Pharmacokinetics and time-course ofD2 receptor occupancy induced byatypical antipsychotics in stabilizedschizophrenic patients

AM Catafau Experimental Medical Sciences, Clinical Pharmacology Discovery Medicine, Psychiatry Centre of Excellence for DrugDiscovery, GlaxoSmithKline, Barcelona, Spain and Verona, Italy; Centre for Imaging in Psychiatry, CRC-Mar, Hospital del Mar,Barcelona, Spain.

MM Penengo Centre for Imaging in Psychiatry, CRC-Mar, Hospital del Mar, Barcelona, Spain.

G Nucci Clinical Pharmacokinetics Modelling & Simulation, Psychiatry, GlaxoSmithKline, Verona, Italy.

S Bullich Centre for Imaging in Psychiatry, CRC-Mar, Hospital del Mar, Barcelona, Spain.

I Corripio Psychiatry Department, Red de Enfermedades Mentales-Trastornos Afectivos y Psicóticos(REM-TAP Network) Hospital de Sant Pau, Barcelona, Spain.

E Parellada Psychiatry Department, Clinic Schizophrenia Program, Hospital Clinic i Provincial, Barcelona, Spain.

C García-Ribera Mental Health and Drug Addiction Department, Hospital del Mar and Psychiatry Institute, IMAS, Barcelona,Spain.

R Gomeni Clinical Pharmacokinetics Modelling & Simulation, Psychiatry, GlaxoSmithKline, Verona, Italy.

E Merlo-Pich Experimental Medical Sciences, Clinical Pharmacology Discovery Medicine, Psychiatry Centre of Excellence for DrugDiscovery, GlaxoSmithKline, Barcelona, Spain and Verona, Italy.

for the Barcelona Clinical Imaging in Psychiatry Group.

Journal of Psychopharmacology22(8) (2008) 882–894

© 2008 British Associationfor PsychopharmacologyISSN 0269-8811SAGE PublicationsLos Angeles, London,New Delhi and Singapore10.1177/0269881107083810

Original Papers

Abstract

The 123I-IBZM SPECT measured D2 receptor occupancy (D2RO) in chronically dosed, stabilized schizophrenic patients and its relationship withantipsychotic (AP) pharmacokinetics (PK) over time is still unclear. The aimsof this study were: 1) To define the relationship between striatal D2 receptoroccupancy (D2RO) and plasma concentration (CP) in stabilized schizophrenicpatients on clinically relevant doses using 123I-IBZM SPECT; 2) To investigatethe time course of AP-induced D2RO and corresponding CP. Forty-six schizo-phrenic patients on their clinically required doses of risperidone, olanzap-ine, clozapine or quetiapine were included. D2RO and CP were measured overtime following a sparse-sampling experimental design, and individual PKand D2RO-time profiles were estimated using a population approach.Observed striatal D2RO and CP ranges were 28–75% and 9.4–60.5 ng/mL forrisperidone, 22–84% and 8.6–89.5 ng/mL for olanzapine, 5–53% and

41.6–818.2 ng/mL for clozapine and 0–64% and 37.9–719.6 ng/mL for que-tiapine. A PK–D2RO relationship was found for the four APs. D2RO patternover time was stable for risperidone, olanzapine and clozapine but fluctuating for quetiapine. Stabilized schizo-phrenic patients show a wide range of both D2RO and CP at clinically effective doses of the four AP, suggesting that clinical response tothese AP may be maintained with D2RO below 65%. D2RO patterns over timediffer between AP. These results should be considered for accurate interpretation of D2RO measurements, proper design of studies andoptimization of drug regimens for patients on AP treatment.

Keywordsschizophrenia, D2 receptor, PK-PD model, atypical, antipsychotics, SPECT

Corresponding author: Ana M. Catafau, M.D., GlaxoSmithKline, Psychiatry CEDD, Torre Mapfre, Villa Olimpica, La Marina, 16-18, Pl. 9 B y C 08005 – Barcelona, Spain.

Email: ana.m.catafau@gsk.com

at Biblioteca de la Universitat Pompeu Fabra on December 12, 2008 http://jop.sagepub.comDownloaded from

PK/PD model for atypical antipsychotics using SPECT 883

Introduction

It has been claimed that antipsychotic drugs (AP) exert their actionby means of blockade of D2 receptors (Seeman et al., 1976).Positron emission tomography (PET) and single photon emissioncomputed tomography (SPECT) allow the in vivo assessment ofAP induced-D2 receptor occupancy (D2RO) (Nyberg et al., 1995;Kufferle et al., 1997; Dresel et al., 1998; Gefvert et al., 1998;Kapur et al., 1999; Tauscher et al., 1999). However, despite exten-sive literature on this subject, the degree of D2RO associated tosymptom control using second-generation AP in schizophrenia,their brain and plasma kinetic relationships, as well as the influenceof the scanning time within the interdose interval on the D2ROmeasurements are still poorly understood.

The relationship between D2RO measured by PET or SPECTand clinical response is a controversial issue. For haloperidol andraclopride, the likelihood of clinical response seems to increase sig-nificantly as D2RO exceeds 65–70% (Nordstrom et al., 1993;Nyberg et al., 1995; Kapur et al., 2000). However, this has not beenreplicated for second-generation AP (Pilowsky et al., 1992, 1996;Nordstrom et al., 1995; Busatto et al., 1995; Knable et al., 1997;Gefvert et al., 1998, 2001; Tauscher et al., 1999; Bernardo et al.,2001). Moreover, despite good clinical efficacy, a high variability inD2RO has been reported for both clozapine (16–75%) (Schereret al., 1994; Pilowsky et al., 1996; Pickar et al., 1996; Tauscheret al., 1997; Kapur et al., 1999) and quetiapine (0–76%) (Gefvert et al., 1998; Kapur et al., 2000). Thus, the D2RO requiredfor therapeutic response for the second-generation AP remainsundefined. Despite the SPECT technique is more widely available,data on PK–D2RO–clinical response relationship is far more lim-ited with SPECT than with PET.

Classically, AP plasma concentrations (CP) have been used indrug development as guide for dose selection and dosage intervals.However, it has been recently claimed that D2RO measurementsmay be more accurate for these purposes, particularly for those APwhere dissociation of brain and plasma kinetics has been reported,such as risperidone and olanzapine (Tauscher et al., 2002). Forother second-generation AP such as clozapine and quetiapine, it isstill needed to test if D2RO and CP measures can be correlated orare otherwise dissociated. Optimal CP levels required for increasedlikelihood of response to treatment have been suggested forrisperidone (Darby et al., 1997; Olesen et al., 1998; Mauri et al.,2001; Raggi et al., 2004; Hiemke et al., 2004), olanzapine (Olesen and Linnet, 1999; Perry et al., 2001; Fellows et al., 2003;Raggi et al., 2004; Hiemke et al., 2004), clozapine (Perry et al.,1991; Miller et al., 1994; Raggi et al., 2004; Hiemke et al., 2004) and quetiapine (Raggi et al., 2004). PET studies havedescribed a curvilinear hyperbolic relationship between CP andstriatal D2RO for risperidone (Nyberg et al., 1999; Kapur et al.,1999), olanzapine (Kapur et al., 1998, 1999) and clozapine (Kapuret al., 1999). In contrast, only one study has reported suchPK–D2RO relationship using SPECT for risperidone (Knable et al.,1997), although significant correlations have been reportedbetween AP dose in mg/kg body weight and striatal uptake orD2RO for risperidone (Knable et al., 1997; Lavalaye et al., 1999;Schmitt et al., 2002), olanzapine (Lavalaye et al., 1999; Tauscher

et al., 1999; Meisenzahl et al., 2000) and clozapine (Scherer et al.,1994).

Time of scanning after last AP dose intake may account forvariability in reported D2RO for a given AP and dose, as it has beensuggested using PET for AP with low affinity for the D2 receptor, suchas clozapine and quetiapine (Gefvert et al., 1998, 2001; Jones et al.,2000; Kapur et al., 2001). However, in most published studies scanswere obtained at a single time point after medication intake, typically12–14 h post-dose. To the best of our knowledge, no data on SPECTD2RO variability within the interdose interval is yet available.

SPECT studies on chronically dosed, stabilized patients are stillneeded to identify the degree of D2RO necessary for efficacy.Moreover, knowledge of the influence of the scanning time withinthe AP interdose interval could be relevant for an adequate interpre-tation of D2RO measurements. Any relationship between brain andplasma kinetics could be of value for dosing schedules.

In this study, a sparse-sample experimental design and populationmodelling approach were used with the following aims: first, to definethe relationship between striatal D2RO measured by 123I-IBZMSPECT and CP in stabilized schizophrenic patients on risperidone,olanzapine, clozapine and quetiapine (PK–D2RO relationship).Second, to evaluate the time-course of the D2RO profile, and third, tocompare the profiles of D2RO and CP over time for these AP.

Methods and materials

Subjects

Demographic and clinical characteristics of the patient populationare summarized in Table 1. A total of 46 patients (32 male,29.5 � 6.5 years) with DSM-IV-TR diagnosis of schizophrenia,schizophreniform or schizoaffective disorders were included in thestudy. Thirty-six of the included patients met the criteria for schizo-phrenia (86% paranoid, 8% undifferentiated, 3% disorganized and3% residual subtypes), seven for schizophreniform disorder andthree for schizoaffective disorder. All patients were required to beresponding to AP medication, defined as scoring �65 on thePositive and Negative Syndrome Scale (PANSS) (Kay, 1987) andscoring �4 on the Clinical Global Impression Scale (CGI) (Guy,1976). Severity of extrapyramidal symptoms (EPS) was assessed atinclusion by means of the Extrapyramidal Symptom Rating Scale(ESRS) (Chouinard, 1980). The subgroups of patients for each APdid not differ by age (F � 1.49, dF � 3, P � 0.23), length of illness(F � 0.19, dF � 3, P � 0.90), PANSS score (F � 0.58, dF � 3,P � 0.63), CGI score (F � 0.50, dF � 3, P � 0.69), or ESRS score(F � 0.47, dF � 3, P � 0.70). All patients had been on monother-apy for at least two months with one of the following AP at clinicallyrequired doses: risperidone (n � 12; 1.5–7 mg/day), olanzapine(n � 12; 5–30 mg/day), clozapine (n � 12; 50–600 mg/ day) andquetiapine (n � 10; 200–800 mg/day). EPS were absent or verymild in all cases. All patients had normal physical examination, 12-lead electrocardiogram, clinical chemistry and hematology findings.Pregnancy was excluded in females by means of a urine pregnancytest (Clearview HCG II; Unipath Limited). Patients had abstainedfrom taking any adjunctive medication apart from the prescribed

at Biblioteca de la Universitat Pompeu Fabra on December 12, 2008 http://jop.sagepub.comDownloaded from

884 PK/PD model for atypical antipsychotics using SPECT

AP 14 days prior to the SPECT examination, except for contracep-tives or benzodiazepines with a maximum dose equivalent to 3 mglorazepam. Only three patients on risperidone and two on quetiapinereceived benzodiazepines. None of the patients had received anydepot AP in the year before, and none was taking anticholinergicsfor drug-induced EPS. Smoking was allowed up to 20 cigarettes perday. Absence of illegal drugs of abuse was assessed by a urine drugscreening (InstaCheck Drug Screen Test, Applied Biotech/ForefrontDiagnostics, San Diego, USA). None had concurrent substanceabuse or dependence (with the exception of nicotine).

The study was approved by the Ethics Committees from allthe participating centres and the Spanish Ministry of Health, andwritten informed consent was obtained in all cases prior to inclusionin the study.

Study design

An open-labelled, randomized, parallel group, sparse-sampledesign was used. A sparse sampling design is a study design inwhich a limited number of samples are collected in each subjectand where the sampling times differ from subject-to-subject. Thisstrategy permits to properly characterize the entire D2RO time-course in the population of the individuals evaluated accounting forthe independent contribute of each subject. Each patient receivedonly one tracer injection and was scanned once due to dosimetrylimitations. A non-linear mixed effects modelling was applied tothe population data to estimate the average time-course. Then anempirical Bayesian approach was used to estimate the individualPK, and the individual D2RO time course. Each group of patientson each treatment regimen were subdivided in four subgroups ofn � 2-3 and allocated randomly to undergo the SPECT scan atone time point during the interdose interval, defined based on

known PK for each AP, that is the time of maximum AP CP (Tmax);before Tmax (pre-Tmax); at Tmax; early after-Tmax, and late after-Tmax(Tables 2 and 4). Blood samples were drawn during SPECT scanningfor AP plasma levels measurement (see Plasma Analysis section).

123I-IBZM SPECT imaging procedure

SPECT was performed using a Prism 3000S camera (PhilipsMedical Systems, The Netherlands) fitted with ultrahigh-resolutionfanbeam collimators. Subject preparation included administration ofpotassium perchlorate (8 mg/kg) before 123I-IBZM injection to min-imize radiation exposure to the thyroid. Four SPECT/MRI markers,each filled with 0.054 MBq of 123I and olive oil, were stuck on themastoids and the corners of the eyes. 123I-IBZM (AmershamHealth/Cygne, 185.2 � 13.3 MBq) was injected intravenouslyflushed with saline. A 30-min scan was acquired starting 90 minafter radioligand injection. Each frame was collected using a 360o

circular orbit, step and shoot mode every 3o, matrix 128 � 128 pix-els. A venous catheter for blood sampling was inserted on the armopposite to the radioligand injection in all subjects.

All subjects underwent a T-1 weighted 3D MRI scan for co-registration and region of interest (ROI) drawing. A superconductive1.9 Tesla system (Prestige 2T, General Electric) equipped with ahead coil was used. Images were acquired with the following param-eters: repetition time, 25 ms; echo time, 6 ms; flip angle, 28°; fieldof view, 25 � 25 cm; matrix size, 256 � 256; section thickness,2 mm with no interslice gap; and number of excitations, 1.

Image processing

Images were reconstructed using a filtered-backprojection algorithmwith a Butterworth filter (exponent � 5.0; cuttoff frequency � 0.4

Table 1 Main demographic and clinical characteristics of the patient population at inclusion

AP Group Dose n Age Gender DSM-IV-TR Illness Treatment PANSS CGI ESRS(mg/day) Diagnosis (n) length length

(years) (years)

Risperidone 4.8 � 1.8 12 28.3 � 4.2 10M, 2F S.p.(12) 5.2 � 3.2 2.5 � 2.0 42.9 � 7.5 2.1 � 0.7 3.4 � 2.5Olanzapine 12.9 � 6.8 12 28.2 � 7.2 7M, 5F S.p.(8); Sph.D.(2); 5.3 � 4.2 2.2 � 1.6 45.6 � 7.0 2.4 � 0.7 4.0 � 2.5

S.d.(1); S.r.(1)Clozapine 304.2 � 180.2 12 29.4 � 6.5 10M, 2F S.p.(5); Sph.D.(3); 6.7 � 4.7 4.2 � 3.3 46.3 � 10.7 2.1 � 0.8 2.7 � 3.8

Sa.D.(2); S.u.(2)Quetiapine 600.0 � 235.7 10 32.8 � 7.7 5M, 5F S.p.(6); Sph.D.(2); 6.5 � 8.9 1.3 � 1.0 42.3 � 8.2 2.1 � 1.0 3.0 � 2.3

Sa.D.(1); S.u.(1)

Total – 46 29.5 � 6.5 32M, 14F S.p.(31); Sph.D.(7); 5.9 � 5.4 2.6 � 2.4 44.4 � 8.4 2.2 � 0.8 3.3 � 2.8Sa.D.(3); S.u.(3);S.d.(1); S.r.(1);

All values expressed as mean � standard deviation; n: number of patients; M: male; F: female; DSM-IV-TR: Diagnostic and Statistical Manual of Mental Disordersrevised version; PANSS: Positive and Negative Syndrome Scale; CGI: Clinical Global Impresion; ESRS: Extrapyramidal Symptoms Rating Scale. For DSM-IV-TRDiagnosis, values represent classification (number of patients in each class between brackets), according to: S.p.: Schizophrenia paranoid type; S.d.:Schizophrenia disorganised type; S.r.: Schizophrenia residual type; S.u.: Schizophrenia undifferentiated type; Sph.D.: Schizophreniphorm disorder; Sa.D.:Schizoaffective disorder.

at Biblioteca de la Universitat Pompeu Fabra on December 12, 2008 http://jop.sagepub.comDownloaded from

PK/PD model for atypical antipsychotics using SPECT 885

Table 2 Reported PK parameters and SPECT scan time selection after last dose intake for each AP

AP Group Dose Reported PK parameters SPECT scan time after last AP dose intake (h)regimen

Tmax (h) T1/2 (h) Pre-Tmax Tmax Early after-Tmax Late after-Tmax

Risperidone* once/day 1.5 22 0.5 1.5 12 22Olanzapine once/day 6 30 2 4-5 17 22Clozapine twice/day 2.1 12 1 2 4 10Quetiapine twice/day 1.5 6 0.5 1.5 4 10

*Risperidone active moiety, i.e. risperidone � 9-OH-risperidone.AP: antipsychotic; Tmax: time of maximum plasma concentration; T1/2: mean plasma half-life.

cycle/pixel). Attenuation correction was performed using Chang’salgorithm (� � 0.1 cm�1). SPECT and MRI scans were registeredusing external markers manually identified on the MRI and SPECTscans. A rigid body transformation was estimated automatically byminimizing the sum of squared distances between the correspon-ding marker positions. Striatum and occipital cortex ROI weremanually drawn on the MRI.

Data analysis

The tissue ratio method was used for quantification. The specific tonon-specific partition coefficient, representing the ratio of specificbinding in the target region to non-specific binding in the referenceregion was measured by means of specific uptake ratios (SUR)from each subject as SUR � [(S-O)/O], where S and O were theaverage counts in the striatum and occipital cortex, respectively.Individual D2RO values were directly calculated from the experi-mental data as D2ROobserved � 100 · [(0.95–SURAP)/0.95], whereSURAP is the SUR obtained in patients on medication and 0.95 themean SUR value from a group of healthy volunteers previouslystudied (n � 28, 23 M, 26.2 � 5.1 years, Catafau et al., 2005), whodid not differ in age with the patients included in the present study(F � 1.51, df � 4, P � 0.21). These D2ROobserved were the phar-macodynamic (PD) input for the PK–PD analysis. PK/D2RO mod-eling approaches were used to evaluate the relationship between CPof each AP and their respective brain-induced D2RO.

PK–PD analysis

Three PK samples were drawn from each patient across the SPECTprocedure 30 min apart. Population-PK (Pop-PK) modelling wasundertaken using non-linear mixed-effect modelling as imple-mented in NONMEM (version V) to assess population PK of thefour AP tested. Inter-subject variability and residual error were alsoassessed. Empirical Bayesian methods were used to estimate individual PK parameters. For risperidone-treated patients the CPreported and modelled was the active moiety (risperidone plus itsactive metabolite 9-hydroxy-risperidone).

It has been shown that it is possible to obtain unbiased estimatesof population model parameters using sparse sampling when prior

information on drug PK are used in the data-modelling procedure(Ette and Williams, 2004). Using this approach, the CPs of eachAPs were analyzed by fixing the population estimates of volume ofdistribution (Vd), absorption rate constant (ka) and their interindi-vidual variability to the population values reported in the literature(Huang et al., 1993 for risperidone; Callaghan et al., 1999 for olan-zapine; Jerling et al., 1997 for clozapine and Kimko et al., 2000 forquetiapine). In all cases the time course of drug CP was modelledusing a one-compartment open model with first-order absorptionand elimination assuming steady state dosing. The following priorinformation was included in the Pop-PK: for risperidone, the non-compartmental PK parameters (area under the curve AUC, plasmahalf-life T1/2 and Tmax) reported by Huang et al. (1993) were usedto extrapolate the compartmental parameters Vd, clearance (CL)and ka. For olanzapine, PK values reported by Callaghan et al.(1999) for average and standard deviation for CL and Vd were used.Average ka was selected in order to reproduce the Tmax reported. For clozapine and quetiapine, the reported Pop-PK values (Jerling et al., 1997; Kimko et al., 2000) were used.

PK/D2RO modelling approaches were based on a direct linkmodel between CP and D2RO. The relationships between AP CPand D2RO were assessed using sigmoidal Emax model:

where CP was the average AP plasma concentration duringthe SPECT scan, D2ROmax the theoretical maximum displacementfor the particular AP, EC50 the CP associated with 50% of themaximum displacement, and the slope factor. After a preliminaryPK–PD assessment it was decided to fix the slope factor to 1 for allthe AP except quetiapine, for which visual inspection of thePK–D2RO relationship suggested a sharper transition between CPassociated with approximately zero D2RO and levels associatedwith moderate (around 50%) D2RO. Given the sparse samplingstrategy adopted, it was possible to estimate inter-subject variabil-ity only for the potency parameter (EC50). The measurement errorwas fixed as additive with a standard deviation of 7% D2RO, whichis half of the estimated test-retest variability studied in healthysubjects (Catafau et al., 2008).

at Biblioteca de la Universitat Pompeu Fabra on December 12, 2008 http://jop.sagepub.comDownloaded from

886 PK/PD model for atypical antipsychotics using SPECT

Table 3 Pop-PK parameters

AP Group CL (L/h) ka (h�1) Vd (L)

Estimate CV (%) Estimate CV (%) Estimate CV (%)

Risperidone 6.4 45 3.59 46 140 39Olanzapine 19.5 58 0.6 32 1150 75Clozapine 37.9 28 1.37 24 719 62Quetiapine 96 59 2.5 80 380 10

CL: clearance; ka: absorption rate constant; Vd: distribution volume; CV: inter-subject coefficient of variation, in percentage.

PK and D2RO profiles over time were compared. Using thePK–PD model parameters, simulated D2RO curves (D2ROSim) weregenerated for a single representative dose for each AP, that is6 mg/day for risperidone; 15 mg/day for olanzapine; 450 mg/day forclozapine and 650 mg/day for quetiapine.

Plasma analysis

Risperidone, 9-hydroxy-risperidone, olanzapine, clozapine and que-tiapine CP was measured at 90, 105 and 120 min after radioligandinjection, using a validated solid phase extraction – liquidchromatography – mass spectrometry/mass spectrometry method(Dear et al., 1998) by SAS analytics (Stoneleigh Deer Park,Stareton, Kenilworth, Warks). Limits of quantification were 5 ng/mLfor risperidone and 9-hydroxy-risperidone, 1 ng/mL for olanzapine,20 ng/mL for clozapine and 15 ng/mL for quetiapine.

Statistical analysis

Statistical analyses were implemented using SPSS v.12.0 (SPSS Inc.,Chicago, Ill.). Factorial ANOVA was used to compare demographicand clinical characteristics at inclusion among groups. A probabilityvalue of 0.05 was selected as significance level.

Results

PK analysis

In two patients, one on risperidone 3 mg/day and one on quetiapine200 mg/day, the CP were found to be below the limit of quantification.These two patients were not included in the PK and PK–PD datasets.Observed CP ranges were: risperidone 9.4–60.5 ng/mL; olanzapine8.6–89.5 ng/mL; clozapine 41.6–818.2 ng/mL and quetiapine37.9–719.6 ng/mL.

Pop-PK parameters, together with their inter-subject variability(CV%) for each AP are shown in Table 3.

PK–PD analysis

Striatal D2RO ranged from 28% to 75% for risperidone, 22–84%for olanzapine, 5–53% for clozapine and 0–64% for quetiapine.Individual D2RO values and corresponding CP and time of

scan after last dose administration for each patient are presented inTable 4. Model predicted D2ROCp as a function of CP are shown inFigure 1, and Pop-PK–PD parameters for each AP are shownin Table 5. PK–D2RO curves for risperidone, olanzapine andclozapine were hyperbolic with � 1, whereas for quetiapine wassigmoidal, with � 3. In agreement with the corresponding affini-ties for the D2 receptor, risperidone and olanzapine showed higherpotency at the D2 receptor, that is required lower CP to occupy 50%of the available D2 receptors (EC50) than clozapine and quetiapine.Moreover, risperidone and olanzapine showed similar EC50,whereas quetiapine showed approximately double EC50 than cloza-pine (Table 5). Risperidone showed moderate variability inestimated potency (individual EC50 from 8.3 to 42.8 ng/mL) andpopulation PK–D2RO predictions were in good agreement withobserved data (r2 � 0.60). Olanzapine showed low inter-subjectvariability in potency (individual EC50 from 17.2 to 26.1 ng/mL)and population model predictions were in very good agreementwith observed data (r2 � 0.97). Clozapine showed lower estimatedD2ROmax than risperidone and olanzapine (Table 5); high inter-subject variability in potency (EC50 ranging from 59 to 1418 ng/mL)and population PK–D2RO predictions were not in good agreementwith the observed data (r2 � 0.10). Quetiapine showed slightlyhigher estimated D2ROmax than clozapine but lower than risperi-done and olanzapine (Table 5); inter-subject variability in potencywas moderate (EC50 ranging from 386 to 856 ng/mL) and thepopulation PK–D2RO predictions were in good agreement withthe observed data (r2 � 0.60).

Co-registered MRI/SPECT images from two representativepatients on each AP treatment at Tmax and late after-Tmax are shownin Figure 2. D2ROSim patterns over time for risperidone, olanzapineand clozapine showed a maintained D2RO, ranging between 45 and68% for risperidone 6 mg/day, 53–61% for olanzapine 15 mg/dayand 32–43% for clozapine 450 mg/day. These were dissociatedfrom the PK profile over time. However, D2ROSim patterns overtime matched the corresponding PK curves for quetiapine at alldoses (Figure 3). Mean maximum D2ROSim at peak quetiapine CPwas 45% at 650 mg/day, with D2RO reaching 0% during the dosinginterval.

Discussion

PK analysis

A wide range of CP was observed at therapeutic doses of risperi-done, olanzapine, clozapine and quetiapine in clinically stabilizedschizophrenic patients, consistent with reports of high interindivid-ual PK variability for these AP (Huang et al., 1993; Byerly et al.,1996; Darby et al., 1997; Kimko et al., 2000; Harvey et al., 2001;Raggi et al., 2004; Hiemke et al., 2004).

Population parameters for the four APs and their inter-subjectvariability were in good agreement with published data for risperi-done (Huang et al., 1993; Byerly et al., 1996; Darby et al., 1997;Burns, 2001), olanzapine (Jerling et al., 1997; Callaghan et al.,1999; Burns, 2001; Chiu et al., 2004; Sidhu et al., 2006), clozapine(Byerly et al., 1996; Guitton et al., 1998; Burns, 2001) and quetiap-ine (Gefvert et al., 1998; Kimko et al., 2000; DeVane and Nemeroff,

at Biblioteca de la Universitat Pompeu Fabra on December 12, 2008 http://jop.sagepub.comDownloaded from

PK/PD model for atypical antipsychotics using SPECT 887

Tabl

e 4

Indi

vidu

al D

2RO

valu

es, C

P, a

nd s

can

time

afte

r la

st d

ose

adm

inis

trat

ion

Risp

erid

one

Olan

zapi

neCl

ozap

ine

Quet

iapi

ne

Patie

ntC P

Scan

D 2R

O(%

)Pa

tient

C P

Scan

D 2R

O (%

)Pa

tient

C PSc

an

D 2RO

(%)

Patie

nt

C PSc

anD 2R

O (%

)nu

mbe

r(n

g/m

L)tim

e (h

)*nu

mbe

r(n

g/m

L)tim

e (h

)*nu

mbe

r(n

g/m

L)tim

e (h

)*nu

mbe

r(n

g/m

L)tim

e (h

)*

160

.522

7512

26.7

256

2436

0.7

1042

3637

.910

02

34.2

0.5

3913

33.2

1765

2541

.61

3337

719.

64

493

23.2

1272

1437

.022

8426

132.

44

3938

134.

80.

50

458

.01.

545

1589

.54.

576

2720

7.9

240

3940

7.3

1.5

05

14.3

1.5

4616

22.9

4.5

5028

481.

02

4140

585.

30.

519

645

.112

7117

78.4

275

2929

9.4

105

4159

1.2

464

738

.60.

565

1816

.017

4930

476.

31

4342

324.

51.

513

832

.622

4819

19.5

2243

3128

3.1

410

4311

6.7

108

99.

40.

528

208.

64.

522

3216

3.3

115

4461

7.0

1.5

5510

53.7

1.5

5121

13.7

1734

3381

8.2

436

1114

.222

4222

24.7

2242

3471

4.8

253

2316

.42

3335

294.

310

20

The

two

patie

nts

wit

h C P

belo

w li

mits

of q

uant

ific

atio

n ar

e no

t in

clud

ed (

see

text

). C

P: P

lasm

a co

ncen

trat

ion;

D2R

O: D

2re

cept

or o

ccup

ancy

. ** Hou

rs a

fter

last

ant

ipsy

chot

ic d

ose

adm

inis

trat

ion.

at Biblioteca de la Universitat Pompeu Fabra on December 12, 2008 http://jop.sagepub.comDownloaded from

888 PK/PD model for atypical antipsychotics using SPECT

between 10–60 ng/mL for risperidone (Darby et al., 1997; Olesenet al., 1998; Mauri et al., 2001; Hiemke et al., 2004; Raggi et al.,2004), 8–80 ng/mL for olanzapine (Olesen and Linnet, 1999; Perryet al., 2001; Fellows et al., 2003; Raggi et al., 2004; Hiemke et al.,2004) and between 350 and 600 ng/mL for clozapine (Perry et al.,1991; Miller et al., 1994; Raggi et al., 2004; Hiemke et al., 2004).For quetiapine, plasma concentrations ranging between 50 and400 ng/mL have been suggested by Raggi et al. (2004) to beassociated with clinical efficacy. In the present sample of stabilizedschizophrenic patients, 75% of the patients on risperidone and olan-zapine, but only 25% on clozapine and 30% on quetiapine showedCP within these suggested ranges. For clozapine, this may beexplained because some patients were responding to low doses. Forquetiapine, the reference range could be strongly influenced by thePK sampling time since the PK and the associated D2RO peak totrough fluctuations may be highly variable due to its short half-life.

PK–PD analysis

A wide range of D2RO was observed at therapeutic doses of risperi-done, olanzapine, clozapine and quetiapine in clinically stabilized

Figure 1 PK–D2RO model: D2ROobserved (unfilled circles) and model predicted D2ROCp (thick lines) with upper and lower 95% confidence inter-val bounds (thin lines) in response to increasing CP. CP,R�9OHR, CP,O, CP,C and CP,Q are CP for risperidone active moiety, olanzapine, clozapine andquetiapine, respectively. Model estimated D2ROMAX is represented in the figure (dashed lines) and shown in table 5.

Table 5 Pop-PK-PD parameters

AP Group D2ROmax EC50 (unitless) r2

(%) (ng/mL)

Estimate* Estimate CV (%) Estimate*

Risperidone 93 23.1 56 1 0.60Olanzapine 100 22.7 20 1 0.97Clozapine 68 257 101 1 0.10Quetiapine 74 583 40 3 0.60

*Fixed values (CV not available). AP: antipsychotic; D2ROmax: percentage of maxi-mal attainable D2RO; EC50: CP inducing 50% of the D2ROmax; : curve shape param-eter; CV: inter-subject coefficient of variation in percentage; r2: proportion of thetotal variance in D2RO that is explained by the variation in CP.

2001; Burns, 2001; Cheer and Wagstaff, 2004), and validated thesparse sampling method.

The range of plasma concentration required for increased likeli-hood of treatment response has been previously reported to lie

at Biblioteca de la Universitat Pompeu Fabra on December 12, 2008 http://jop.sagepub.comDownloaded from

PK/PD model for atypical antipsychotics using SPECT 889

Figure 2 Top: Co-registered MRI/SPECT transaxial slice at the basal ganglia level from a healthy volunteer, representing normal 123I-IBZM distribution in the brain. Please note the predominant radioligand uptake in the striatum and almost negligible uptake in theoccipital cortex. Bottom: Transverse views of co-registered MRI/SPECT images showing the 123I-IBZM uptake in two representative patientson each AP treatment, corresponding to scans performed at Tmax and late after-Tmax. The colour scale on the right represents increasing degrees of radioligand uptake with lighter colours.

schizophrenic patients. Risperidone and olanzapine showed similarPK–D2RO profiles both in terms of D2ROMAX and EC50 (Figure 1),suggesting that D2RO-equipotent CP of risperidone and olanzapinecould be in the ratio of 1:1 ng/mL, as previously reported byKapur et al. (1999). CP inducing D2RO � 50% for risperidone(26.9 ng/mL) and olanzapine (22.7 ng/mL) are within the reportedtherapeutic CP ranges for these AP, suggesting that 123I-IBZMSPECT measured D2RO around 50% may be enough for clinicalefficacy. Moreover, less than 25% on risperidone and olanzapineand none on clozapine or quetiapine showed D2RO above 65%,value claimed as a lower threshold for AP efficacy for haloperidol-treated patients with first-episode schizophrenia by means of11C-Raclopride PET (Kapur et al., 2000). According to thePK–D2RO relationships found in the present study, the CP neededto reach 65% D2RO for olanzapine (42.2 ng/mL) is higher than theCP observed in 10 out of 12 patients, and for risperidone(53.7 ng/mL) is close to the highest observed CP value (Figure 1).At least for risperidone and olanzapine, a lower D2RO (i.e., 50%)seems a more reasonable marker of AP efficacy. Interestingly, using123I-IBZM SPECT Frankle et al. (2004) reported 69 � 8% and55 � 11% D2RO induced by risperidone 6 mg/day and olanzapine10 mg/day, respectively, and that at these doses, 48% D2RO would

be necessary to reduce stimulation of D2 receptors by dopamine tolevels similar to those measured in healthy subjects. Our studyresults suggest that clinical response to these AP may be main-tained with 123I-IBZM measured D2RO values below 65%.However, neither an efficacy threshold model has been empiricallyconfirmed in atypical AP, nor a direct comparison of the D2RO val-ues measured with PET and SPECT in the same patient has beenmade. Moreover, it is unclear if the threshold for inducing responseis the same as the one for maintaining it, that is in chronicallydosed, stabilized patients (Kapur et al., 1999; Nyberg et al., 1995).

For clozapine, large interindividual variability in D2RO hasalready been reported (Nordström et al., 1995a; Pickar et al.,1996), probably due to the high interindividual variations in plasmaunbound fraction for this AP and its high interindividual CP vari-ability (Darby et al., 1997). Clozapine showed the lowest varianceexplained by CP in the PK–D2RO model (10%). This may beexplained by the irregular distribution of experimental data at lowCP values, that is below 350 ng/mL, as shown in Figure 1. Similardistributions can be found in previous SPECT (Pickar et al., 1996)and PET (Kapur et al., 1999) studies. This observation supports the350 ng/mL value as the lower limit of the optimal CP range reportedfor clozapine efficacy in treatment-refractory schizophrenics.

at Biblioteca de la Universitat Pompeu Fabra on December 12, 2008 http://jop.sagepub.comDownloaded from

890 PK/PD model for atypical antipsychotics using SPECT

D2RO versus dose (mg/kg) profiles found in this study matchwith those reported by Dresel et al. (1999) and Meisenzahl et al.(2000) for risperidone and by Schmitt et al. (2002) for olanzapine.Interestingly, these authors used very similar methodology and thesame equipment to the one used in this study. D2RO versus dose(mg/kg) profiles reported by other groups using different equipment(Scherer et al., 1994; Pickar et al., 1996; Knable et al., 1997; Dreselet al., 1998; Lavalaye et al., 1999; Tauscher et al., 1999) are also inagreement, although a trend to lower D2RO values is sometimes seenin this study. To the best of our knowledge, there are no publishedSPECT studies reporting D2RO versus dose or CP for quetiapine.

Sigmoidal Emax models have been mainly reported using PET.D2ROMAX results of this SPECT study are in agreement with thosereported using PET by Kapur et al. (1999) for risperidone, olanza-

pine and clozapine, and with the only SPECT report that was foundfor risperidone in the literature search (Knable et al., 1997).Although D2ROMAX for olanzapine was 100%, it was lower forrisperidone and particularly for clozapine and quetiapine. This maysuggest partial saturability of striatal D2 receptors by these AP or,more likely, be a consequence of the low D2RO induced by the CPseen in clinical practice. Higher CP than those seen at clinically relevant doses are expected to induce complete striatal occupancy(Nyberg et al., 2002). EC50 values reported in PET studies areusually lower (3.2–10.3 ng/mL) (Kapur et al., 1998, 1999;Remington et al., 1998; Nyberg et al., 1999), than the ones foundin the present study for risperidone and olanzapine(22.7–23.1 ng/mL). Methodological differences (e.g., radioligand,image processing and quantification) are more likely to account for

Figure 3 Comparison of the steady state CP over time (left) and corresponding D2ROSim (right), plotted against time after dose intake. Curves represent mean (continuous line) and percentiles 5 and 95 (dashed lines) for a single representative dose for each AP. CP,R�9OHR, CP,O, CP,C and CP,Q are CP for risperidone active moiety, olanzapine, clozapine and quetiapine, respectively.

at Biblioteca de la Universitat Pompeu Fabra on December 12, 2008 http://jop.sagepub.comDownloaded from

PK/PD model for atypical antipsychotics using SPECT 891

this difference than biological factors. The D2RO versus CP profilesfound in the present study match with values reported by Talviket al. (2001) for clozapine and by Kapur et al. (2000a), Gefvertet al. (2001) and Tauscher-Wisniewski et al. (2002) for quetiapine,although no PK–D2RO relationship was reported by these authors.

D2RO over time was stabilized and dissociated from PK forrisperidone, olanzapine and clozapine. Olanzapine showed themaximum stability, followed by risperidone and clozapine. In con-trast, quetiapine D2RO over time was variable and associated withits PK levels over time (Figure 3). These results are in agreementwith 11C-Raclopride PET studies by Tauscher et al. (2002a), Joneset al. (2000) and Kapur et al. (1999) reporting either a slow declineor a sustained occupancy over time for risperidone and olanzapine.For clozapine, D2ROSim in this SPECT study adequately predictedthe D2RO level reported by Jones et al. (2000) 48 h after drugintake (0–18% versus 15%) and times of peak occupancy found byNordström et al. (1995) (1.6–3.2 versus 3–6 h). Similarly, transienthigh striatal D2RO within 1–3 h of the administration of quetiapine,declining rapidly after dose intake has been reported using 11C-Raclopride PET by Jones et al. (2000), Gefvert et al. (2001), Kapuret al. (2000a) and Tauscher-Wisniewski et al. (2002), in agreementwith the results of the present SPECT study. In some cases, a trendto lower D2RO values was predicted by our study in comparison toD2RO reported by other authors at the same times, which may beexplained by underlying differences of the PET and SPECT proce-dures. Further studies are planned to clarify this issue.

Theoretically, the combination of the rate of AP dissociation fromthe D2 receptor and their plasma T1/2 are likely responsible for the main-tained or transient-peak D2RO time course (Kapur and Seeman, 2000;Tort et al., 2006). Based on the relationship between CP and D2ROfound for the four AP, higher variation in D2RO is expected in thesteeper part of the PK–D2RO curve. Many of the patients on risperi-done and olanzapine in this study were in this ascending part of thecurve. Thus, the prolonged D2RO seen with olanzapine might be betterexplained by the long plasma T1/2 found for this AP. For risperidone, thepresence of the long-acting active metabolite 9-hydroxy-risperidone,responsible of the majority of the CP of the active moiety and withacknowledged longer T1/2 than the parent compound, may have con-tributed to the prolonged D2RO (Schotte et al., 1996; Seeman, 2005).

The potential influence of the degree of protein binding for eachof the four APs studied on the modelled data should be discussed.Risperidone, olanzapine and clozapine are equally highly bound toplasma proteins (90%) while the protein binding of quetiapine is of83% (Jann, 2002). Based on these data and assuming that only thefree fraction of the drug crosses the blood/brain barrier, the esti-mated relative ratio between the EC50 for risperidone, olanzapineand clozapine provides an estimate of the relative potency of thesethree compounds, assuming a similar blood-to-brain ratio.However, a fair comparison of the quetiapine potency with thepotency of the other APs would require a correction of the modelestimated EC50 by increasing this value about 8%.

Limitations of the studySeveral potential limitations of the SPECT procedure applied inthis study deserve discussion. First, D2RO should ideally be calcu-

lated as the percentage difference between the available number ofD2 receptors measured after treatment with respect to those initiallyavailable in a drug naïve or free condition in the same subject.However, scanning a patient in these conditions is difficult and toooften unfeasible. For this reason, calculation of D2RO using themean receptor availability value from a control group (oftenhealthy volunteers) is extensively used in the literature either usingSPECT (Knable et al., 1997; Dresel et al., 1999; Tauscher et al.,1999) or PET (Gefvert et al., 1998; Kapur et al., 1999; Talvik et al.,2004). Second, there is evidence using [11C]raclopride PET ofapproximately 5% decrease per decade in D2 receptor availabilityabove the age of 30 years (Rinne et al., 1993; Antonini et al., 1993).123I-IBZM specific uptake was not corrected by age in this study.However, this should not affect the results since there were nostatistical differences in age between the control group and thepatient groups. Therefore, the potential bias in the D2RO values dueto lack of age-correction, if any, should be small. Finally, a singlebolus injection and ratio method at pseudoequilibrium was used forquantification. Ideally, a bolus � constant infusion approach or amodel-based multiple-scan method would more accurately estimatethe binding potential. However, the single bolus and ratio methodat pseudoequilibrium is widely used because the radioactive doseadministered is kept to a minimum, subjects are more likely to par-ticipate in a single-scan experiment, there are far smaller errors inhead repositioning and the whole experiment is conducted in onesession, thus avoiding possible diurnal or longer term variations inreceptor measures (Laruelle, 2000). Furthermore, the bolus injec-tion approach yields acceptable results and has been extensivelyused in the SPECT and PET literature (Pilowsky et al., 1996;Dresel et al., 1999; Tauscher et al., 1999; Lavalaye et al., 1999;Nyberg et al., 1999; Talvik et al., 2001).

ConclusionsA wide range of both observed CP and 123I-IBZM SPECT meas-ured D2RO is obtained at therapeutic doses of risperidone, olan-zapine, clozapine and quetiapine in clinically stabilized schizo-phrenic patients. The findings in this study suggest that clinicalresponse to these APs may be maintained with D2RO valuesbelow 65%. The relationship between PK and striatal D2RO isadequately described by a PK–D2RO sigmoidal Emax model forthe four AP. D2RO patterns over time differ between AP, being sta-bilized and dissociated from PK for risperidone, olanzapine andclozapine, but variable and PK-associated for quetiapine. Theseresults should be considered for accurate interpretation of D2ROmeasurements, proper design of studies and optimization of APdrug regimens.

AcknowledgementsThe authors wish to thank the Psychiatry CEDD and Clinical PharmacologyDiscovery Medicine/Experimental Medicine, GlaxoSmithKline andMarató-TV3 exp. 011410 for funding the study; the Medical Departmentfrom GlaxoSmithKline Spain, particularly Juan Lahuerta and LuisGuardiola-Gómez for their support and all the patients who participated inthis study.

at Biblioteca de la Universitat Pompeu Fabra on December 12, 2008 http://jop.sagepub.comDownloaded from

892 PK/PD model for atypical antipsychotics using SPECT

ReferencesAntonini A, Leenders K L (1993) Dopamine D2 receptors in normal human

brain: effect of age measured by positron emission tomography (PET)and [11C]raclopride. Ann NY Acad Sci 695: 81–85

Bernardo M, Parellada E, Lomena F, Catafau A M, Font M, Gomez J CLopez-Carrero C, Gutierrez F, Pavia J, Salamero M (2001) Double-blindolanzapine vs. haloperidol D2 dopamine receptor blockade in schizo-phrenic patients: a baseline-endpoint. Psychiatry Res 107: 87–97

Burns M J (2001) The pharmacology and toxicology of atypical antipsy-chotic agents. Clin Toxicol 39: 1–14

Busatto G F, Pilowsky L S, Costa D C, Ell P J, Verhoeff N P L G, Kerwin R W (1995) Dopamine D2 receptor blockade in vivo with thenovel antipsychotics risperidone and remoxipride—an 123I-IBZM sin-gle photon emission tomography (SPET) study. Psychopharmacology(Berl) 117: 55–61

Byerly M J, DeVane C L, Lindsay C (1996) Pharmacokinetics of clozapineand risperidone: A review of recent literature. J Clin Psychopharmacol16: 177–187

Callaghan J T, Bergstrom R F, Ptak L R, Beasley C M (1999) Olanzapine.Pharmacokinetic and pharmacodynamic profile. Clin Pharmacokinet37: 177–193

Catafau A M, Bullich S, Danús M, Penengo M M, Cot A, Abanades S, FarréM, Pavía J, Ros D, n(2008) Test-retest variability and reliability of 123I-IBZM SPECT measurement of striatal dopamine D2 receptor availabil-ity in healthy volunteers and influence of iterative reconstruction algo-rithms. Synapse 62: 62–69

Cheer S M, Wagstaff A J (2004) Quetiapine: A review of its use in themanagement of schizophrenia. CNS Drugs 18: 173–199

Chiu C C, Lane H Y, Huang M C, Liu H C, Jann M W, Hon Y Y, Chang W H,Lu M L (2004) Dose-dependent alterations in the pharmacokinetics ofolanzapine during coadministration of fluvoxamine in patients withschizophrenia. J Clin Pharmacol 44: 1385–1390

Chouinard G, Ross-Chouinard A, Annable L, Jones B D (1980)Extrapyramidal Symptom Rating Scale (ESRS). Can J Neurol Sci 7: 233

Darby J K, Pasta D J, Elfand L, Dabiri L, Clark L, Herbert J (1997)Risperidone dose and blood level variability: Accumulation effects andinterindividual and intraindividual variability in the non-responder patientin the clinical practice setting. J Clin Psychopharmacol 17: 478–484

Dear G J, Fraser I J, Patel D K, Long J, Pleasance S (1998) Use of liquidchromatography-tandem mass spectrometry for the quantitative andqualitative analysis of an antipsychotic agent and its metabolites inhuman plasma and urine. J Chromatogr A 794: 27–36

DeVane C L, Nemeroff C B (2001) Clinical pharmacokinetics of quetiapine.An atypical antipsychotic. Clin Pharmacokinet 407: 509–522

Dresel S, Tatsch K, Dahne I, Mager T, Scherer J, Hahn K (1998) Iodine-123-iodobenzamide SPECT assessment of dopamine D2 receptor occupancyin riperidone-treated schizophrenic patients. J Nucl Med 39: 1138–1142

Dresel S, Mager T, Rossmuller B, Meisenzahl E, Hahn K, Moller H J, Tatsch K(1999) In vivo effects of olanzapine on striatal dopamine D(2)/D(3) recep-tor binding in schizophrenic patients: an iodine-123 iodobenzamide single-photon emission tomography study. Eur J Nucl Med Mol Img 26: 862–868

Ette E I, Williams P J (2004) Population pharmacokinetics I: background,concepts, and models. Ann Pharmacother 38: 1702–1706

Fellows L, Ahmad F, Castle D J, Dusci L J, Bulsara M K, Ilett K F (2003)Investigation of target plasma concentration–effect relationships forolanzapine in schizophrenia. Ther Drug Monit 25: 682–689

Frankle W G, Gil R, Hackett E, Mawlawi O, Zea-Ponce Y, Zhu Z, Kochan L D, Cangiano C, Slifstein M, Gorman J M, Laruelle M, Abi-Dargham A (2004) Occupancy of dopamine D2 receptors by theatypical antipsychotic drugs risperidone and olanzapine: theoreticalimplications. Psychopharmacology (Berl) 175: 473–480

Gefvert O, Bergstrom M, Langstrom B, Lundberg T, Lindstrom L, Yates R(1998) Time course of central nervous dopamine-D2 and 5-HT2 receptorblockade and plasma drug concentrations after discontinuation of queti-apine (Seroquel) in patients with schizophrenia. Psychopharmacology(Berl) 135: 119–126

Gefvert O, Lundberg T, Wieselgren I M, Bergstrom M, Langstrom B,Wiesel F, Lindstrom L (2001) D(2) and 5HT(2A) receptor occupancyof different doses of quetiapine in schizophrenia: a PET study. EurNeuropsychopharmacol 11: 105–110

Guitton C, Abbar M, Kinowski J M, Chabrand P, Bressolle F (1998)Multiple-dose pharmacokinetics of clozapine in patients with chronicschizophrenia. J Clin Psychopharmacol 18: 470–476

Guy W (1976) ECDEU Assessment Manual for Psychopharmacology:Publication ADM 76-338. Washington, DC: US Department of Health,Education, and Welfare

Harvey E J, Taylor D M, Flanagan R J (2001) D2 receptor occupancy underrecommended and high doses of olanzapine. J Psychopharmacol 15:213–214

Hiemke C, Dragicevic A, Gründer G, Hätter S, Sachse J, Vernaleken I,Müller M J (2004) Therapeutic monitoring of new antipsychotic drugs.Ther Drug Monit 26: 156–160

Huang M L, Van Peer A, Woestenborghs R, De Coster R, Heykants J, Jansen A A I, Zylicz Z, Visscher H W, Jonkman J H G (1993)Pharmacokinetics of the novel antipsychotic agent risperidone and theprolactin response in healthy subjects. Clin Pharmacol Ther 54:257–268

Jann M W (2002) Clinical significance of drug binding, protein binding,and binding displacement drug interactions. Psychopharmacology Bull36: 22–41

Jerling M, Merle Y, Mentre F, Mallet A (1997) Population pharmacokinet-ics of clozapine evaluated with the nonparametric maximum likelihoodmethod. Br J Clin Pharmacol 44: 447–453

Jones C, Kapur S, Remington G, Zipursky R B (2000) Transient D2dopamine receptor occupancy in low EPS incidence drugs: PET evidence.Biol Psychiatry 47(Suppl 1): S112

Kapur S, Seeman P (2000) Antipsychotic agents differ in how fast theycome off the dopamine D2 receptors. Implications for atypical antipsy-chotic action. J Psychiatry Neurosci 25: 161–166

Kapur S, Zipursky R B, Remington G, Jones C, DaSilva J, Wilson A A,Houle S (1998) 5-HT2 and D2 receptor occupancy of olanzapine inschizophrenia: a PET investigation. Am J Psychiatry 155: 921–928

Kapur S, Zipursky R B, Remington G (1999) Clinical and theoretical impli-cations of 5-HT2 and D2 receptor occupancy of clozapine, risperidone,and olanzapine in schizophrenia. Am J Psychiatry 156: 286–293

Kapur S, Zipursky R, Jones C, Remington G, Houle S (2000) Relationshipbetween dopamine D(2) occupancy, clinical response, and side effects: adouble-blind PET study of first-episode schizophrenia. Am J Psychiatry157: 514–520

Kapur S, Zipursky R, Jones C, Shammi C S, Remington G, Seeman P (2000a)A positron emission tomography study of quetiapine in schizophrenia—a preliminary finding of anantipsychotic effect with only transientlyhigh dopamine D2 receptor occupancy. Arch Gen Psychiatry 57: 553–559

Kapur S, Remington G (2001) Dopamine D(2) receptors and their role inatypical antipsychotic action: still necessary and may even be sufficient.Biol Psychiatry 50: 873–883

Kay S R, Fiszbein A, Opler L A (1987) The positive and negative syndromescale (PANSS) for schizophrenia. Schizophr Bull 13: 261–276

Kimko H C, Reele S S, Holford N H, Peck C C (2000) Prediction of the out-come of a phase 3 clinical trial of an antischizophrenic agent (quetiapinefumarate) by simulation with a population pharmacokinetic and pharma-codynamic model. Clin Pharmacol Ther 68: 568–577

at Biblioteca de la Universitat Pompeu Fabra on December 12, 2008 http://jop.sagepub.comDownloaded from

PK/PD model for atypical antipsychotics using SPECT 893

Knable M B, Heinz A, Raedler T, Weinberger D R (1997) Extrapyramidalside effects with risperidone and haloperidol at comparable D2 receptoroccupancy levels. Psychiatry Res 75: 91–101

Küfferle B, Tauscher J, Asenbaum S, Vesely C, Podreka I, Brucke T, Kasper S(1997) IBZM SPECT imaging of striatal dopamine-2 receptors in psy-chotic patients treated with the novel antipsychotic substance quetiapinein comparison to clozapine and haloperidol. Psychopharmacology 133:323–328

Laruelle M (2000) Imaging synaptic neurotransmission with in vivo bind-ing competition techniques: A critical review. J Cereb Blood FlowMetab 20: 423�451

Lavalaye J, Linszen D H, Booij J, Reneman L, Gersons B P, van Royen E A(1999) Dopamine D2 receptor occupancy by olanzapine or risperidonein young patients with schizophrenia. Psychiatry Res 92: 33–44

Mauri M C, Laini V, Boscati L, Rudelli R, Salvi V, Orlandi R, Papa P (2001)Long-term treatment of chronic schizophrenia with risperidone: a studywith plasma levels. Eur Psychiatry 16: 57–63

Meisenzahl E M, Dresel S, Frodl T, Schmitt G J, Preuss U W, Rossmuller B,Tatsch K, Mager T, Hahn K, Moller H J (2000) D2 receptor occupancyunder recommended and high doses of olanzapine: an iodine-123-iodobenzamide SPECT study. J Psychopharmacol 14: 364–370

Miller D D, Fleming F, Holman T L, Penny P J (1994) Plasma clozapineconcentrations as a predictor of clinical response: a follow-up study. J Clin Psychiatry 55(Suppl B): 117–121

Nordström A L, Farde L, Wiesel F A, Forslund K, Pauli S, Halldin C,Uppfeldt G (1993) Central D2-dopamine receptor occupancy in relationto antipsychotic drug effects—a double-blind PET study of schizo-phrenic patients. Biol Psychiatry 33: 227–235

Nordström A L, Farde L, Eriksson L, Halldin C (1995) No elevated D-2dopamine receptors in neuroleptic-naive schizophrenic patientsrevealed by positron emission tomography and [C-11]n-methylspiper-one. Psychiatry Res 61: 67–83

Nyberg S, Nordström A-L, Halldin C, Farde L (1995) Positron emission tomog-raphy studies on D2 dopamine receptor occupancy and plasma antipsy-chotic drug levels in man. Int Clin Psychopharmacol 10(Suppl 3): 81–85

Nyberg S, Eriksson B, Oxenstierna G, Halldin C, Farde L (1999) Suggestedminimal effective dose of risperidone based on PET-measured D2 and5-HT2A receptor occupancy in schizophrenic patients. Am J Psychiatry156: 869–875

Nyberg S, Chou Y H, Halldin C (2002) Saturation of striatal D(2) dopaminereceptors by clozapine. Int J Neuropsychopharmacol 5: 11–16

Olesen O V, Linnet K (1999) Olanzapine serum concentrations in psychi-atric patients given standard doses: the influence of comedication. TherDrug Monit 21: 87–90

Olesen O V, Licht R W, Thomsen E, Bruun T, Viftrup J E, Linnet K (1998)Serum concentrations and side effects in psychiatric patients duringrisperidone therapy. Ther Drug Monit 20: 380–384

Perry P J, Miller D D, Arndt S V, Cadoret R J (1991) Clozapine and nor-clozapine plasma concentrations and clinical response of treatmentrefractory schizophrenic patients. Am J Psychiatry 148: 23l–235

Perry P J, Lund B C, Sanger T, Beasley C (2001) Olanzapine plasma con-centrations and clinical response: acute phase results of the NorthAmerican Olanzapine Trial. J Clin Psychopharmacol 21: 14–20

Pickar D, Su T P, Weinberger D R, Coppola R, Malhotra A K, Knable M B,Lee K S, Gorey J, Bartko J J, Breier A, Hsiao J (1996) Individual varia-tion in D2 dopamine receptor occupancy in clozapine-treated patients.Am J Psychiatry 153: 1571–1578

Pilowsky L S, Costa D C, Ell P J, Murray R M, Verhoeff N P, Kerwin R W(1992) Clozapine, single photon emission tomography, and the D2dopamine receptor blockade hypothesis of schizophrenia. Lancet 340:199–202

Pilowsky L S, Busatto G F, Taylor M, Costa D C, Sharma T, SigmundssonT, Ell P J, Nohria V, Kerwin R W (1996) Dopamine D2 receptor occu-pancy in vivo by the novel atypical antipsychotic olanzapine – a 123I-IBZM single photon emission tomography (SPET) study.Psychopharmacology (Berl) 124: 148–153

Raggi M A, Mandrioli R, Sabbioni C, Pucci V (2004) Atypical antipsy-chotics: pharmacokinetics, therapeutic drug monitoring and pharmaco-logical interactions. Current Medicinal Chemistry 11: 279–296

Remington G, Kapur S, Zipursky R (1998) The relationship betweenrisperidone plasma levels and dopamine D2 occupancy: a positronemission tomographic study. J Clin Psychopharmacol 18: 82–83

Rinne J O, Hietala J, Ruotsalainen U, Sako E, Laihinen A, Någren K,Lehikoinen P, Oikonen V, Syvälahti E (1993) Decrease in human striataldopamine D2 receptor density with age: a PET study with [11C]raclo-pride. J Cereb Blood Flow Metab 13: 310–314

Scherer J, Tatsch K, Schwarz J, Oertel W, Kirsch M C, Albus M (1994)Striatal D2-dopamine receptor occupancy during treatment with typicaland atypical neuroleptics. Biol Psychiatry 36: 627–629

Schmitt G J, Meisenzahl E M, Dresel S, Tatsch K, Rossmuller B, Frodl T,Preuss U W, Hahn K, Moller H J (2002) Striatal dopamine D2 receptorbinding of risperidone in schizophrenic patients as assessed by 123I-iodobenzamide SPECT: a comparative study with olanzapine. J Psychopharmacol 16: 200–206

Schotte A, Janssen P F M, Gommeren W, Luyten W H M L, VanGompel P,Lesage A S, DeLoore K, Leysen J E (1996) Risperidone compared withnew and reference antipsychotic drugs: In vitro and in vivo receptorbinding. Psychopharmacology (Berl) 124: 57–73

Seeman P, Lee T, Chau-Wong M, Wong K (1976) Antipsychotic drug dosesand neuroleptic/dopamine receptors. Nature 261: 717–719

Seeman P (2005) An update of fast-off dopamine D2 atypical antipsy-chotics. Am J Psychiatry 162: 1984–1985

Sidhu J, Job S, Bullman J, Francis E, Abbott R, Ascher J, Theis J G(2006) Pharmacokinetics and tolerability of lamotrigine and olanza-pine coadministered to healthy subjects. Br J Clin Pharmacol 61:420–426

Talvik M, Nordstrom A L, Nyberg S, Olsson H, Halldin C, Farde L (2001)No support for regional selectivity in clozapine-treated patients: a PETstudy with [(11)C]raclopride and [(11)C]FLB 457. Am J Psychiatry158: 926–930

Talvik M, Nordstrom A L, Larsen N E, Jucaite A, Cervenka S, Halldin C,Farde L (2004) A cross-validation study on the relationship betweencentral D2 receptor occupancy and serum perphenazine concentration.Psychopharmacology 175: 148–153

Tauscher J, Küfferle B, Barnas C, Asenbaum S, Brücke T, Kasper S(1997) Dopamine-2 receptor imaging in psychotic patients treatedwith quetiapine, clozapine, risperidone and haloperidol. BiolPsychiatry 42: 15S

Tauscher J, Kufferle B, Asenbaum S, Fischer P, Pezawas L, Barnas C,Tauscher-Wisniewski S, Brucke T, Kasper S (1999) In vivo 123I-IBZM SPECT imaging of striatal dopamine-2 receptor occupancy inschizophrenic patients treated with olanzapine in comparison toclozapine and haloperidol. Psychopharmacology (Berl) 141:175–181

Tauscher J, Küfferle B, Asenbaum S, Tauscher-Wisniewski S, Kasper S(2002) Striatal dopamine-2 receptor occupancy as measured with[123I]iodobenzamide and SPECT predicted the occurrence of EPS inpatients treated with atypical antipsychotics and haloperidol.Psychopharmacology (Berl) 162: 42–49

Tauscher J, Jones C, Remington G, Zipursky R B, Kapur S (2002a)Significant dissociation of brain and plasma kinetics with antipsy-chotics. Mol Psychiatry 7: 317–321

at Biblioteca de la Universitat Pompeu Fabra on December 12, 2008 http://jop.sagepub.comDownloaded from

894 PK/PD model for atypical antipsychotics using SPECT

Tauscher-Wisniewski S, Kapur S, Tauscher J, Jones C, Daskalakis Z J,Papatheodorou G et al. (2002) Quetiapine: an effective antipsychotic infirst-episode schizophrenia despite only transiently high dopamine-2receptor blockade. J Clin Psychiatry 63: 992–997

Tort A B L, Souza D O, Lara D R (2006) Theoretical insights into the mech-anism of action of atypical antipscyhotics. Prog NeuropsychopharmacolBiol Psychiatry 30: 541–548

Appendix

Barcelona Clinical Imaging in Psychiatry Group Members

Ana M. Catafau, Experimental Medical Sciences, ClinicalPharmacology Discovery Medicine, Psychiatry Centre of

Excellence for Drug Discovery, GlaxoSmithKline, Barcelona,Spain and Verona, Italy and Centre for Imaging in Psychiatry,CRC-Mar, Hospital del Mar, Barcelona, Spain; María M. Penengo,Santiago Bullich, Centre for Imaging in Psychiatry, CRC-Mar,Hospital del Mar, Barcelona, Spain; Iluminada Corripio, VíctorPérez, Enric Álvarez, Psychiatry Department, Red deEnfermedades Mentales-Trastornos Afectivos y Psicoticos (REM-TAP Network) Hospital de Sant Pau, Barcelona, Spain; EduardParellada, Mireia Font, Psychiatry Department, ClinicSchizophrenia Program, Hospital Clinic i Provincial, Barcelona,Spain; Carles Garcia-Ribera, Purificación Salgado, AlfonsoRodríguez, Salvador Ros, Mental Health and Drug AddictionDepartment, Hospital del Mar and Psychiatry Institute, IMAS,Barcelona, Spain; Javier Perich, Magnetic Resonance Department,CRC-MAR, Hospital del Mar, Barcelona, Spain.

at Biblioteca de la Universitat Pompeu Fabra on December 12, 2008 http://jop.sagepub.comDownloaded from