Population Pharmacokinetics and Dosing Simulations for Aripiprazole 2‐Month Ready‐to‐Use Long‐Acting Injectable in Adult Patients With Schizophrenia or Bipolar I Disorder

A ready‐to‐use (RTU) long‐acting injectable (LAI) formulation of aripiprazole monohydrate for administration once every 2 months, available in 960 mg (Ari 2MRTU 960) or 720 mg doses, has been developed for the treatment of schizophrenia or bipolar I disorder. A previously developed and validated population pharmacokinetic model for characterizing aripiprazole plasma concentrations following administration of oral aripiprazole or aripiprazole once‐monthly (AOM) intramuscular injection was expanded to include the RTU LAI formulation of aripiprazole (Ari RTU LAI). Overall, 8899 aripiprazole pharmacokinetic samples from 1191 adults from 10 clinical trials were included in the final combined analysis data set. Aripiprazole plasma concentration‐time profiles were simulated for various Ari RTU LAI initiation and maintenance scenarios in 1000 virtual patients. Diagnostic plots demonstrated that the final population pharmacokinetic model, which incorporated data for oral aripiprazole, AOM, and Ari RTU LAI, adequately described aripiprazole concentrations following Ari RTU LAI administration. Absorption of Ari RTU LAI was modeled by a parallel zero‐order and lagged first‐order process. Simulations across multiple scenarios were performed to inform dosing recommendations, including various treatment initiation regimens for a 2‐monthly formulation of Ari RTU LAI in patients with or without prior stabilization on oral aripiprazole, and for patients switching from AOM. Additional simulations accounted for missed/delayed doses, cytochrome (CYP) 2D6 metabolizer status, and concomitant use of CYP2D6 or CYP3A4 inhibitors. Overall, simulations across a variety of scenarios demonstrated an Ari RTU LAI pharmacokinetic exposure profile that was comparable to AOM, with a longer dosing interval.

Schizophrenia and bipolar disorder are complex, chronic mental health disorders, with affected individuals typically experiencing multiple relapses of symptoms that are separated by periods of remission. 1,2hile daily oral antipsychotic agents have proven effective for many patients, [3][4][5] achieving and maintaining symptom control relies on their consistent and longterm use. 6,7][10][11] Various strategies for improving adherence to treatment are available, including the use of long-acting injectable (LAI) antipsychotics; 12 LAI formulations offer dosing intervals ranging from 2 weeks to 6 months, 13 reducing the adherence demands on patients, compared with daily dosing. 149][20][21] Aripiprazole is available in a once-daily oral formulation and as an extended-release suspension of aripiprazole monohydrate for monthly intramuscular injection, known as aripiprazole oncemonthly (AOM).3][24] Per the product label, the formulation contains aripiprazole monohydrate supplied in a ready-to-use (RTU) prefilled syringe containing doses of 960 or 720 mg, with the latter dose used in the case of any tolerability concerns, in patients who are known cytochrome P450 (CYP) 2D6 poor metabolizers, and in patients who are taking concomitant strong inhibitors of CYP2D6 or CYP3A4.
Aripiprazole 2-month RTU 960 mg (Ari 2MRTU 960) was developed to extend the maintenance of therapeutic plasma concentrations of aripiprazole from 1 month with AOM 400 mg (AOM 400) to 2 months; 24 this is primarily achieved through an increase in the administered dose, while maintaining minimum aripiprazole concentrations that are similar to those after multiple doses of AOM.Less frequent dosing with Ari 2MRTU 960 (ie, 6 injections per year, versus 12 with AOM) may reduce the medication burden for patients and clinicians and improve adherence to antipsychotic treatment in adult patients with schizophrenia or BP-I. 22,24 part of the development of Ari 2MRTU 960 and Ari 2MRTU 720, a previously developed and validated population pharmacokinetic (popPK) model for characterizing aripiprazole plasma concentrations following administration of oral aripiprazole, or AOM at the gluteal or deltoid sites, was expanded to include the RTU LAI formulation of aripiprazole (Ari RTU LAI).The aims of the popPK model were to characterize the pharmacokinetics (PK) of Ari RTU LAI and to perform simulations to explore aspects of its dosing.The model and simulations, which are reported herein, were presented in poster form at the 2023 Annual Meeting of the American Society of Clinical Psychopharmacology, May 30 to June 2, 2023, Miami, Florida. 25,26

Population PK Model Development
Source Data.Data for oral aripiprazole and AOM injected at the gluteal or deltoid sites were from a previously developed and validated popPK model, as described by Wang et al, 27 which was expanded to include data for the RTU LAI formulation of aripiprazole.Newly added data were from 374 patients with schizophrenia or BP-I, including 240 patients who received at least 1 dose of Ari RTU LAI.In all, the final combined analysis data set comprised 8899 aripiprazole concentration samples from 1191 individuals enrolled in 10 clinical trials.Dosing and sampling information for each trial is detailed in Table S1.
Research protocols for all of the included studies were approved by locally appointed ethics committees/institutional review boards, with voluntary written informed consent obtained from all study participants or their guardians prior to study initiation.A list of specific study sites and institutional review boards is provided in Appendix S1.
Model Development.PK data for the Ari RTU LAI formulation were added to the analysis data for the prior popPK model for oral aripiprazole and AOM, and an absorption compartment specific to the Ari RTU LAI formulation was added.Covariate effects from the prior model were retained, while additional covariates related to the absorption and relative bioavailability of the Ari RTU LAI formulation were explored (ie, injection site and volume).Age, sex, and body mass index were also evaluated as standard demographic variables.
As a first step, potential relationships were investigated graphically by plotting individual random effects versus covariates (scatter plots for continuous covariates, side-by-side box plots for categorical covariates).
Next, formal testing of potential covariate effects on the absorption parameters of the Ari RTU LAI formulation was conducted.The likelihood ratio test was used, with the p-value calculated using the difference in minimum value of objective function ( MVOF) for a prespecified 2-sided hypothesis test of level α (ie, type I error).Testing comprised a modified stepwise approach using forward selection of all statistically significant variables on univariate analysis (P < .01threshold [ MVOF of ≥6.63 for 1 degree of freedom]) and standard stepwise backward elimination that examined the effect of removing 1 covariate at a time (P < .001threshold [ MVOF of ≥10.83 for 1 degree of freedom).Relationships between PK parameters and continuous covariates were evaluated using a power function, while relationships between PK parameters and categorical covariates were evaluated using a proportional function.
Model Evaluation.The goodness of fit of the final model was assessed by standard diagnostic plots (eg, observed concentration versus population predictions [PRED], observed concentration versus individual predictions, conditional weighted residuals [CWRES] versus PRED, and time after first dose).The predictive performance of the final model was evaluated using visual predictive checks (VPCs) by simulating 500 replicates of the combined data set.If significant categorical covariate effects were identified in the final model, simulations were performed to visualize these effects on the concentration-time profile following repeated administration of Ari 2MRTU 960.
Bioanalytical Methods.Across all 10 trials, aripiprazole plasma concentrations were assayed using validated high-performance liquid chromatography with tandem mass spectrophotometry.The calibration curve was linear, with the calibration range varying across the different trials, as follows:

Population PK Model Simulations
Following development of the final popPK model, multiple simulations were conducted to explore aspects of Ari RTU LAI dosing.This included various treatment initiation scenarios, including a switch from AOM 400; prior stabilization on oral aripiprazole 10, 20, or 30 mg; and no prior stabilization on oral aripiprazole.For each initiation scenario, 2 start strategies were considered: an overlap of the first dose of Ari RTU LAI with 14 days of oral aripiprazole 10 or 20 mg, or a 2-injection start (1 dose of Ari 2MRTU 960 and 1 dose of AOM 400 at separate injection sites combined with a single oral dose of aripiprazole 20 mg on treatment Day 1).Ad-ditional simulations included: missed/delayed doses (to evaluate the impact of a lack of adherence, as defined by different scenarios, with Ari 2MRTU 960 doses delayed by 2, 4, 6, and 8 weeks); CYP2D6 metabolizer status (to evaluate lower dosing in CYP2D6 poor metabolizers); drug-drug interactions (to evaluate lower dosing with concomitant use of strong inhibitors of CYP2D6 or CYP3A4); and maintenance dosing (ie, steady state).
For the simulations, a population of 1000 virtual patients was created.Individual PK parameters for the virtual patients were derived from patient covariates and random effects according to the popPK model and were used to simulate individual PK profiles according to each dosing scenario.Model covariates (ie, sex, body mass index, and CYP2D6 metabolizer status, as included in the popPK model) were assigned to the virtual patients by resampling from the total patient population used to develop the final popPK model (n = 1191).To best reflect the target patient population, individual random effects on PK parameters were resampled from subgroups of patients with the most relevant and robust information.Across almost all simulation scenarios, the virtual patient population remained constant, comprising a mixture of normal and poor metabolizers of CYP2D6 as obtained by the resampling procedure, with poor metabolizers making up approximately 5% of the population.The only exception to this was for scenarios that required all patients to be normal or poor CYP2D6 metabolizers; in this situation, the status was changed accordingly for the whole virtual patient population.For simulations involving AOM, the site of administration was the gluteal muscle to allow easier comparison to Ari RTU LAI, which is gluteal only.
Simulation results were presented graphically by plotting median values over time, with reference lines added to mark certain concentration levels of interest (ie, to facilitate interpretation of the results and comparisons across different plots).These included 95 ng/mL (the plasma concentration identified as the lower efficacy threshold for the prevention of impending relapse); this was based on an exposureresponse model analysis for AOM 400 that showed that a patient with schizophrenia and a predicted aripiprazole minimum concentration (C min ) of 95 ng/mL or greater was 4.41 times less likely to relapse than a patient with a C min less than 95 ng/mL. 28Additionally, a value of 95 ng/mL is consistent with 94 ng/mL, which is the median steady-state minimum aripiprazole plasma concentration (C min,ss ) for oral aripiprazole 10 mg 29 (ie, the lowest daily dose of oral aripiprazole approved for the treatment of adults with schizophrenia); it is also broadly similar to a therapeutic threshold of approximately 100-120 ng/mL reported Diagram of the base model describing aripiprazole pharmacokinetics following oral, AOM, and Ari RTU LAI administration.*Vc/F has a different numerical value for the Ari RTU LAI formulation.AOM 400, aripiprazole once-monthly 400 mg; Ari, aripiprazole; CL/F, apparent clearance; F rel,AOM , relative bioavailability of the AOM formulation; F rel,LAI , relative bioavailability of the Ari RTU LAI formulation; K a,deltoid , absorption rate constant for the AOM formulation at the deltoid site; K a,gluteal , absorption rate constant for the AOM formulation at the gluteal site; K a,LAI , absorption rate constant for the Ari RTU LAI formulation; K a,oral , absorption rate constant for the oral formulation; LAI, long-acting injectable; Q 1 /F, apparent intercompartmental clearance 1; Q 2 /F, apparent intercompartmental clearance 2; RTU, ready-to-use; Vc/F, apparent central volume; Vp 1 /F, apparent volume in peripheral compartment 1; Vp 2 /F, apparent volume in peripheral compartment 2.
1][32] A reference line indicating 534 ng/mL (the 75th percentile of simulated maximum concentration at steady state [C max,ss ] values following administration of once-daily oral aripiprazole 30 mg/day) was also included; this functioned as an indicator of a conservative upper limit of a clinically relevant concentration range, consistent with previous research. 33ich PK profiles were simulated by sampling concentrations every 2 hours after dosing for 24 hours following oral dosing and every 24 hours after dosing following AOM or Ari RTU LAI dosing.

Software and Tools
The popPK model was developed using nonlinear mixed effects modeling software (NONMEM Version 7.4.2;Icon Development Solutions).Estimation was performed by first-order conditional estimation with interaction; if satisfactory convergence could not be achieved, estimation was performed by stochastic approximation expectation-maximization followed by importance sampling for evaluating the MVOF.Data preparation, graphical analysis, model diagnostics, and statistical summaries were performed using R (Version 4.0.5 and above; R Foundation for Statistical Computing) and RStudio (Version 1.2.5042 and above; RStudio, Inc.).All simulations were performed using Pumas (Version 2.0; Pumas-AI, Inc.).

Population PK Model Development
Base Model.As a first step, external validation of the prior model (which characterized aripiprazole plasma concentrations following administration of oral aripiprazole or AOM) was undertaken using new AOM 400 data from Study 031-201-00181.The prior model was able to characterize the new data reasonably well, although a small degree of overprediction was observed at treatment initiation in the CWRES versus time panel and the CWRES versus PRED panel (data not shown).
After adding PK data for the Ari RTU LAI formulation to the input file, the model that best fit the data was a 3-compartment model with linear elimination and different absorption models for the oral, AOM, and Ari RTU LAI formulations (Figure 1).Absorption of the Ari RTU LAI formulation was modeled by a parallel zero-order and lagged first-order process that accounted for the complex absorption process inherent to LAIs.Absorption of the Ari RTU LAI formulation was parameterized by the fraction of total dose absorbed by the first-order process rather than the zero-order process (FRAC1), the duration of zero-order absorption, the time lag for the start of the first-order process, the absorption rate constant (K a ) for the first-order process (K a RTU), and bioavailability of the Ari RTU LAI formulation relative to the oral formulation (F relative RTU).Except for apparent central volume (which is different for the RTU LAI formulation versus the other formulations of aripiprazole), all other disposition parameters (ie, apparent clearances and peripheral volumes) were shared between the formulations and were held fixed to the values in the prior model.Interindividual variability for deposition and RTU LAI absorption parameters, as well as residual variability, were reestimated.
Covariate Analysis (in Addition to Retained Covariates).The effect of injection site on K a RTU, sex on K a RTU, and sex on FRAC1 were found to be statistically significant on univariate analysis and were carried forward.The effect of injection site on K a RTU was dropped from the final covariate model, as only a minority of injections were administered at the deltoid site (the intended injection site for Ari 2MRTU 960 or Ari 2MRTU 720 is gluteal), and in a sensitivity analysis, parameter estimates for the base model without the deltoid injection data were very similar to the base model with all data.As such, only the effects of sex on FRAC1 and on K a RTU were retained in the final covariate model, with neither removed in a backwards elimination step.
Final Model.Parameter estimates for the final model are shown in Table S2; parameters that were not fixed were estimated with percent relative standard errors of less than 15%.
Per the final model, for typical individuals, the first peak in concentration occurred at 6.75 days (162.1 hours) after dosing for the Ari RTU LAI formulation.The fractions absorbed by the first-order and zeroorder processes were 79.3% and 20.7%, respectively, in women and 62.7% and 37.3%, respectively, in men.First-order absorption started after a lag of 17.5 days (419 hours).In women, the K a was 0.00127/hours, corresponding to an absorption half-life of 22.7 days.In men, the K a was 0.00242/hours, corresponding to an absorption half-life of 11.9 days.Following the first dose, the second peak occurred at 41 days after dosing in a typical woman and at 32 days after dosing in a typical man.At steady state, the second peak occurred at 33 days after dosing in a typical woman and at 28 days after dosing in a typical man.F relative of the Ari RTU LAI formulation was 1.58 versus the oral formulation and 1.06 versus the AOM formulation.
Standard goodness-of-fit diagnostic plots for the final model are shown in Figure 2; overall, the final model was able to adequately describe aripiprazole concentrations following administration of the Ari RTU LAI formulation.The slight overprediction at treatment initiation that was noted in the external validation of the prior model with the AOM 400 data from Study 031-201-00181 was not present in the fit of the final model.
The predictive performance of the model was considered adequate based on prediction-corrected VPC (pcVPC) and nonparametric VPC.Overall, the pcVPC showed good predictive performance of the final model (Figure 3), as evidenced by distributions of the observed data that were comparable with the 90% prediction intervals for the individual predictions.
Impact of Covariates.Regarding retained covariates, K a RTU (ie, the absorption rate constant) was 90.7% higher in men versus women, while FRAC1 (ie, the fraction of first-order absorption) was 56.0% lower in men versus women.
The impact of covariates on secondary PK parameters was evaluated by simulation of typical PK profiles following repeated administrations of Ari 2MRTU 960 mg every 8 weeks until steady state.The results of simulations to demonstrate the effects of sex on aripiprazole steady-state exposure parameters (ie, average concentration at steady state [C avg,ss ], C max,ss , and C min,ss ) according to CYP2D6 metabolizer status are shown in Table S3, while a graphical assessment of the effect of retained covariates is shown in Figure S1.Although men had higher peak concentrations (12% higher in normal metabolizers and 6% higher in poor metabolizers) and lower trough concentrations (22% lower in normal metabolizers and 10% lower in poor metabolizers) of aripiprazole than women, no difference in C avg,ss was observed between the sexes.As expected, simulated aripiprazole concentrations were higher in poor versus normal metabolizers of CYP2D6; based on simulated data (Table S3), C avg,ss in CYP2D6 poor metabolizers was approximately 2 times higher (97.4%) than that in normal metabolizers.

Population PK Model Simulations
The final popPK model was used to simulate aripiprazole plasma concentration-time profiles as a means of exploring aspects of Ari 2MRTU dosing, reflecting various scenarios that might arise in clinical practice.Overall, from initiation through Week 32, plasma concentrations of aripiprazole were comparable between Ari 2MRTU 960 and AOM 400 (Figure 4).
Treatment Initiation.Across multiple treatment initiation scenarios, Ari 2MRTU 960 resulted in plasma concentrations of aripiprazole that were comparable to AOM 400.This was the case for: initiation of Ari 2MRTU 960 following a switch from AOM 400 (Figure 5A); initiation of Ari 2MRTU 960 plus 14 days of overlapping oral aripiprazole in patients previously stabilized on oral aripiprazole 10 or 20 mg/day (Figure 5B, C); initiation of Ari 2MRTU 960 plus 14 days of overlapping oral aripiprazole in patients not previously stabilized on oral aripiprazole (Figure 5D  administration of AOM 400 on Day 0 plus 14 days of overlapping oral aripiprazole 20 mg/day in patients not previously stabilized on oral aripiprazole (Figure S2).
Scenarios involving a 2-injection start (ie, Ari 2MRTU 960 plus AOM 400 at separate injection sites plus a single dose of oral aripiprazole 20 mg, all on the first day of treatment) resulted in more stable plasma concentrations of aripiprazole, compared with a single dose of Ari 2MRTU 960 plus 14 days of oral aripiprazole 10 or 20 mg/day.This was the case for patients with prior stabilization on oral aripiprazole 10 mg/day (Figure S3A), 20 mg/day (Figure S3B), and 30 mg/day (Figure S3C) and patients without prior stabilization on oral aripiprazole (Figure S3D).
Missed/Delayed Doses.Missed/delayed doses were simulated to explore the impact of imperfect adherence with Ari 2MRTU 960 (Figure 6).In scenarios in which Ari 2MRTU 960 dosing was delayed by 2, 4, 6, and 8 weeks, trough median plasma concentrations at the time of dosing were 191, 144, 110, and 80.1 ng/mL, respectively.At 56 days after the delayed dose had been administered, trough median plasma concentrations had increased to 231, 217, 205, and 196 ng/mL, respectively (in the last case, the dose was given with 14 days of overlapping oral aripiprazole 10 mg/day).
Concomitant Long-Term Use of a CYP2D6 or CYP3A4 Inhibitor.Ari 2MRTU 720 plus a concomitant CYP3A4 strong inhibitor resulted in median steadystate plasma concentrations that were comparable to Ari 2MRTU 960 without a concomitant CYP3A4 or CYP2D6 strong inhibitor.In contrast, administration of Ari 2MRTU 720 with a concomitant CYP2D6 strong inhibitor resulted in higher median steady-state plasma concentrations versus Ari 2MRTU 960 without a concomitant CYP3A4 or CYP2D6 strong inhibitor (Figure S5).
Maintenance Dosing (Steady State).Simulated steadystate PK parameters for Ari 2MRTU 960 and Ari 2MRTU 720 are shown in Table 1.The median steadystate terminal elimination half-life was 29.4 days for both Ari 2MRTU 960 and Ari 2MRTU 720.

Discussion
PopPK analyses are a well-established quantitative means of explaining variability in drug concentrations among individuals, including as a result of intrinsic factors, extrinsic factors, and drug-related factors that affect drug disposition. 34Such analyses are frequently undertaken as part of the drug development process to inform dosing regimens to be tested in future clinical trials.PopPK models can also be used to simulate PK profiles for scenarios that are beyond those that have been clinically studied.
This publication describes the development and use of a popPK model to characterize the PK of an LAI formulation of aripiprazole that is administered every 2 months.Based on the current analysis, aripiprazole PK following administration of Ari RTU LAI was best characterized by a linear 3-compartment model with parallel zero-and lagged first-order absorption.In an analysis of potential extrinsic and intrinsic covariates affecting the absorption characteristics of Ari RTU LAI, sex was identified as a statistically significant covariate; specifically, sex was found to affect the fraction and rate constant of first-order absorption, with faster absorption observed in men versus women.The finding of sex as a statistically significant covariate of Ari RTU LAI absorption was consistent with AOM absorption in an earlier model on which the current model was based; 28 however, a direct comparison was not possible due to the different structures of the absorption models (ie, first-order absorption for AOM versus parallel zero-and first-order absorption for Ari RTU LAI).Although sex was found to affect both the fraction and rate constant of the first-order absorption for Ari RTU LAI, it was not expected to be clinically relevant as a result of a combination of these effects.This was confirmed by simulations of typical PK profiles following repeated administration of Ari 2MRTU 960 that showed no difference in C avg,ss values in men versus women.The effect of CYP2D6 metabolizer status was retained as a covariate from the prior model, with simulated C avg,ss found to be approximately 2-fold (97.4%) higher in poor versus normal CYP2D6 metabolizers.Reference lines represent the plasma concentration identified as the lower efficacy threshold for the prevention of impending relapse (95 ng/mL), 28 and 75th percentile of simulated C max,ss following a daily dose of oral aripiprazole 30 mg (534 ng/mL).AOM 400, aripiprazole once-monthly 400 mg; Ari 2MRTU 960, aripiprazole 2-month ready-to-use 960 mg.The final popPK model was used to conduct simulations to evaluate dosing of Ari 2MRTU 960 and Ari 2MRTU 720 across multiple realistic clinical situations, with results demonstrating that the PK exposure profile for Ari RTU LAI was comparable to AOM, while also sustaining efficacious plasma concentrations for the entire 2-month dosing interval.The simulations accounted for various treatment initiation scenarios, which accounted for different prior treatments (ie, whether or not patients were stabilized on oral aripiprazole or AOM 400).The use of overlapping oral aripiprazole for 14 days, or a 2-injection start with Ari 2MRTU 960 plus AOM and a single dose of oral aripiprazole, were also simulated as a means of more rapidly Reference lines represent the plasma concentration identified as the lower efficacy threshold for the prevention of impending relapse (95 ng/mL), 28 and 75th percentile of simulated C max,ss following a daily dose of oral aripiprazole 30 mg (534 ng/mL).AOM 400, aripiprazole once-monthly 400 mg; Ari 2MRTU 960, aripiprazole 2-month ready-to-use 960 mg; C max,ss , maximum concentration at steady state.reaching effective exposures, given the slow absorption of Ari 2MRTU 960.Simulations also accounted for Ari 2MRTU 720 mg in poor metabolizers of CYP2D6 and in long-term users of strong CYP2D6 and CYP3A4 inhibitors.This is important since CYP2D6 metabolizer status is known to influence the PK of aripiprazole, and per the label for AOM, a dose of 300 mg every 4 weeks is recommended in CYP2D6 poor metabolizers.Additionally, CYP2D6 and CYP3A4 are both involved in the elimination of aripiprazole. 35Finally, missed/delayed doses were simulated to assess the impact of imperfect adherence with a dosing interval of every 2 months.With a 6-week delay, simulated median aripiprazole concentrations remained above 95 ng/mL; however, this was not the case when dosing was delayed by 8 weeks.Consistent with these findings, it is recommended in the product label that, if more than 14 weeks have elapsed since the last injection of Ari 2MRTU 960 (ie, if there has been a delay of more than 6 weeks from when the next injection was scheduled), the next administered injection should be administered with 14 days of concomitant oral aripiprazole.The apparent central volume was notably different for the Ari RTU LAI formulation (2035 L) versus the other formulations of aripiprazole (93.4 L).However, this finding should be interpreted with caution since the absorption of Ari RTU LAI is very slow, with the terminal phase being absorption driven.This situation-where the rate of absorption is slower than the rate of elimination-is known as flip-flop kinetics. 36lthough not directly comparable, data show that, for long-acting intramuscular formulations with flip-flop kinetics, the computed volume of distribution can be considerably larger than that with intravenous administration. 37While flip-flop kinetics also applies for the AOM formulation, the difference in the absorption parameterization (ie, first-order absorption for AOM versus parallel zero-and first-order absorption for Ari RTU LAI) creates a situation in which the model accommodates each formulation differently.As such, the Ari RTU LAI formulation, with its more complex dual-peak profile, cannot be directly compared to AOM or the oral formulation in terms of the apparent central volume of distribution.It is important to note that differences in the apparent central volume of distribution have no impact on the simulation results.Individual PK parameters for the virtual patients were derived from the patients' covariates and random effects to simulate individual PK profiles according to each dosing scenario, with individual random effects on PK parameters resampled from subgroups of patients with the most relevant and robust information.This means that concentrations contributed by the oral and AOM formulations were simulated using resampled parameters relevant to these formulations; similarly, concentrations contributed by the Ari RTU LAI formulation were simulated using resampled parameters relevant only to the Ari RTU LAI formulation.
The data presented here provide clinicians with information regarding a variety of dosing scenarios for a 2-monthly RTU LAI formulation of aripiprazole.It is impractical/infeasible to include all potential dosing scenarios in efficacy and safety trials, but popPK modeling and simulation addresses this, by generating PK profiles for scenarios beyond those specifically studied as part of the clinical trial program.In this manner, modeling and simulations act as "virtual" clinical trials, allowing more scenarios to be investigated than could be conducted in real-life clinical trials.Nonetheless, any simulations presented here must be viewed in context with the approved use of Ari 2MRTU 960 and Ari 2MRTU 720, according to patient type, prior treatment, and/or country/region.

Conclusion
In conclusion, a popPK model was developed to characterize the PK of Ari 2MRTU 960, which was then used to perform model-based simulations exploring various aspects of dosing of this formulation.The final model adequately described the PK of plasma aripiprazole concentrations following administration of Ari 2MRTU 960; the model also established estimates for key popPK parameters and identified sources of variability in drug exposure.Simulations across a variety of dosing regimens that could reasonably be encountered in clinical practice demonstrated a PK exposure profile that was comparable to AOM, with efficacious plasma concentrations sustained across the entirety of each 2-month dosing interval.
, E); initiation of Ari 2MRTU 960 on Day 28 following administration of AOM 400 on Day 0 plus 14 days of overlapping oral aripiprazole in patients previously stabilized on oral aripiprazole 20 mg/day (Figure 5F); and initiation of Ari 2MRTU 960 on Day 28 following Population prediction (ng/mL)

Figure 2 .
Figure 2. Goodness-of-fit plots for the final aripiprazole population pharmacokinetic model.Circles represent observed data.Solid black lines in the figures in the upper panel represent a line of unity.Solid black lines in the figures in the lower panel indicate the zero intercept.Black dashed lines with conditional weighted residuals equal to ±4 are provided as reference lines to check for potential outliers.Gray dashed lines represent a locally estimated scatterplot smoother.AOM, aripiprazole once-monthly; Ari RTU LAI, aripiprazole ready-to-use long-acting injectable.

Figure 3 .
Figure 3. Prediction-corrected visual predictive checks for the final aripiprazole population pharmacokinetic model.Dashed lines represent observed data (median, red; 5th and 95th percentile, black).Solid lines represent simulated data (median, red; 5th and 95th percentile, black).Shaded regions represent the 95% confidence interval of the simulated median (red) and the 5th and 95th percentile (blue).AOM, aripiprazole once-monthly; Ari RTU LAI, aripiprazole ready-to-use long-acting injectable.

Figure 4 .
Figure 4. Simulated median aripiprazole concentration-time profiles from initiation to Week 32.For comparative purposes, the figure includes a red line, representing simulation results for the same patients treated with AOM 400 (with the same oral 10 mg overlap during initiation).Reference lines represent the plasma concentration identified as the lower efficacy threshold for the prevention of impending relapse (95 ng/mL),28 and 75th percentile of simulated C max,ss following a daily dose of oral aripiprazole 30 mg (534 ng/mL).AOM 400, aripiprazole once-monthly 400 mg; Ari 2MRTU 960, aripiprazole 2-month ready-to-use 960 mg.

Figure 5 .
Figure 5. Simulated median aripiprazole concentration-time profiles following initiation of Ari 2MRTU 960 across various prior treatment scenarios.For comparative purposes, each figure includes a red line, representing simulation results for the same patients treated with AOM 400 (with the same oral overlap, where relevant).Reference lines represent the plasma concentration identified as the lower efficacy threshold for the prevention of impending relapse (95 ng/mL),28 and 75th percentile of simulated C max,ss following a daily dose of oral aripiprazole 30 mg (534 ng/mL).AOM 400, aripiprazole once-monthly 400 mg; Ari 2MRTU 960, aripiprazole 2-month ready-to-use 960 mg; C max,ss , maximum concentration at steady state.

Table 1 .
Simulated Key Steady-State Pharmacokinetic Parameters for Ari 2MRTU 960 and Ari 2MRTU 720 28mulated median aripiprazole concentration-time profiles following Ari 2MRTU 960 as scheduled or after a delay of 2, 4, 6, or 8 weeks in patients already stabilized on Ari 2MRTU 960.Reference lines represent the plasma concentration identified as the lower efficacy threshold for the prevention of impending relapse (95 ng/mL),28and 75th percentile of simulated C max,ss following a daily dose of oral aripiprazole 30 mg (534 ng/mL).Ari 2MRTU 960, aripiprazole 2-month ready-to-use 960 mg; C max,ss , maximum concentration at steady state.