University of Groningen. Translational PKPD modeling in schizophrenia Pilla Reddy, Venkatesh

Transcription

1 University of Groningen Translational PKPD modeling in schizophrenia Pilla Reddy, Venkatesh IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Publication date: 2012 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Pilla Reddy, V. (2012). Translational PKPD modeling in schizophrenia: linking receptor occupancy of antipsychotics to efficacy and safety Groningen: s.n. Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date:

2 Pharmacokinetic-Pharmacodynamic Modelling of Severity Levels of Extrapyramidal Side Effects with Markov Elements Venkatesh Pilla Reddy, 1 * Klas J. Petersson, 2 * Ahmed Abbas Suleiman, 2 An Vermeulen, 3 Johannes H. Proost, 1 and Lena E. Friberg 2. Submitted 1 Division of Pharmacokinetics, Toxicology and Targeting, University of Groningen, Groningen, the Netherlands 2 Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden 3 Advanced PKPD Modelling and Simulation, Janssen Research & Development, a Division of Janssen Pharmaceutica NV, Beerse, Belgium * The first and second authors contributed equally to the paper CHAPTER

3 202 ABSTRACT A major problem in the treatment of schizophrenic patients with current antipsychotic drugs, mainly acting as dopamine-2 receptor antagonists, is the occurrence of side effects such as Extra-Pyramidal-Symptoms (EPS). Metaanalyses of summary data of EPS occurrence, and receptor occupancies inferred from mean plasma concentrations, have shown the incidence of EPS to rise when receptor occupancy is above approximately 80%. In this analysis, individual longitudinal EPS data from 2630 patients participating in one of seven different trials and treated with haloperidol, paliperidone, ziprasidone, olanzapine, JNJ , or placebo, was analyzed using a compartment-based probability model with Markov elements. The developed pharmacokinetic-pharmacodynamic model describes the longitudinal changes of spontaneously reported EPS-related adverse events and their severity levels rated by clinicians. Individual steadystate concentrations and occupancy levels were found to be predictors for EPS. The results confirm 80 % occupancy as a level of increased EPS occurrence rates, also at the individual level.

4 PKPD Modelling of EPS Side Effects 203 INTRODUCTION Schizophrenia is a chronic psychiatric condition that includes positive symptoms (e.g. delusions, hallucinations), negative symptoms (e.g. lack of motivation, social withdrawal), and cognitive symptoms (e.g. lack of attention, loss of memory). These symptoms can severely affect the social, educational, occupational and health related functioning of the individual. [1] It has been hypothesized that schizophrenia is associated with excessive dopamine activity in the mesolimbic dopaminergic neurons (causing positive psychotic symptoms) and reduced dopamine activity in the mesocortical dopaminergic neurons (causing negative and cognitive symptoms). [2] Antagonism at the dopamine D 2 receptor is the common mechanism of action for currently available treatment of schizophrenia. [3] However, D 2 receptor occupancy (D 2 RO) higher than 80% in other regions than the mesolimbic and mesocortical dopaminergic pathways, e.g., nigrostriatal pathway in the brain increases the risk of adverse effects (AEs) such as extrapyramidal symptoms (EPS). [4] Extrapyramidal side effects of antipsychotic drugs (APs) include acute movement disorders such as Parkinsonism, akathisia, dystonia and chronic movement disorders such as tardive dyskinesia and tardive dystonia. In addition to being uncomfortable to patients, these side effects could behave as confounding factors and may lead to an underestimation of efficacy of the APs. [5] In drug development EPS symptoms are often accessed using rating scales [e.g. Simpson Angus Scale (SAS) and Extrapyramidal Symptom Rating Scale (ESRS)] or by occurrences spontaneously reported by patients, the severity of which are graded by clinicians at the time of planned visits. Atypical antipsychotics (ATAPs) were introduced to the clinic in an attempt to lower the EPS incidence compared to the typical antipsychotics (e.g. haloperidol) and to improve the negative and cognitive symptoms of schizophrenia. One of the claims while marketing ATAPs is the absence of EPS side effects. [6,7] Recently, a meta-analysis of ATAPs vs. haloperidol using a central-tendency statistical approach to evaluate such a claim was performed and it demonstrated that at least some ATAPs do produce EPS and that there are differences between drugs in their tendency to produce EPS. [7] Previously described assessments of EPS either looked at mean differences in either Simpson-Angus Scale or EPS-incidence on a study level [8] or indirect measurements such as the prescription of anti-eps treatment. Until now, no papers have been published linking individual data of EPS to predictors such as plasma concentrations or receptor occupancy using pharmacokinetic-pharmacodynamic (PKPD) models developed describing the longitudinal aspect of EPS events following antipsychotic treatment. However, de Ridder and Vermeulen [] modeled the time to the first treatment-emergent EPSrelated event using hazard models, without considering the severity of the EPS events and presented the results in a book chapter. []

5 204 The discrete nature of observations of events or discrete type scales as the ones used for EPS assessments calls for likelihood type models. In the field of PKPD models, one of the approaches that could be used is the proportional odds model. However, this model does not assume dependency between observations and the profiles generated from simulations can be physiologically implausible with individuals moving more frequently between the EPS severity levels than observed. Markov elements would be needed to handle that property of the data. Another approach would be to assess the time-to-eps of a certain severity grade using survival models but these would not take into account that patients stay in one state for several observations nor that the probability of states are correlated. A compartmental Markov type of model is capable of estimating parameters considering time factor and hence eliminating the need of having frequent observations at equal spaces of time needed in ordinary Markov models or ordered categorical models. [10] In addition, we can scale time (e.g. compute probability for one day and scale to X days). Markov type of model allows predicting the probability of observing a patient at any of the EPS severity grades depending on what severity grade he or she had at the last observation and the time elapsed since the last observation. Thus, Markov modelling approach can help to quantify and compare the potency of different APs to induce EPS AEs over a time. In this analysis, the aim was to develop a compartmental, population-based mixed-effects Markov model to describe the probability of EPS incidence and EPS severity as a function of dose, individual steady-state drug exposure (C ss ) or predicted dopamine D 2 RO of APs. All available ordered categorical EPS data after the administration of placebo, four ATAPs (paliperidone extended release, ziprasidone, and JNJ , olanzapine) and haloperidol from seven clinical studies were used. METHODS Studies, Patients and EPS scores EPS modelling was performed using individual-level data obtained from six randomized, placebo-controlled, clinical trials, and one open-label study conducted in schizophrenic patients. The study durations ranged from 4 to 12 weeks, following oral administration of one of five APs, namely, haloperidol, olanzapine, paliperidone extended release, ziprasidone, and JNJ EPS-related AEs included side effects that were reported spontaneously by patients over the study period. The important AEs documented were akathisia, hyperkinesia, dystonia, Parkinsonism, dyskinesia, tardive dyskinesia, hypokinesia, and tremor. Since the total number of events was too low for some AEs to draw conclusions about the exposure-response relationship, and since not all trials reported on all types of EPS-related AEs, the model was developed without discriminating between the different types of EPS-related AE.

6 PKPD Modelling of EPS Side Effects 205 According to the design of the studies, the clinicians consulted patients at least once a week during the total study duration. Some studies included two visits during the first week of the trial. Patients were directed to record the incidence of EPS, the duration of the EPS episode and resolution of any EPS, the severity of which was graded by clinicians on a scale from 0 to 3 (0: no EPS, 1: mild EPS, 2: moderate EPS and 3: severe EPS). Transitions were thus possible on any day and not only on visit days. If the patient dropped out before the end of the study, all available EPS scores until the day of dropout were included in the analysis. If several manifestations of EPS were experienced simultaneously, but differing in grade, the grade was set to the highest grade. In all trials, patients were hospitalized for the first 2-3 weeks of the study period. Data were analyzed by non-linear mixed-effects modelling using the LAPLACE method in NONMEM 7, level 1.2, [11] with front-end interface PsN [12] R version 2.11 and higher versions were used for data management purposes and generating graphic outputs. [13] Pharmacokinetics A patient-specific steady-state concentration (C ss ) was calculated using post-hoc Bayesian estimates of clearance obtained from previously developed PK models. [14-17] The typical apparent clearance (CL/F) estimates were 88 l/h for haloperidol, 14 l/h for paliperidone, 20 l/h for olanzapine, 54 l/h for ziprasidone, and 25 l/h for JNJ The calculated individual C ss was then linked to the timecourse of EPS incidence. C ss was set at 0 for subjects receiving placebo. EPS model structure A first-order Markov chain type model was used. Figure 1 shows the model structure and corresponding differential equations used to model the EPS incidence rates. EPS events were categorized on an ordered categorical scale based on the intensity of EPS-related AEs. Because there were few observations on severe EPS, the moderate and severe EPS states were combined and consequently a three-state scale was used. A Markov transition model with three compartments (Figure 1) was used and the rate constants of the change in probability over time of an observation being in the different states were estimated. Each compartment represented the probabilities of a different state (no EPS, mild EPS, and moderate/severe EPS). The probability at t=0 was set to 1 for the observed EPS state. The probabilities were allowed to distribute between the compartments and the first-order rate constants between the compartments were estimated. After an observation, all compartments were reset to zero and the compartment corresponding to the observed EPS state of the patient was initialized with a 100 % probability. Effects of time, drug exposure, receptor occupancy, and covariates on transition probabilities were evaluated in the Markov model as a proportional relation on the transition rate parameters. Drug and covariate effects were tested on each transition parameter separately, and in combination.

7 206 Table 1. Overview of clinical trials in subjects with schizophrenia included in the development of the EPS model Route of Administration Drug/Dose #Subjects % EPS rate* Study Design Paliperidone ER 6,,12 mg SCH-303 Phase-III, double-blind, parallel, placebo controlled, randomized 6 weeks Oral/once daily Olanzapine 10 mg Placebo Paliperidone ER 6,12 mg 1 16 SCH-304 Phase-III, double-blind, parallel, placebo controlled, randomized 6 weeks Oral/once daily Olanzapine 10 mg Placebo 7 6 Paliperidone ER 3,,15 mg SCH-305 Phase-III, double-blind, parallel, placebo controlled, randomized 6 weeks Oral/once daily Olanzapine 10 mg, Placebo Phase-III, double-blind, parallel, Ziprasidone 40, 80 mg weeks Oral/twice daily placebo controlled, randomized Placebo 2 20 Ziprasidone 20,60,100 mg Phase-III, double-blind, parallel, placebo controlled, randomized 6 weeks Oral/twice daily Haloperidol 10 mg 85 5 Placebo JNJ SCH-2002 Phase-II, double-blind, parallel, placebo controlled, randomized 6-12 Weeks Oral/twice daily Olanzapine 15 mg (QD) 3 15 Placebo LMU Open-label study 4 weeks Oral/once daily Haloperidol mg 58 5 Study Duration Demographics included in this analysis: Gender: males (55 %) and females (45 %); Race: Caucasians (67 %), Blacks (18 %), Asians (7 %) and others (6 %); Geographical region: United States of America (~ 41 %), Eastern Europe (~ 35 %), Asia (~10 %) and other regions (~13 %) and concomitant medications for treating EPS: anti-cholinergic drugs (12 %), beta-blockers (1.8 %) and benzodiazepines (1 %). *EPS rate: [number of subjects with treatment-emergent EPS-related AEs (irrespective of severity)/ total number of subjects randomized in a trial or treatment arm]

8 PKPD Modelling of EPS Side Effects 207 Fig. 1. EPS compartmental model structure with possible transitions (no EPS, mild EPS and pooled moderate/severe, defined as compartment 1, 2, and 3, respectively). The forward rate transitions constants KF 12 and KF 23 were allowed to be influenced by drug exposure or D 2 RO. Placebo effect and drug effect Various functions such as constant, linear, exponential and Weibull functions were explored for describing a change over time in the transition probability rate constants. The contribution of drug effect to the transition probabilities were modeled after the placebo model was built. The antipsychotic drug exposure was related to the EPS incidence with increasing complexity using linear, E max and sigmoid E max models and tested on top of parameters estimated for the placebo group. To characterize the relationship between EPS incidence and dopamine D 2 RO levels, the following relationship was used to calculate D 2 RO for each patient: Where, RO max is the maximum receptor occupancy, Kd is the plasma level of antipsychotic drug associated with 50 % of RO max. The values of Kd for D 2 receptor binding were obtained from literature, where E max models were used to fit D 2 RO and plasma drug concentrations data obtained from PET studies. The Kd values for D 2 receptor binding (assuming RO max 100%) were 0.56 ng/ml for haloperidol, 12 ng/ml for olanzapine, 4. ng/ml for paliperidone, 33 ng/ml for ziprasidone, [18] and 15 ng/ml for JNJ [1] Covariate model Influences of patient and study specific covariates were evaluated as possible explanatory variables for the individual variability in the model parameters (Table 1). The potential effect of concomitant medications for treating EPS was tested using the data from paliperidone trial where exact time of start and stop of these medications were known. Covariates were included in the model using different functional forms like linear, piece-wise linear, power and exponential functions. [12] Only non-missing, clinically relevant, and uncorrelated covariates that met the pre-defined statistical criteria (p <0.05) were included in the EPS model. Final model selection and model evaluation During initial model building, differences in the objective function value (OFV; difference denoted as ΔOFV: 3.84 as a cut-off value) between the models together with % RSE (from NONMEM covariance step) of the parameters were used to

9 208 guide the selection of the model. A bootstrap analysis (1000 samples) was carried out, for both the drug exposure-eps and the receptor occupancy-eps models, to check the precision of the estimates and to construct a 5 % confidence interval for each of the parameter estimates. The adequacy of the model was investigated with simulation-based visual predictive checks (VPC). Briefly, lower and upper prediction intervals of the proportions of patients at each EPS state at each planned observation as well as transitions between the different EPS states were constructed using 100 simulated replicates of the initial dataset and compared with the corresponding observed proportions or transitions. RESULTS EPS data The EPS dataset consisted of 16,58 observations (rated as no EPS, mild, moderate or severe) collected from 2,630 subjects with acute schizophrenic condition. Of these subjects, 54 subjects received placebo; 867 subjects paliperidone (3-15 mg/day); 437 subjects olanzapine (10 or 15 mg/day); 143 subjects haloperidol ( mg/day); 285 subjects ziprasidone ( mg/day) and 301 received JNJ [20] treatment (10-30 mg/day). Table 1 summarizes the trial designs and the observed EPS incidence rates of APs that were used in this study. The calculated individual-level C ss and the predicted D 2 RO range (minmax) were as follows; for haloperidol (1-4 ng/ml; 55-8 %); olanzapine (-76 ng/ ml; %); paliperidone ( ng/ml; -5 %); ziprasidone ( ng/ml; 38-8 %); JNJ (8-178 ng/ml; 35-2%). Structural model for EPS incidence To be able to calculate the probabilities of observing each EPS severity grade at any time point, the probabilities were determined by a Markov model with a compartmental structure using differential equations (Figure 1). To model the EPS occurrence, a placebo model was first developed based on the placebo data alone. Thereafter the combined placebo and drug effects model was built by estimating placebo and drug effect parameters. Following the start of the study, the decrease of the transition rates as a function of time over the study period was best described by the exponential model (Eqs. 1 and 2). A linear model described the relationship between the drug exposure (Eq 3) and the effect of APs on the transition rate constants reasonably well. Modelling the drug effect, using dose or Css, as proportional to the placebo effect resulted in a better model fit and lower relative standard errors of the parameter estimates, compared to additive drug effects. E max model was the best to link D 2 RO and EPS events (Eq 4).

10 PKPD Modelling of EPS Side Effects 20 (Eq. 1) (Eq. 2) (Eq. 3) (Eq. 4) Differential equations used to model the probability of observing EPS event da(1)/dt=kb 21 A(2) - KF 12 A(1) da(2)/dt=kf 12 A(1) + KB 32 A(3)- KB 21 A(2) - KF 23 A(2) da(3)/dt =KF 23 A(2) - KB 32 A(3) KF 12 =Base 12 exp (-k 12 time) (Drug Effect) KB 21 = Base 21 exp (-k 21 *time) KF 23 =Base 23 exp (-k 23 *time) (Drug Effect) KB 32 = Base 32 exp (-k 32 *time) KF xy is the forward transition probability rate constant at a given time and treatment; KB xy is the backward transition probability rate constant at time t. Base xy the transition probability rate constant at the start of the study; x: present EPS state; y: future EPS state; t (1/2) = ln2 k xy ; t 1/2 : half-life describing the change in KF xy or KB xy with time; Drug effect is the effect of antipsychotic drug on KF xy ; EFF is the drug dependent slope parameter in relation to drug exposure; E max represents the maximum drug effect on the transition rate from state x to state y; RO 50 is the model parameter that corresponds to compound independent D 2 RO required for 50% of E max effect; A(1), A(2), A(3) represents the compartments for EPS probabilities. Inclusion of inter-individual variability (IIV) as an exponential function (log-normal distribution) to describe differences between patients sensitivity to treatment effects resulted in an improved model fit by objective function, however the model was numerically unstable and parameters were estimated with less precision and simulation properties worsened, probably due to nonnormal distributions of random effects that frequently arise for categorical type models. Allowing for semi-parametric distributions [21] did not improve simulation properties. Therefore, the model was built without IIV in the typical parameters. Visit 4, scheduled to be between day 14 and day 21 (except LMU

11 210 study), corresponds to the point in the study at which subjects were allowed to enter into an outpatient status. The improvement in the model fit and in visual predictive checks showed that it was important to allow some parameters to have different values before and after visit 4 (approximately day 16). The effect of predictors was only significant to estimate on the forward transition rate constants [i.e. transitions from no EPS to mild (KF 12 ) and from mild to moderate/severe EPS (KF 23 )]. Based on the NONMEM OFV value, C ss was better than dose (decrease of ~45 OFV units) in describing the drug-eps relationships. The final EPS model structure (based on all drugs simultaneously) was able to describe the observed EPS occurrence adequately with acceptable low relative standard errors. To further characterize the relationship between D 2 RO by different APs and the emergence of EPS, predicted D 2 RO levels associated with C ss were linked to the transition rate constants by either using a linear function (Eq 3) or sigmoidal function (Eq 4) with Q transformation of the D 2 RO: Q=D 2 RO/(100-D 2 RO). The advantage of this transformation is that it allows Q to have a value from zero to infinity for D 2 RO from 0 to 100% and allowing the Drug Effect to reach its maximum effect (E max ) for D 2 RO approaching 100% (Eq. 4). This parameterization resulted in statistically better and numerically more stable estimation than directly linking the absolute D 2 RO levels to the transition rate constants. The parameter estimates of the EPS model when using C ss and D 2 RO as predictors are shown in Table 2. The dopamine D 2 RO needed for half of the maximum effect of APs on the transition rate was found to be 86.5 %. The posthoc forward transition rate constants for each individual vs. predicted individual D 2 RO (based on C ss and K d ) for different antipsychotic drugs are depicted in Figure 2. It is evident that haloperidol exhibited higher EPS occurrence rates when compared to ATAPs despite the same level of D 2 RO. Olanzapine was found to be less associated with EPS events than placebo treatment, causing difficulties in estimating the effect of olanzapine on EPS transitions. The EPS incidence was higher (15 %) for the olanzapine arm compared to the placebo treatment in one of the studies (SCH-2002), but individual plasma olanzapine profiles were not available from this study, which may have contributed to difficulties in estimating the effect of olanzapine exposure effect on the probability of EPS. Covariate model Covariates were included in the model based on their statistical significance and clinical effect. The final EPS model included only one covariate (regional difference: Eastern Europe vs. other) which was added as a proportional linear term to the corresponding transition probability. Residing in Eastern Europe increased the chances of improvement of the EPS side effects since it increased backward movement probability rate constants (KB 21 and KB 32 ) by ~100 % (Table 2). Not all covariate information was available for all trials to test the influence

12 PKPD Modelling of EPS Side Effects 211 of covariates on the probability of EPS. Only paliperidone trials had important covariate information such as co-medication and complete demographic details. Based on only the paliperidone data, administration of anti-cholinergic drug for the treatment of EPS was found to be an influential covariate, decreasing the transition probability rate constant for movement from a mild EPS state to moderate/severe EPS states by ~80 %. Model evaluation The final models (Table 2) were evaluated using a bootstrap analysis with 1000 samples. 5% confidence intervals were constructed for each of the parameters and none of these overlapped zeros. The population estimates of the final model are in close agreement with the median values of the 52 successful bootstrap replicates. Visual predictive checks (VPCs) using 100 simulated data sets were Fig. 2. Increase in forward rate constants for all individuals by the final model vs. predictions of drug occupancy based on individual Css. All doses for ATAPs and only lower doses of haloperidol (<10 mg/day) were used.

13 212 Table 2: Parameter estimates (5% CI from bootstrap analysis) of the final EPS models using C ss and D 2 -receptor occupancy as predictors of EPS. D 2 Receptor Occupancy (E max Model) Steady-State Exposure (Linear Model) Predictor Parameters <=Day 16 > Day 16 <=Day 16 > Day 16 Base 12 (day-1) ( ) ( ) Base 21 (day-1) ( ) # ( ) # ( ) # ( ) # Base 23 (day-1) ( ) ( ) # 0.36 ( ) ( ) # Base 32 (day-1) ( ) # ( ) # t ½ for KF 12 (days) 8.8 ( ) # 8. ( ) # t ½ for KF 23 (days) 2. ( ) 8.8 ( ) # 2.7 ( )# 8. ( ) # t ½ for KB 21 and KB 32 (days) 22 ( ) 20 ( ) EFF: Haloperidol 2.4 ( ) - EFF: Paliperidone ( ) - EFF: Ziprasidone ( ) - EFF: JNJ ( ) - Effect of residence in Eastern Europe on KB 21 and KB ( ) 1.08 ( ) Emax * ( ) RO 50 (%) * ( ) γ (shape parameter) - 1.3* ( )Φ OFV 7740 ( ) 763 ( ) #: shared parameter estimates; : the covariate effect was expressed as a proportional increase in the parameter value by the covariate. *: a single drug effect parameter was estimated. Φ: γ does not appear to be significantly different from 1 as 1 is included in confidence interval: 1.3* ( )

14 PKPD Modelling of EPS Side Effects 213 made to evaluate the predictive ability of the model. The model was capable of satisfactorily predicting the proportions of the patients experiencing each of the EPS states across time (Figure 3). The number of transitions simulated from the model was found to be reasonably consistent with the observed number of transitions as shown for the paliperidone data (supplementary file figure 1). DISCUSSION This work describes a PKPD modelling approach that successfully characterized the temporal course of the proportions of patients experiencing different grades of EPS following exposure to one of 5 different antipsychotic drugs or placebo. The dependency of a future EPS event on the current state was accounted for by using a compartmental structure, defined by differential equations, for the probabilities of observing each severity grade at any time point (Fig. 1). [10,22] The probabilities were thereby continuous with respect to time, and coupled to the observed scores through a probabilistic model. Markov elements can also be incorporated in a proportional odds model, [23] however, that type of model is discrete in time and the sampling times and frequencies can greatly influence the parameter estimates. [24] The analyzed studies included self-reported EPS events, and consequently the EPS observations were not equally spaced in time. Therefore, the compartmental Markov modelling approach was chosen and was shown to handle the dependency between successive observations of ordered adverse events well. The simultaneous fit of the model to all data resulted in one set of parameter estimates that to an adequate degree could simulate the proportions of patients experiencing EPS after treatment with one of four different compounds studied in seven different trials (Figure 3). The simulations of the transitions between different states show that the model cannot recreate exactly every transition seen in the data, but in relation to the huge differences between the occurrence of the different transition types, from 0.1 % to above 0%, we conclude that the overall pattern of transitions can be described by the model (supplementary file; Figure 1). A linear model with respect to steady-state concentrations was better than using the predicted D 2 -receptor occupancy as a direct link to EPS. Looking at figure 2, this finding concurs with what de Greef et al. [5] found on the meta-data level, where the incidence rate increases at occupancy >80%. A 100% increase in forward rate constants corresponds to a 50% decrease of the half-time to having an event, compared to placebo. The model dependent on the D 2 -receptor occupancy can however be used to predict EPS for a new D 2 -receptor antagonist using the drug K d and information on the typical PK profile. The model can also simulate the longitudinal correlations seen in EPS AE data. It has been suggested that ATAPs induce EPS [5] at a lower rate than typical antipsychotics. A model-based approach can help to investigate and quantify

15 Fig. 3. Visual predictive checks showing the proportions of the patients at each EPS state on planned visits in the observed data (red lines) with the corresponding 5% prediction intervals as constructed from 100 simulations (light blue region) for different antipsychotics using Css as a predictor. 214

16 PKPD Modelling of EPS Side Effects 215 these hypotheses and to compare the potency of different APs to induce EPS by using the same model structure. The EPS occurrence rate was approximately 16 times higher for haloperidol than for ATAPs. Haloperidol exhibited higher EPS occurrence rate even at the lower doses of 10 mg/day (Figure 2). Regional differences in the EPS incidence might be an important factor explaining the differences between studies and need to be taken into consideration. Study region was found to be an important covariate where patients studied in Eastern Europe had a lower incidence and shorter average event lengths of EPS than other countries. The reason for this is not clear but regional differences in study results are not uncommon, different attitudes towards reporting adverse events could be one reason. The present EPS model did not consider the dropouts, as in dropout modelling; a standard assumption is that within a subject the risk of dropout is independent of the risk for the EPS event or whatever the effect modeled. If this is not the case, the parameter estimates of the effect model could be slightly biased. [25] However, EPS was not the major reason for discontinuation of treatment, but primarily discontinuation was due to lack of efficacy and subject s choice. It is an important task to ascertain a dose and dosing regimen that would provide adequate exposure maintaining target D 2 RO of % with a low risk of AEs such as EPS. Overall, this data analysis indicated that olanzapine was well tolerated in patients with schizophrenia as it was not possible to estimate a slope parameter for the effect (EFF in Eq 3). However, the proportions of patients with EPS at a given time point, and therefore the absolute numbers of events were relatively low for both placebo and olanzapine in this study and hence the parameter estimates were uncertain. It may be that olanzapine has a favorable EPS profile [26] due to binding to receptors such as 5-HT 2A, but that olanzapine would truly have lower incidence rates than placebo seems unlikely. In conclusion, a model was successfully developed that can characterize the temporal course of the proportions of patients experiencing different grades of EPS using a compartmental model structure incorporating Markov elements. To our knowledge, this approach has not been used for characterizing side-effect data in general and no population models have been developed for EPS side effects. The approach could be used for any adverse event with severity grading and a reported duration. The model quantifies the relationship between the occurrence of EPS and movement to higher EPS states and the different predictors of exposure (dose, C ss ), as well as receptor occupancy of the different APs. ACKNOWLEDGEMENTS Klas J. Petersson was supported by a grant from Janssen Research & Development (Beerse, Belgium). Venkatesh Pilla Reddy was supported financially by the Dutch Top Institute Pharma (Leiden, The Netherlands; within the framework of project no. D2-104 (Mechanism-based PK-PD modelling platform).

17 216 We thank Dr. Jing Liu (Pfizer Global Research and Development, Groton, USA) for her contribution to this article via Top Institute Pharma. We also thank Professor Dan Rujescu (Department of Psychiatry, Ludwig Maximilians University, Munich, Germany) for sharing the haloperidol data. We thank Professor Geny M.M. Groothuis, Professor Meindert Danhof, and Dr. Magdalena Kozielska for their scientific inputs. Supplementary file: Fig. 1. Visual predictive check of Markov model transitions for paliperidone treatment. The dashed black lines represent the 10 th and 0 th percentiles of the simulated number of transitions. Green and red solid line indicates the median observed and simulated transitions, respectively.

18 PKPD Modelling of EPS Side Effects 217 REFERENCES 1. Mueser KT, Mcgurk SR. Schizophrenia. Lancet 2004; 363(426): Howes OD, Montgomery AJ, Asselin MC, et al. Elevated Striatal Dopamine Function Linked to Prodromal Signs of Schizophrenia. Arch Gen Psychiatry 200; 66(1): Seeman P. All Roads to Schizophrenia Lead to Dopamine Supersensitivity and Elevated Dopamine D2High Receptors. CNS Neurosci Ther 2011; 17(2): Horacek J, Bubenikova-Valesova V, Kopecek M, et al. Mechanism of action of atypical antipsychotic drugs and the neurobiology of schizophrenia. Cns Drugs 2006; 20(5): de Greef R, Maloney A, Olsson-Gisleskog P, et al. Dopamine D(2) Occupancy as a Biomarker for Antipsychotics: Quantifying the Relationship with Efficacy and Extrapyramidal Symptoms. AAPS J 2011; 13(1): Sernyak M, Rosenheck R. Experience of VA psychiatrists with pharmaceutical detailing of antipsychotic medications. Psychiatric Services 2007; 58(10): Rummel-Kluge C, Komossa K, Schwarz S, et al. Second-generation antipsychotic drugs and extrapyramidal side effects: A systematic review and metaanalysis of head-to-head comparisons. Schizophr Bull Matsui-Sakata A, Ohtani H, Sawada Y. Pharmacokinetic-pharmacodynamic analysis of antipsychotics-induced extrapyramidal symptoms based on receptor occupancy theory incorporating endogenous dopamine release. Drug Metab Pharmacokinet 2005; 20(3): De Ridder F, Vermeulen AM. Balancing efficacy and safety in the clinical development of an atypical antipsychotic, paliperidone extended-release. In: Kimko HC, Peck CC, editors. Clinical Trial Simulations: Applications and Trends. AAPS Advances in the Pharmaceutical Sciences Series, Bergstrand M, Soderlind E, Weitschies W, et al. Mechanistic Modeling of a Magnetic Marker Monitoring Study Linking Gastrointestinal Tablet Transit, In Vivo Drug Release, and Pharmacokinetics. Clin Pharmacol Ther 200; 86(1): Beal S, Sheiner, LB, Boeckmann A, and Bauer RJ. NONMEM User s Guides. (18-200), Icon Development Solutions, Ellicott City, MD, USA, Lindbom L, Pihlgren P, Jonsson N. PsN-Toolkit - A collection of computer intensive statistical methods for nonlinear mixed effect modeling using NONMEM. Comput Methods Programs Biomed 2005; 7(3): R Development Core Team (2011). R: A language and environment for statistical computing. R Foundation for Statistical Computing, V., Austria. URL Samtani MN, Vermeulen A, Stuyckens K. Population pharmacokinetics of intramuscular paliperidone palmitate in patients with schizophrenia: a novel once-monthly, long-acting formulation of an atypical antipsychotic. Clin Pharmacokinet 200; 48(): Hoeben E, Neyens M, Mannaert M, et al. Population pharmacokinetic analysis of JNJ , a specific and fastdissociating D2 antagonist for the treatment of schizophrenia. [abstract no. 206]. 20th Population Approach Group in Europe; 2011 Jun 07-10; Athens, Greece [online]. Available from URL: default.asp?abstract=206. [Accessed 2011 Dec 26]. 16. Pilla Reddy V, Kozielska M, Johnson M, et al. Population pharmacokinetic and pharmacodynamic modeling of haloperidol in patients with schizophrenia using Positive and Negative Syndrome Rating Scale. Submitted to J Clin Psychopharmacol. 17. Pilla Reddy V, Kozielska M, Johnson M, et al. Pharmacokinetic and pharmacodynamic modeling of

19 218 antipsychotic drugs in patients with schizophrenia Part 1: the use of PANSS total score and clinical utility. Submitted to Schizophr Res. 18. Nucci G, Gomeni R, Poggesi I. Modelbased approaches to increase efficiency of drug development in schizophrenia: a can t miss opportunity. Expert Opin Drug Discov 200; 4(8): te Beek ET, de BP, Moerland M, et al. In vivo quantification of striatal dopamine D2 receptor occupancy by JNJ using [11C]raclopride and positron emission tomography. J Psychopharmacol te Beek ET, Moerland M, de BP, et al. Pharmacokinetics and central nervous system effects of the novel dopamine D2 receptor antagonist JNJ J Psychopharmacol Petersson KJF, Hanze E, Savic RM, et al. Semiparametric Distributions With Estimated Shape Parameters. Pharm Res 200; 26(): Lacroix BD, Sargentini-Maier ML, Karlsson MO, et al. Simultaneous modeling of the three ACR improvement thresholds - 20, 50 and 70% - in rheumatoid arthritis patients treated with certolizumab pegol. PAGE 1 (2010) Abstr 1811 [ 23. Bizzotto R, Zamuner S, De Nicolao G, et al. Multinomial logistic estimation of Markov-chain models for modeling sleep architecture in primary insomnia patients. J Pharmacokinet Pharmacodyn 2010; 37(2): Snelder N, Diack C, Benson N, et al. A proposal for implementation of the Markov property into a continuous time transition state model in NONMEM. [abstract no. 1536].18th Population Approach Group in Europe; 200 Jun 23-26; St. Petersburg, Russia [online]. Available from URL: PAGE 18 (200) Abstr 146 [ org/?abstract=1536]. [Accessed 2011 July 26]. 25. Bjornsson MA, Friberg LE, Simonsson USH. Performance of non-linear mixed effects models with and without taking informative dropout into account. [abstract no. 2055]. 20th Population Approach Group in Europe; 2011 Jun 07-10; Athens, Greece [online]. Available from URL: o rg/defa u l t. a s p? a b s t ra c t = [Accessed 2011 Dec 26]. 26. Carlson CD, Cavazzoni PA, Berg PH, et al. An integrated analysis of acute treatment-emergent extrapyramidal syndrome in patients with schizophrenia during olanzapine clinical trials: Comparisons with placebo, haloperidol, risperidone, or clozapine. J Clin Psychiatry 2003; 64(8):88-06.