI NESBITT, PHARM.D.

ABSTRACT Previous studies of serotonin transporter protein (5HTPR) indexed in platelets by 3H-imipramine demonstrate reduction in children with comorbid obsessive-compulsive disorder (OCD) and Tourette's syndrome (TS). Objective: To use the 5HTPR selective ligand 3H-paroxetine and homogeneous diagnostic groups to reevaluate these findings. Method: Platelet kinetic binding parameters were evaluated using standard techniques from medication-free child and adolescent patients with OCD (n

= 18), with TS

(n

= 10), and normal controls

(n

= 19). Results:

Baseline binding capacity (Bma, )

was significantly reduced in patients with OCD (1,342 :!: 952 fmol/mg protein; p < .01) compared with normal controls (2,486 :!: 1309 fmol/mg) and TS patients (2,420 :!: 1,069 fmol/mg; p < .05). Among OCD patients who were subsequently treated on an open-label basis with selective serotonin reuptake inhibitor (SSRI), Bma, values at baseline differentiated between responders (1,718 :!: 1,041 fmol/mg) and nonresponders (802 :!: 713 fmol/mg protein; p < .05). Response to SSRI was greatest in patients with a positive family history of OCD. Among responders (n

= 10), baseline Yale-Brown

Obsessive Compulsive Scale and Bma, were positively correlated (r = .76, P = .01), as was Clinical Global Impression (r= .67, p

= .03). Conclusions:

Platelet 5HTPR capacity (Bmax ) is reduced in children and adolescents with OCD, but

not in those with TS. 5HTPR may be an indirect measure of basal serotonergic tone. J. Am. Acad. Child Ado/esc. Psychiatry, 1996,35(12):1647-1656. Key Words: serotonin transporter, platelet, obsessive-compulsive disorder, Toure-

tte's syndrome, 3H-paroxetine.

Neurobiological studies suggest that central serotonin (5-HT) dysfunction may be a feature of obsessivecompulsive disorder (OCD) and possibly of Tourette's syndrome (TS) (Insel, 1992; Leckman et al., 1995; Singer and Walkup, 1991). Evidence of 5-HT involvement comes from measurement of 5-HT breakdown products in the CSF, the use of 5-HT challenges with chemical probes and dietary manipulation, and the assessment of peripheral markers of central 5-HT function (e.g., platelet studies) (Stein and Uhde, 1995). OCD occurs commonly in association with TS, and

Accepted May 21, 1996 Dr. Sallee, Mr. Beach, and Mr. Sethuraman are with the Department of Psychiatry and Behavioral Sciences, Medical University of South Carolina, Charleston. Dr. Richman is with the Department ofPsychology and Sociology, Georgia Southwestern College, Americanus. Dr. Nesbitt is Director of the Clinical Research Center, Department of Internal Medicine, University of South Alabama School ofMedicine, Mobile. Reprint requeststo Dr. Sallee,Medical University ofSouth Carolina, Institute of Psychiatry, 171 Ashley Avenue, Charleston, SC 29425.

0890-8567/96/3512-1647$03.00/0©1996 by the American Academy of Child and Adolescent Psychiatry.

evidence suggests that some forms may represent tic spectrum-related alternative expressions of the same genetic diathesis (Pauls and Leckman, 1986). There are, however, phenomenological evidence (George et al., 1993; Holzer et al., 1994), neurochemical and neuroendocrine findings (Leckman et al., 1994, 1995), and pharmacological data (McDougle et al., 1990, 1993, 1994) that highlight the differences between tic spectrum OCD and non-tie-related OCD. These distinctions are clinically important in that treatment responsivity of OCD to selective serotonin reuptake inhibitors (SSRIs) may be different in the two groups (McDougle et al., 1993, 1994). Collateral data supporting 5-HT in OCD come from challenge studies to exacerbate OCD symptoms with oral m-chlorophenylpiperazine (m-CPP) (Goodman et al., 1995; Zohar and Insel, 1987), or with metergoline during SSRI treatment (Benkelfat et al., 1989), but not every study is positive (Goodman et al., 1995). CSF studies in OCD are likewise mixed, with 5-hydroxyindoleacetic acid (5-HIAA), the principal

j, AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 35:12, DECEMBER 1996

1647

SALLEE ET AL.

5-HT metabolite, elevated (Insel et al., 1985) or unchanged (Leckman et al., 1995; Lydiard et al., 1990) compared with control. CSF 5-HIAA, however, is known to decrease in concert with clomipraminerelated symptom reduction in OCD (Thoren et al., 1980). Evidence for 5-HT involvement in TS is less convincing, although early CSF studies reported low 5-HIAA (Butler et al., 1979; Cohen et al., 1978). Recent studies have failed to replicate this finding (Leckman et al., 1995). Postmortem TS brain shows low levels of 5-HT and its precursor tryptophan in the basal ganglia (Anderson et al., 1992). Peripheral blood studies of plasma tryptophan (Leckman et al., 1984) and whole blood serotonin show reductions in TS (Comings et al., 1990). As serotonin transmission may differ along an age continuum (Seifert et al., 1980; Young et al., 1984), it is important to study children and adolescents with OCD and TS. Serotonergic abnormalities in children with OCD have been evaluated primarily with peripheral blood markers, either whole blood 5-HT or indirectly through platelet-related parameters such as 3H_ imipramine. Direct examinations of whole blood 5-HT levels in juvenile OCD (Hanna et al., 1991, 1995) have not distinguished these patients from controls, yet Hanna et al. (1991) determined that positive family history of OCD led to significantly higher 5-HT than "sporadic" OCD or controls. Hanna et al. (1995) found that whole blood 5-HT correlates negatively with aggressive scores on the Child Behavior Checklist and that OCD with disruptive behavior has significantly lower 5-HT levels. 3H-Imipramine binding, previously used to characterize serotonin transporter protein (5HTPR) on platelets of adolescents and children with major depression (Ambrosini et al., 1992; Rehavi et al., 1984), indexes many binding sites, only one of which is 5HTPR. Normally the selectiviry of binding demanded in such studies is carried by the index ligand, but 3H-imipramine binding selectivity is defined by a reference ligand (clomipramine or desipramine) for nonspecific baseline. At least two of the positive studies in OCD (Weizman et al., 1986, 1992) defined 3H-imipramine binding using clomipramine, while most of the negative studies used desipramine baselines. At least seven 3H-imipramine binding studies have been completed in OCD adults, with four studies lacking significant findings and three showing reduced binding capacity (BmaJ

1648

J.

(Stein and Uhde, 1995). Weizman et al. (1992) found decreased 3H-imipramine binding capacity (BmaJ in platelets from children with TS and comorbid OCD. The data suggest a 5HTPR abnormality but are confounded by use of a nonspecific index ligand and mixed diagnostic groups (TS+OCD). It is unknown now whether reduced 5HTPR density reported (Weizman et al., 1992) is a characteristic of OCD, TS, or comorbid TS+OCD due to the small sample size comparison group with TS only. The evidence for 5-HT involvement in OCD is strong, stemming in part from the effectiveness of SSRI in the disorder in adults, but child studies are limited (DeVane and Sallee, 1996). There are four trials of SSRI in youths suffering from OCD reported in the literature (DeVane and Sallee 1996), only one of which was placebo-controlled (Riddle et al., 1992). Of the four studies, three examined fluoxetine (10 to 50 mg/ day) (Geller et al., 1995; Riddle et al., 1990) and one evaluated fluvoxamine (100 to 300 mg/day) (Apter et al., 1994). Clinical improvement ranged from 44% to 74% across all studies but was highest in the retrospective chart review of Geller and colleagues (1995). Among prospective studies, the response rate is not encouraging, but sample sizes were small (n = 36 across all prospective studies). In the one controlled trial of fluoxetine (n = 14), study duration was 20 weeks with a crossover at week 8. Scores on the YaleBrown Obsessive Compulsive Scale (YBOCS) decreased 44% after fluoxetine treatment compared with a 27% decrease with placebo, but tolerability was high (Riddle et al., 1992). No report of sertraline effectiveness in this population exists in the literature, but a shorter half-life and lack of interference with cytochrome P-450 IID6 argue for utility in treatment of OCD in children and adolescents (DeVane and Sallee, 1996). As response to SSRI may differ between OCD spectrum versus non-tic-related OCD, we have attempted to define response to sertraline in an openlabel 6-week trial in youths suffering from non-ticrelated OCD. As the OCD group we have selected is negative for comorbid TS, it would afford the opportunity to define platelet 5HTPR in homogeneous diagnostic groups of TS, normal control, and non-ticrelated OCD. The index ligand 3H-paroxetine labels 5HTPR selectively, is resistant to interference from circulating plasma concentrations of antidepressant (Mellerup and Plenge, 1990), and has been used in

AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 35:12, DECEMBER 1996

SEROTONIN TRANSPORTER IN OeD OR TOURETTE'S

platelet studies of major depression (Owens and Nemeroff, 1994) and anxiety (Iny et al., 1994). We chose 3H-paroxetine to define kinetic binding parameters of platelets from medication-free OCD, TS, and normal control subjects. Guided by Weizman et al. (1992), our specific hypothesis was that OCD would demonstrate reduced binding capacity (Bmax ) of 5HTPR in this population. In an exploratory fashion, we also wished to evaluate any association between OCD family history, symptom severity, onset, and 5HTPR parameters. An OCD subgroup was concurrently enrolled in an openlabel safety study of sertraline, where patients were evaluated by clinical response at week 6 relative to their baseline 5HTPR kinetic parameters. The rationale of this approach was to determine whether 5HTPR kinetic parameters could be related to either symptom severity or treatment outcome at week 6 to sertraline.

Exclusion Criteria. Specific exclusion criteria were as follows: (l) pregnant or nursing females (serum human chorionic gonadotropin-negative); (2) females of childbearing potential who were sexually active and not using contraception; (3) patients meeting DSM-III-R criteria for both OCD+TS, bipolar disorder, major affective disorder, panic and/or generalized anxiety, schizophrenia, anorexia/bulimia, or autism; (4) patients currently (within 6 months) known to abuse or to be dependent on any drug, including alcohol, or those who tested positive on urine drug screen; and (5) patients requiring concomitant therapy with a psychotropic drug of any kind or any drug with a psychotropic component. Half of the TS patients also had comorbid ADHD (n = 5), and four patients with OCD also had ADHD. Comorbid conditions excluded included TS for the OCD group and OCD for the TS group. At least four OCD patients had failed prior SSRI therapy (Huoxerine n = 3, paroxetine n = 1, clomipramine n = 2). To the OCD group only, a multiple ascending-dose pharmacokinetic trial of sertraline was offered after baseline assessment and platelet sampling; of 18 OCD patients, 17 agreed to participate (see "Serrraline Treatment Protocol"). At least six of the TS patients had been previously treated with neuroleptic medications. All patients were medication-free for at least 2 weeks prior to platelet assessment, with the exception of Huoxerine- and neuroleptictreated patients, who were medication-free for at least 4 weeks.

METHOD

Assessment of OGD and TS

Subjects Nineteen controls (mean age 12.7 ::!:: 3.9 years), 18 OCD patients (mean age 13.4 ::!:: 2.7 years), and 10 TS patients (mean age 12.2 ::!:: 2.2 years) agreed to participate in the study. Patients were elicited from tertiary outpatient psychiatric clinics specializing in either anxiery disorder or tic and Tourette's disorders. After informed consent procedures were administered, patients and controls received a complete medical and psychiatric evaluation. Included were a physical examination and a structured psychiatric interview by a child psychiatrist (F.R.S.). Evaluation procedures included the Schedule for Affective Disorders and Schizophrenia for School-Age Children (Chambers et al., 1985) modified for the anxiety disorders. The distribution of male/female and prepubertal to adolescent subjects is illustrated in Table 1. Controls (n = 19) of approximately the same age, sex, and race as the TS and OCD patients were also evaluated. There were 12 male and 7 female controls, of which 4 were Afro-Americans and 15 were Caucasians (mean age 12.7 ::!:: 3.9 years) (Table 1). Normal controls were solicited from a pool of subjects participating in metabolism and endocrinological evaluations at the health center and were compensated for their participation. Controls with anxiety (particularly OCD), depression, substance abuse, tic or Tourerte's disorder, or a family history of OCD, TS, or chronic motor tic disorder were specifically excluded from participation. No control had either an active Axis I diagnosis or a positive family history of OCD, TS, or chronic motor tic. Inclusion Criteria. Specific inclusion criteria were as follows: (1) male and female outpatients aged 9 through 17 years; (2) DSMIII-R primary diagnosis of either Tourette's disorder or OCD (patients with comorbid TS+OCD were not included; comorbid attention-deficit hyperactivity disorder on Axis I was allowed); and (3) normal baseline laboratory studies, negative urine drug screen, and normal results on physical examination.

Tic severity assessment was made with standardized and validated instruments (Walkup et al., 1992), including the Tourette Syndrome Global Scale (Harcherik et al., 1984), the Yale Global Tic Severity Scale (Walkup er al., 1992), and the Clinical Global Impression (CGI) Tic Severity Scale (Cohen et al., 1985). These tic severity measures have high interrater and intertest reliability and are highly correlated with one another (Kurian et al., 1993; Walkup er al., 1992). OCD severity assessment included the Child Yale-Brown Obsessive Compulsive Scale and the CGI-OCD severity scale. These OCD severity measures have well-known psychometric properties, and they exhibit high interrater reliabilities (Goodman et al., 1989).

Sertraline Treatment Protocol This study was designed as an open-label, fixed ascending-dose study primarily to detect pharmacokinetic and safety parameters. OCD patients (n = 17) were treated to a maximum tolerable dose or 200 mg/day sertraline. Pharmacokinetic steady-state dose and area under the plasma concentration curve were obtained at weekly intervals for the first 4 weeks, and then finally at week 6. Initial treatment was started at 50 mg/day as a single dose with the evening meal and increased weekly as tolerated by 50 mg/week until 200 mg/day or maximum tolerability was achieved. Patients were maintained on 200 mg/day during weeks 5 and 6 and were evaluated clinically by YBOCS and CGI at endpoint (week 6). Clinical ratings were performed by raters who were blind to pharmacokinetic data. Responders were defined as having (2-point drop in CGI severity rating from baseline at week 6. There were no dropouts, and all but one patient received the full 200-mg/day dose by week 6.

Platelet Preparation and Binding Assay Platelet samples were collected from both patients and controls between 1 P.M. and 6 P.M. during the months of March through

]. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 35:12, DECEMBER 1996

1649

SALLEE ET AL.

TABLE 1 Subject Demographics and Clinical Characteristics Findings Characteristic

Normal Controls (n = 19)

Current age, mean (::'::SD) SES, mean (::'::SD)a Sex, No. (%) Male Female Prepubertal Adolescent Comorbid ADHD YBOCS, mean (::'::SD) TSGS, mean (::'::SD) YGTSS, mean (::'::SD) CGI severity OCD, mean (::'::SD) CGI severity TS, mean (::'::SD) Duration of illness (months), mean (::'::SD) Family history OCD TS/CMT

12.7 (3.9) 36.7 (10.9) 12 7 8

(63) (37)

(n

TS = 10)

12.2 (2.2) 35.9 (14.1) 7 3

(70) (30)

6

4 5

11

OCD = 18)

(n

13.4 38.1 11

7 7

(2.7) (9.1) (61) (39)

11

4 18.2

(5.6)

4.7

(1.1)

14.9

(7.5)

26.7 (10.4) 44.3 (7.6) 4.3 (0.7) 31.4 (20.8)

o o

3 2

6 0

Note: TS = Touretre's syndrome; OCD = obsessive-compulsive disorder; SES = socioeconomic status; ADHD attention-deficit hyperactiviry disorder; YBOCS = Yale-Brown Obsessive Comulsive Scale; TSGS = Tourette Syndrome Global Scale; YGTSS = Yale Global Tic Severity Scale; CGI = Clinical Global Impression; CMT = chronic motor tic disorder. a Hollingshead index.

August 1993 and during the same period in 1994. Baseline samples were obtained from patients before initiation of pharmacotherapy. Platelets were obtained and processed by a modification of methods published elsewhere (Sallee et al., 1995). Briefly, 40 mL of blood was collected in Vacutainer EDTA tubes at room temperature. The blood was diluted 1:1 with modified Hanks balanced salt solution at room temperature, and 40 mL of this mixture was layered over 10 mL of Ficoll Paque in a 50-mL conical tube. After a 45-minute spin in a Beckman T]-6 centrifuge (4,000 x g, 18-° to 20°C), the platelet-containing layer was harvested with a glass pipette and counted by fluorescence-activated cell sorter analysis in selected samples. Lymphocytes were removed by a 10-minute centrifugation (4,000 x g, 4°C), and the supernatant-containing platelets were subjected to further centrifugation (16,000 x g, 4°C for 10 minutes) in a Sorvall model RC28S centrifuge with a SS34 rotor. The platelet pellet was resuspended in buffer (50 mmollL Tris-HCl, 20 mmollL EDTA, 150 mmollL NaCI, pH 7.5 at 4°C) with an electric homogenizer. The suspension was centrifuged once again (40,000 x g, 4°C), yielding 95% of the total platelet population, which was then stored at -70°C until assay. 3H-Paroxetine binding was performed by using a modification the technique of Langer et al. (1985). Platelet membranes were prepared by suspending the pellet in hypotonic buffer (5 mmoll L Tris, 5 mmollL EDTA, pH 7.5 at 4°C) by homogenization causing lysis. Platelet membranes were then sequentially suspended in 70 mmollL Tris, pH 7.5 at 4°C, followed by centrifugation (40,000 x g, 4°C) and final suspension in incubation buffer (50 mmollL Tris, 5 mmollL KCl, 120 mmollL NaCl, pH 7.5) at room temperature for assay. Platelet membrane suspension (100 (L) was incubated in the presence of 3H-paroxetine (New England Nuclear NET-869; specific activity 28 Ci/mmol) at concentrations between 0.1 and 6 nmollL and brought to a final volume of 5

1650

mL with assay buffer. Six concentrations of 3H-paroxetine were used in each assay in the presence and absence of clomipramine (l (moIlL) to define nonspecific binding. Solutions were vortexed and allowed to incubate for 2 hours at room temperature. After mixing and incubation, the assay was stopped with the addition of 5 mL of ice-cold buffer. Material was collected on Whatman GF/B filter paper using a Brandel M24-R. Filter paper sections were put into scintillation vials with 10 mL of Opti-Fluor and counted in a Packard liquid scintillation counter. Data from six points were entered into EBDA software for KD (L'affiniry) and Bm", parameter estimates followed by LIGAND analysis (Munson and Rodbard, 1980). Assay precision was 10%, and interassay variability was less than 15%. Protein analysis of platelet samples was accomplished by using a modification of the Lowry er al. (1951) method to decrease interference from Tris buffer (Sigma Chemical).

Data Analysis A one-way analysis of covariance (ANCOYA) was performed to test for significant differences among the three treatment groups for platelet binding kinetic parameters Bm ", and KD including age as a covariable. If this procedure yielded a significant p value, multiple comparisons among the groups were performed using the Bonferroni correction for multiple comparison. It was observed that for the platelet binding kinetic parameter KD values, the distribution was skewed to the right side of the mean for all three groups (controls, TS, and OCD). To account for this distribution, nonparametric tests were performed initially by using the KruskalWallis test (the nonparametric equivalent of a one-way analysis of variance [ANOYA]). If groups were significantly different, subsequent pairwise comparisons using the Mann-Whitney test were performed. To address the influence of sex, ANCOYA (and/

j. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 35:12, DECEMBER 1996

SEROTONIN TRANSPORTER IN OCD OR TOURETTE'S

or Kruskal-Wallis test when appropriate) was done comparing kinetic parameters by sex and group status for the three groups with age as a covariable. This preliminary analysis of the factor sex was negative, so that subsequent analyses addressed diagnostic grouping only. OCD patients who subsequently took part in the sertraline trial were divided by clinical response defined at week 6 as a drop in CGI score from baseline by (2 points. Student's t test was performed to compare platelet kinetic parameters by sertraline response status. Linear regression of clinical parameters (YBOCS, CGI) on 3H-paroxetine platelet Bm" was also done to assess any relationship to OCD symptom severity.

5000 ,---.....,---"""T----r-------,

l::

'Q)

e

Q..

0)

:::E

<, III

3000

Q)

E

o .....

2000

LL

Demographic characteristics of both the OCD and TS patient samples and the control group are illustrated in Table 1. Socioeconomic status and age were similar across all study groups. Variables of sex distribution and pubertal status were roughly equivalent except for a preponderance of prepubertal TS patients versus control and OCD. TS patients had longer duration of illness (31.4 ± 20.8 months) compared with OCD patients (14.9 ± 7.5 months). TS and OCD patients exhibited similar levels of severity as measured by the CGI (4.3 ± 0.7 versus 4.7 ± 1.1, respectively). Family history of OCD and TS or chronic motor tic disorder is illustrated in Table 1. Of the 18 OCD patients enrolled, 9 had "early-onset" OCD, with onset of symptoms before age 14. Among patients with positive family history of OCD, four of six exhibited "early onset. " The role of age and sex on kinetic parameters (KD, BmaJ was examined by a two-way ANCOVA across all diagnostic groups. No direct or interactive effect of sex and diagnosis on the kinetic parameters was found. For the covariable age, the test of the hypothesis of non-null slope did not attain significance for any group. Examining the data set as a whole, among males (n = 30), 3H-paroxetine KD and Bmax values were 0.09 ± 0.07 nmol/L and 1,664 ± 1,104 fmol/mg protein, respectively, while among females (n = 17) KD was 0.14 ± 0.21 nmol/L and Bmax was 2,380 ± 1,248 fmol/mg protein. In prepubertal subjects (n = 21) 3H_ paroxetine KD was 0.09 ± 0.17 nmol/L and Bmax was 2,228 ± 1,110 fmol/mg protein, while among adolescents (n = 26) the mean KD was 0.13 ± 0.12 nmol/L and Bmax was 1,752 ± 1,364 fmol/mg protein. The platelet 3H-paroxetine Bmax of controls, OCD patients, and TS patients is presented in Figure 1. Binding capacity (Bmax) by ANOVA revealed significant J.

I

•

"0

E Q)

RESULTS

• •

4000

X

ro

E co

1000

•

-+I

I

_.••

•

• •

OCD

TS

• I •

•• •• •• •• •

•

CONTROL

Fig. 1 Platelet JH-paroxetine binding capacity (Bma ) expressed as ferntomolesper milligram of protein in control subjects, medication-free obsessivecompulsive disorder (OCD) patients, and medication-free Touretre's syndrome (TS) patients. Median values for eachgroup represented by horizontal line in scatter plot. OCD Bma different from control (p :0; .OI) and TS (p:O; .OI).

differences by diagnostic group (p = .006). Post hoc analysis (Bonferroni-corrected) revealed that OCD Bmax was significantly reduced compared with controls (1,342 ± 952 versus 2,486 ± 1,309, P = .003) and compared with TS patients (1,342 ± 952 versus 2,420 ± 1,069, P = .01) (Table 2). Binding affinity (1/KD) by group (ANOVA) was not significant. KD was not normally distributed but skewed to the right. When analyzed by nonparametric methods (Kruskal-Wallis), KD suggested a significant group effect (p < .05), with OCD showing decreased affinity compared with control (0.16 ± 0.20 nmol/L versus 0.09 ± 0.12 nmol/L, respectively, p < .05) and compared with TS (0.16 ± 0.20 nmol/L versus 0.06 ± 0.04 nmol/L, respectively, p < .05) (Table 2). Mfinity data should be interpreted with caution because of its distribution. Platelet 5HTPR kinetic parameters were not differentiated in the OCD group by family history of OCD or presence of "early onset." Of the 18 OCD patients examined at medicationfree baseline, 17 subsequently completed the safety and pharmacokinetic study of sertraline. One remaining patient who declined enrollment was treated clinically with fluoxetine and was judged to be a responder at 6 weeks by criteria. Exploratory analyses within the

AM. ACAD. CHILD ADOLESC. PSYCHIATRY. 35:12, DECEMBER 1996

1651

SALLEE ET AL.

TABLE 2 Platelet 3H-Paroxetine Binding in Children and Adolescents with TS and OCD KD (nmollL) Group Control TS OCD SSRI responders SSRI nonresponders

No.

Mean

19 10 18 10 7

0.09 0.06 0.16 0.18 0.12

Bm ", (fmollmg) Mean

SD

2,486 2,420 1,342 1,718 802

0.12 0.04 0.20"",b* 0.26 0.10

Note: TS = Touretre's syndrome; OCD = obsessive-compulsive disorder; KD = l/affinity constant; Bm ", capacity; SSRI = selective serotonin reuptake inhibitor. a OCD significantly different from control. b OCD significantly different from TS. C SSRI nonresponders significantly different from responders among OCD patients. * p < .05; ** P < .01.

OCD group (n = 17) were then done to detect relationships between the variables of clinical response, family history, onset, and 5HTPR. Sertraline response by criteria was 59% (10/17), with total response in the original sample (11/18) to SSRl suggesting an effect of positive family history of OCD. Patients with a positive OCD family history were fivefold more likely to respond (5/6 responders) than those with negative family history (6/12 responders) (95% confidence interval for odds ratio = 0.62, 38). In the subgroup of sertraline-treated OCD patients (n = 17), significant differences by Student's t test in baseline platelet 3H_ paroxetine Bmax between responder (n = 10) and nonresponder (n = 7) groups were demonstrated. Nonresponders, compared with responders, had significantly reduced baseline Bmax (802 ± 713 versus 1,718 ± 1,041 fmol/mg protein, p < .05). Exploratory analysis attempting to correlate baseline YBOCS score and Bmax among sertraline responders (n = 10) was positively correlated (r = .76, P = .OI), as was CGI (r = .67, P = .03) (Fig. 2). Neither YBOCS (17.4 ± 6.7 responders versus 19.1 ± 4.4 nonresponders) nor CGI distinguished between groups at baseline. Nonresponders failed to show a correlation between clinical severity and platelet kinetic parameters.

The present study represents the first attempt to substantiate platelet 5HTPR abnormalities in OCD or TS by using the index ligand 3H-paroxetine. Because of the selectivity of 3H-paroxetine for 5HTPR, the present study points to decreased binding capacity

1652

1,309 1,069 952",,*,b* 1,041 713'* =

binding

(Bmax ) and possibly affinity (1/KD) in medication-free children and adolescents with OCD. It suggests that previous findings of reduced Bmax in children who had comorbid TS+OCD indexed by 3H-imipramine (Weizman et al., 1992) may have tapped the OCD dimension rather than TS. Weizman et aI. (1992) found decreased 5HTPR sites in platelets from children with TS and comorbid OCD (n = 9), while the same abnormality was absent in their TS-only (n = 8) group, compared with controls (n = 9). Adolescents with OCD, compared with age-matched controls and adult OCD patients (Weizman et aI., 1986), demonstrate reduced platelet 3H-imipramine binding in both OCD groups (adolescent and adult OCD). One caveat to '2 :§

3500

e Q.

3000

01

2500

(5

2000

~ ..... Gl

E

~Gl

0

1500

!:!;.

1000

~

500

E

0

!Xl

0

6

DISCUSSION

SD

8

10 12 14 16 18 20 22 24 26 28 Baseline YBOCS Score

Fig. 2 Linear regression of baseline Yale-Brown Obsessive Compulsive Scale (YBOCS) score against baseline medication-free platelet binding capacity (Bmu ) value for obsessive-compulsive disorder (OCD) patients who subsequently responded to sertraline. Response was defined at week 6 of open-label sertraline as a drop in Clinical Global Impression severity score from baseline by ~2 points. Bmu positively correlated with YBOCS (n = 10; r = .76, P = .01).

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 35: 12, DECEMBER 1996

SEROTONIN TRANSPORTER IN OeD OR TOURETTE'S

the present study's exclusion of TS+OCD is that some TS patients may later have a bout of OCD, as its expression may occur later than childhood (Rasmussen and Eisen, 1990). The TS-only group's lacking the 5HTPR abnormality in the present study mirrors findings ofWeizman et al. (1992), although the sample size in both studies is small. By extension, these findings taken together could be further evidence that TS in the presence of comorbid OCD is distinct from TS without OCD (Pauls et al., 1995). The present findings do not directly address whether 5HTPR platelet kinetic abnormalities could also be found in comorbid TS+OCD, as this group is specifically excluded. In a large study of OCD adults (N = 20), Bastani et al. (1991) examined platelet 5-HT uptake and 3H_ imipramine binding and found significantly reduced Bmax in OCD patients compared with normal controls with decreased affinity for 5-HT uptake (Km). These ·same authors sought to correlate platelet kinetic binding parameters with clinical severity of OCD. The Bastani et al. (1991) study used the Maudsley ObsessiveCompulsive Inventory, which negatively trended with the affinity (1/KD) of platelet 3H-imipramine binding. Negative studies of platelet 3H-imipramine binding in adult OCD include those of Insel et al. (1985), Black et al. (1990), Kim et al. (1991), and Vitello et al. (1991). Vitello et al. (1991) were unable to substantiate OCD-related 3H-imipramine binding abnormalities, but they did note significantly increased 5-HT uptake capacity (Vmax) in OCD patients versus controls. Black et al. (1990), in a large study of adults with OCD (N = 22), could find no platelet 3H-imipramine binding abnormalities except after treatment with clomipramine. While adult OCD studies appear mixed, similar studies of childhood OCD show reduced 3H-imipramine platelet binding capacity (BmaJ (Weizman et al., 1986, 1992). This inconsistency may be due to several factors: (1) that child and adolescent OCD as in Weizman et al. (1992) may be more genetically loaded, more tic-related, and of "early onset" (Leonard et al., 1993; Pauls et al., 1995); (2) small sample sizes; (3) use of nonselective 3H-imipramine as an index ligand that sometimes fails to correlate with 3H-paroxetine and 5HTPR; (4) contamination by circulating plasma levels ofSSRI and tricyclic antidepressants; (5) potential age-, sex-, season-, and stress-related variable interaction with 5HTPR. The present study addresses most of these factors, with the possible exception of stress.

J.

Iny et al. (1994) suggest that decreased platelet 5HTPR indexed by 3H-paroxetine is a stress response variable, and they describe its reduction in anxiety disorders. Plasma cortisol level was not, however, correlated with kinetic parameters of either 3H-imipramine or 3H-paroxetine binding (Iny et al., 1994). Bmax of 3H-paroxetine was correlated (r = - .72, P < .01) with the Hassles scale and with anxiety ratings (Iny et al., 1994). Reduced 3H-paroxetine Bmaxwas negatively correlated with state-dependent anxiety score in posttraumatic stress disorder (Arora et al., 1993), and Bmax is also reduced in panic (Faludi et al., 1994). In the present study, while the OCD group overall lacked any association between OCD symptoms and 3H_ paroxetine binding parameters, the sertraline responder subgroup exhibited a significant positive association between baseline YBOCS score and Bmax. If the present findings were entirely stress-related, there should have been a negative correlation. In the sertraline responder group alone there may exist a relationship between 5HTPR capacity and symptom severity at baseline as both may be indirect measures of 5-HT basal tone (T Bmax and increased OCD symptoms reflect compensatory response to J, 5-HT tone) (Jacobs and Fornal, 1995). For the OCD group as a whole, reduced 5-HT tone is reflected in reduced 5HTPR Bmax relative to normal controls. In an analogous system, a gene knockout of the dopamine transporter results in compensatory down-regulation of the entire dopamine transmission system, which then becomes exquisitely sensitive to exogenous dopamine (Giros et al., 1996). Upon initiation of SSRI therapy or in response to 5-HT challenge (e.g., m-CPP), OCD symptoms should be exacerbated if the 5-HT system is hypersensitive. Among clomipramine nonresponders, most of whom exhibit "early onset," OCD symptoms are exacerbated by initial clomipramine infusion at baseline, indicating 5-HT system hypersensitivity (Mundo et al., 1995). "Early onset" is determined in part by genetic loading of OCD (Bellodi et al., 1992; Pauls et al., 1995) and is associated with 5-HT hypersensitivity and decreased system plasticity over time (Mundo et al., 1995). Leonard et al. (1993), in a 2to 7-year follow-up of 54 children and adolescents with OCD, found that lack of response to 5 weeks of clomipramine therapy (perhaps indicating 5-HT hypersensitivity) is the most powerful predictor of poor OCD outcome, while comorbid tics and/or presence

AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 35:12, DECEMBER 1996

1653

SALLEE ET AL.

of parental Axis I psychiatric diagnosis are less predictive. In Leonard et al. (1993), age of onset accounted for less than 1% of the variance of OCD severity at follow-up. Our sample suggests that positive family OCD history may be associated with positive SSRI response. Although factors of genetic OCD loading, "early onset," and poor response to treatment appear to be interrelated, their relationship to 5-HT system plasticity has yet to be elucidated. Stress may play more of a role in TS. We obtained an indirect measure of stress, the lymphocyte cytosolic glucocorticoid receptor (GCu ) level (Yehuda et al., 1995), in a subsample of TS patients reported here, and we found it to be negatively associated with 3H_ paroxetine platelet Bmax (n = 7; r = -.73, P = .06) (Sallee et al., unpublished data). This same negative correlation of GCu and 5HTPR is also present in adolescent depression, where stress is thought to play a role (Sallee et al., unpublished data). The issue of season, sex, and age effect on platelet 5HTPR parameters is limited by using the selective ligand 3H-paroxetine (Andersson and Marcusson, 1990), but Klompenhouwer et al. (1990) have noted reduced Bmax in males compared with females with this ligand. Bastani et al. (1991) found a negative effect of age on 3H-imipramine KD and age in normal controls (n = 32; ( = - .38, P = .03) but not in adults with OCD, and no effect of sex in either group. Black et al. (1990) found a positive association of 3H_ imipramine-related Bm ax and age among adults with OCD but not in controls. In the present study neither sex, pubertal status, nor age effect could be determined among the group as a whole. Season was controlled for at the outset, with age used as a covariate in the main analyses. In previous studies of platelet 3H_ imipramine binding in child and adolescent major depressive disorder, Ambrosini et al. (1992) failed to find an effect of pubertal status. Our own studies of 3H-paroxetine binding parameters in controls and adolescents with major depressive disorder have failed to find any association with sex, age, or pubertal status, which appears to be in concert with the current study. The main site of pharmacological action of the SSRI class is central 5HTPR, and any abnormality at this site may have implications for treatment responsivity. The platelet is now recognized as a noninvasive model of the 5-HT nerve terminal, sharing biochemistry, gene sequences of 5HTPR, and functionality (Owens and

1654

Nemeroff, 1994). Platelet 5HTPR has been cloned and characterized (Ramarnoorthy et al., 1993) and can be readily identified by 3H-paroxetine binding (Cheetham et al., 1993; Marcusson et al., 1989). 3H_ Paroxetine binding is specific for 5HTPR on platelets and neuronal tissue (Cheetham et al., 1993), is not influenced by circulating tricyclic antidepressant or SSRI plasma levels (Mellerup and Plenge, 1990), and has recently been used to characterize 5HTPR in hyperserotonemic relatives of autistic children (Cook et al., 1993). Recent data would suggest that 3H_ paroxetine binding is highly correlated with 14C-serotonin platelet uptake and may reflect the platelet's capacity to uptake 5-HT (Maguire et al., 1993). As a result of the present findings, decreased numbers of available 5HTPR binding sites for SSRI can be ascribed to OCD. It is unconfirmed whether treatment response, or lack thereof, is attributable to low 5HTPR capacity (Bmax ) . It is unlikely that OCD or SSRI responsivity is attributable to primary protein structure of 5HTPR. Lesch et al. (1995) have recently analyzed the primary 5HTPR structure using platelet 5HTPR messenger ribose nucleic acid (mRNA) from depressed and bipolar patients. These authors used polymerase chain reaction amplification and complementary deoxyribose nucleic acid (cDNA) sequencing to discover that no proteincoding region errors could be found to explain pathogenesis. Given the present findings of reduced 5HTPR capacity, it would be important to determine the relative levels of 5HTPR mRNA available among the diagnostic groups (TS, OCD, control) in the present study. The preliminary finding of significantly lower 5HTPR capacity in SSRI nonresponders awaits definitive evaluation within the context of a double-blind treatment trial.

REFERENCES Ambrosini PJ, Metz C, Arora RC, Lee J, Kregel L, Meltzer HY (1992), Plareler imipramine binding in depressed children and adolescenrs. ] Am Acad Child Adolesc Psychiatry 31:298-305 Anderson GM, Pollack ES, Chatterjee D, Leckman JF, Riddle MA, Cohen DJ (1992), Postmortem analysis of subcortical monoamines and amino acids in Tourerte syndrome. Adv NeuroI58:123-133 Andersson A, Marcusson J (1990), [3H]Paroxetine binding in human platelets in relation to age and sex. Neurobiol Aging 11:615-618 Apter A, Rarzoni G, King RA et aI. (1994), Fluvoxamine open-label trearmenr of adolescent inpatients with obsessive compulsive disorder or depression. ] Am Acad Child Adolesc Psychiatry 33:342-348 Arora RC, Fichtner CG, O'Connor F, Crayton JW (1993), Paroxetine binding in the blood platelets of post-traumatic stress disorder parienrs. Life Sci 53:919-928

]. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 35:12, DECEMBER 1996

S E R OT O N I N TRAN SPORTER I N OCD O R TOURETTE ' S

Bastani B, Arora RC, Meltzer HY (199 I) , Serotonin uptake and imipramine bin ding in the blood platelets of obsessive-compulsive disorde r patie nts.

Insel TR, Muel ler EA, Alterman I, Linnoila M , Murphy D L (l985) , Obsessi ve-com pu lsive diso rde r and serotonin: is there a con tradiction?

Bioi Psychiatry 30:131-139 Bellodi L, Sciuto G, Dia feria G, Ronchi P, Smeraldi E (1992), Psychiatric disorders in the fam ilies of patients with obsessive-com pulsive disorder.

Psychiatry Res42:111- 120 Benkel fat C , Murph y DL, Zohar J er al. (1989), C lomipramine in obsessivecompulsive disorder. Arch Gen Psychiatry 47:840- 848 Black OW, Kelly M, Myers C , No yes R (1990), Tritiated imipramine binding in obsessive compulsive volunteers and psychiatricall y no rmal controls. Bioi Psychiatry 27:319-327 Butle r 1], Koslow SH, Seifert W E, C ap rioli RM, Singer H S (1979 ), Biogenic am ine metabolism in Tourett e synd rome. Ann NeuroI 6:37- 39 Ch ambers WJ, Puig-Antich J, Hirsch M et al. (l985), The assessment of affective disord ers in ch ildren an d adole scents by semistr uc tu red interview. Arch Gen Psychiatry 42:696- 702 Ch eetha m SC, Viggers JA, Slater NA, H eal OJ, Buckett WR (l993) , [JH]Paroxetine bind ing in rat frontal cortex strongly corr elates wit h [JH]5-HT uptake: effect of ad minis tration of variou s an tidepressant treatments. N europharmacology 32:737-743 Co hen OJ , Leckm an JF , Shaywitz BA (1985), T he T ouretre syndrome and other tics. In: The Clinical Guide to Child Psychiatry. New York: Free Press C ohen DJ , Shaywitz BA, Caparul o B, You ng JG , Bowers MB (1978), Chronic m ultiple tics of G illes de la Toure rre's d isease: CSF acid monoamine metabolites after pro ben ecid administra tion. Arch Gen

Psychiatry 35:245-250 Comings DE, H imes JA, Comings BG (l990) , An epid em iologic stud y of Tourette's syndro me in a single school distr ict. J Clin Psychiatry 51 :463- 469 Cook EH, Aro ra RC, An derson GM er aI. (l993), Platelet seroton in studies in hype rsero tonergic relatives of children with aut istic disorder. Life

Sci 52:2005-2015 DeVa ne C L, Sallee FR (1996), Selective serotonin reu prake inh ibitors in child and ado lescent psychopharmacology: a review of published experie nce. J Clin Psychiatry 57:55-66 Faludi G , T ekes K, T othfalusi L (l 994), Comparative stu dy of platelet ' H -paroxetine and ' H -im ipramine binding in panic d isorder patients and health y controls. J Psychiatry Neurosci 19:109- 113 G eller DA , Biederm an J, Reed ED et aI. (199 5), Sim ilariti es in response to Huoxetine in the trea tment of chi ldren and ado lescents with obsessivecom p ulsive d isor der . J Am Acad Child Adolesc Psychiatry 34:36- 44 George MS , Trimble MR, Ring HA , Sallee FR, Robertso n MM (199 3), O bsessions in obsessive-compulsive disor der with and without Gi lles de la Tourette 's synd rome , Am J Psychiatry 150:93-97 G iros B, Jab er M,Jones SR, Wigh tman RM, Caron M G (1996) , Hyperlocomotion and indifference to cocaine an d amphetamine in m ice lacking the dopamine transporter. Nature 379:606- 6 12 Goodman WK, M cDougle C], Price LH et al. (l995), m-C hlorophenylpiperazine in patients with obsessive-compulsive disorder: absence of sym p to m exacerba tion. Bioi Psychiatry 38:138-1 49 Go odman WK, Price LH , Rasm ussen SA et al. (l 98 9) , The Yale-Brow n Obsessive C o m p u ls ive Scale : I I. V ali dity. A rch Gen Psychiatry 46:10 12- 10 16 Hanna G L, M cCracken JT , Ca ntwell DP (l991) , Pro lact in in ch ildhood obsessive-compulsive d isor der: clin ical correlates and response to clomipramine. JAm Acad Child Adolesc Psychiatry 30:173-178 Hanna GL , Yuwiler A, Co ates J K (l99 5), W hole blood sero tonin and disruptive behaviors in juvenile obsessive-com pu lsive disorder. J Am

Acad Child Adolesc Psychiatry 34:28-35 Harcherik O F, Leckman JF , D etlor J , Cohen OJ (1984), A new instrument for clinical studies of Tourett e's syndrome. JAm Acad Child Psychia-

try 23:153- 160 Ho lzer J , Goodman WK , Price LH , Bear L, Leckman JF , H eninger GR (1994) , Obsessive compulsive disorde r with and without a chronic tic disorder: a comparison of sym p toms in 70 pat ient s. Br J Psychiatry 164:469-473 Insel T R (l992) , T oward a neu roanatom y of obsessive-compulsive disorder.

Arch Gen Psychiatry 49:739-744

Bioi Psychiatry 20:1174-1188 In y LJ, Pecknol d J , Su ranyi-C adotr e BE et al. (l994) , Studies of a neuroch emical link betw een depression, anxiety, and stress from [JH) imipramine and [' H ]paroxeti ne bin ding on h uman platelets. Soc Bioi

Psychiatry 36:281- 291 Jacob s BL, Fornal CA (l995) , Serotonin and behavior: a gen eral hypothesis. In: Psychobiology: The Fourth Generation ofProgress, Bloom FE, Kupfe r OJ , eds. New York: Raven Press Kim SW , Dysken MW, Pande y GN, Davi s GM (1991) , Platelet ' H im ipram ine bindi ng sites in obsessive-compulsive behavior. Bioi Psychia-

try 30:467- 474 Klompenhouwer JL , Fekkes 0 , van Hulst At"' , Mo leman P, Pepplinkhuizen L, M uld er PG H (l990) , Seasonal variation s in binding of'H -paroxetine to blood platel ets in healthy volunteers: indications for a gende r differ ence. Soc Bioi Psychiatry 28:509- 517 Kurian R, Como P, Deeley C , M cD ermott M , M cDerm ott MP (1993) , A pilot controlled study of Huoxerine for obsessive-com pulsive sym pto ms in children with T our erte' s syndrome. Clin Neuropbarmacol i S:167-172 Langer SZ, Scho emaker H , Segonzac A (1985), [' H]Paroxetine binding to human platelets: clues to a selectivity paradox from temperature stu d ies. Br J Phannacol 96:303-309 Leckman JF, Anderson GM, Co hen et aJ. (1984), Whole blood serotonin and tryp top han levels in Toure tte 's disorder: effects of acute an d ch ron ic clon idine treatment. Life Sci 35:2497- 2503 Leckman J F, Goo dman WK, And erson GM er al. (1995), Cerebrospinal fluid biogenic amines in obsessive compulsive disorde r, T ourerre's disorder, and healthy controls. Neuropsycbopbarmacology 12:73- 86 Leckman JF, Goodman WK, North WG et al, (1994 ), Elevated levels of CSF oxytocin in obsessive comp ulsive disorder: com parison wit h Tourerrc' s sy n d ro m e and h ealthy control s . Arch Gen Psychiatry 51:7 82-792 Leonard HL, Swedo SE, Lenane MC et aI. (l993), A 2- to 7-year followup study of 54 obsessive-co m pulsive chi ldren and ado lescents. Arch

Gen Psychiatry 50:429-439 Lesch KP, Gross J , Franzek E, Wolozin BL, Riede rer P, Murphy DL (1995) , Primary struc ture of the seroto ni n transporter in unipolar depression and bipo lar disorde r. Bioi Psychiatry 37:215-223 Lowry D H , Rosebrough HJ, Farr AL, Randall RJ (195 1), Prote in measurement with Folin phenol reagent. J Bioi Cbem 193:265- 275 Lydiard RB, Ballenge r JC , Ellinwoo d E e t al. (l990), CSF mo noamine metabolites in obsessive-compulsive disor der. Poster p resented at the Am erican Psychiatr ic Association meetin g, N ew York, M ay 5 Maguire K, Tuckwell V, Pereira A, D ean B, Singh B (1993), Sign ificant corr elation betw een 14C -5-HT uptake by and ' H -paroxetine bind ing to platele ts from healthy volu ntee rs. Bioi Psychiatry 34:356- 360 Ma rcusson JO, An dersson A, Backstrom I (1989), Drug inhibition indicates a single-site model of the 5- HT up take site/antidep ressant bin ding site in rat and human brain. Psychopharmacology 99:17-21 Mc Do ugle q , Goodman WK, Leckma n JF, Barr LC, Heninger GR, Price LH (l 99 3), The efficacy of fluvoxamine in obsessive com p ulsive disor der : effects of comorbi d chron ic ti c disorder. J Clin Psy chopharm acol 13:3 54- 3 58 Mc Dougle q, Go odman WK, Leckman JF, Lee NC, H en inger GR, Price LH (1994) , Haloperidol addition in Huvoxam ine- refrac tory obsessivecompulsive disorder: a do uble-b lind , place bo-cont rolled study in pati ents with and without tics. Arch Gen Psychiatry 5 1:302- 308 McDougle C] , Goo dman WK, Price LH er al. (1990), Ne urolep tic add itio n in Huvoxarn ine-refracrory ob sessive com pulsive disorde r. A m J Psychia-

try 146:652-654 Me llerup ET , Plenge P (l990), W hy some dep ressed pat ients may have low platelet ' H -imipram ine bind ing. Acta Psychiatr Scand 82:330-334 Mundo E, Deliodi L, Sme raldi E (1995), Effects of acu te int ravenous clomipramine on obsessive-comp uls ive symptoms an d resp on se to chronic treatme nt . Bioi Psychiatry 38:525- 531 M unson PJ, Rodbard 0 (1980) , A versarile comput erized approach for characterizatio n of ligand-bind ing systems. Anal Biocbem 107:220-239

j . AM. ACA D . C H I LD ADO LESC . PSYC H IATRY , 35:12 , DEC EM BER 19 96

1655

SALL EE ET AL. Owens MJ, Nemero ff C B (1994), Role of serotonin in the pathophysiology of depression: focus on the serotonin transpott er. Clin Chem 40:288 -295 Pauls DL, Alsobrook JP, Phil M, Goodman W, Rasmussen S, Leckman JF (1995), A family study of obsessive-compulsive disorder. A m I Psychiatry 152:76-84 Pauls DL, Leckman JF (1986), The inheritance of Gilles de la Tourette's syndrome and associated behaviors: evidence for an autosomal dominant transmission. N Engl; M ed 315 :993-997 Ramamoorrhy S, Bauman AL, Moore KR et al. (1993), Antidepressant-. and cocaine-sensitive human serotonin transporter: molecular cloning, expressio n, and chromosomal localization. Proc Natl Acad Sci USA 90:2542-2546 Rasmussen SA, Eisen JL (1990), Epidemiology of obsessive-compulsive disorderv j' Clin Psychiatry 51(2 suppl} :10- 13 Rehavi M, Weizman R, Carel C, Apter A, T yano S (1984), High-affinity 3H _ imipramine binding in plarelers of children and adolescents with major affective disorders. Psychiatry Res 13:31-39 Riddle MA, Hardin MT, King R et al. (1990), Fluoxetine treatment of children and adolescents with T ourette's and obsessive compulsive disorders: preliminary clinical experience. ; Am Acad Child Adolesc Psychiatry 29 :45- 48 Riddle MA, Scahill L, King RA er al. (1992), Double-blind trial of f1uoxetine and placebo in children and adolescents with obsessive-compulsive disorder. ; Am Acad Child A dolesc Psychiatry 3 1:1062- 1069 Sallee FR, Nesbitt L, Do ugherry D,-H ilal R, Nandagopal VS, Sethu raman G (1995), Lymphocyte glucocorticoid receptor: predictor of sertraline response in adolescent MDD . Psychopharmacol Bull 3 1:339- 345 Seifert WE, Foxx JL, Buder IJ (1980), Age effect on dopamine and serotonin metabolite levels in cerebrospinal fluid. An n NeuroI8 :38-42 Singer HS, Walkup JT (1991), T ourette syndrome and other tic disorders: diagnosis, pathophysiology, and treatment. M edicine 70:15-32

1656

Stein MB, Uhde TW (1995), Biology of anxiery disorders. In: The American Psychiatric Press Textbook ofPsychopharmacology, Scharzburg AF, Nemeroff CB, cds. Washington , DC: American Psychiatric Press Thoren P, Asberg M, Berrilsson L et al. (1980), Clomipramine treatment of obsessive- compulsive disorder: II. Biochemical aspects. Arch Gen Psychiatry 37 :1289- 129 5 Vitello B, Shimon H , Behar D, Stoff D, Bridger WH, Friedman E (1991), Platelet imipramine binding in obsessive-compulsive patients. A cta Psychiatr Scand 84:2 9- 32 Walkup JT , Rosenberg LA, Brown J, Singer H (1992), T he validiry of instruments measuring tic severity in Tourerte's syndrome. ; A m Acad Child Adolesc Psychiatry 3 1:472-477 Weizman A, Carmi M , Her mesh H et al. (1986), H igh-affiniry imipramine binding and serotonin uptake in platelets of eight adolescent and ten adult obsessive-compulsive patients. Am; Psychiatry 143:335-339 Weizman A, Mandel A, Barber Y er al. (1992), Decreased platelet imipramine binding in Touretre syndrome children with obsessive-compulsive disorder. BioI Psychiatry 31:705- 711 Yehuda R, Boisoneau D, Lowy MT, Giller EL (1995), Dose-response changes in plasma cottisol and lymphocyte glucocorticoid receptors following dexamethasone administration in combat veterans with and withour posttraum atic stress disorder. A rch Gen Psychiatry 52:583 -593 YoungJ G, Cohen DJ, Anderson GM , Shaywitz BA (1984), Neurotransmitter ontogeny as a perspective for studies of child development and pathology. In: The Psychobiology ofChildhood, G reenhill 1, Shopsin B, eds. New York: Spectrum Publications, pp 51-84 Zohar J, Insel T R (1987), Ob sessive-compulsive disorder: psychobiological approaches to diagnosis, treatment, and pathophysiology. BioI Psychiatry 22:667- 687

]. AM . ACAD . C H ILD ADOLESC. PSYCHI ATRY , 3 5: 12 , DECEMBER 19 96