An alternative start regimen with aripiprazole once-monthly in patients with schizophrenia: population pharmacokinetic analysis of a single-day, two-injection start with gluteal and/or deltoid intramuscular injection

Abstract

Background

The single-injection start regimen for aripiprazole once-monthly 400 mg (AOM 400) in patients with schizophrenia requires a single intramuscular injection in the gluteal or deltoid site and 14 days of concurrent oral therapy. A simplified, single-day regimen of two injections at separate gluteal and/or deltoid injection sites, together with a single 20-mg dose of oral aripiprazole on the 1st day, was assessed.

Patients and methods

A previously developed population-pharmacokinetic (popPK) model for characterizing aripiprazole PK following oral administration and gluteal intramuscular depot injection was expanded to include deltoid injection. Simulations were conducted to assess PK profiles following various (including two-injection) start regimens. Postmarketing data on patients who received higher-than-recommended AOM doses were used to assess overall safety/tolerability.

Results

The two-injection start regimen with a single concurrent oral dose displayed a comparable PK profile to the single-injection start regimen with concurrent 14-day oral administration in simulations. The safety assessment indicated the two-injection start regimen was unlikely to be associated with safety concerns beyond those expected with a single-injection start regimen.

Conclusion

These data support use of the two-injection start regimen in clinical practice to reduce reliance on daily oral administration and optimize the therapeutic benefits of AOM 400 in patients with schizophrenia.

Keywords:

• Antipsychotic treatment
• long-acting injectable
• pharmacokinetic modeling

View correction statement:

Correction

Introduction

Schizophrenia is a chronic, disabling, and progressive psychiatric disorder characterized by recurrent psychosis, with the majority of patients experiencing multiple relapses that involve exacerbations of acute illness^1,2. Adherence to antipsychotic treatment is crucial to the success of long-term maintenance treatment; however, nonadherence is common in this patient population^3–5. Nonadherence has been associated with increased risk of relapse and hospitalization, and nonadherence and discontinuation of medication have been identified as the most significant predictors of relapse^4,6–8. Long-acting injectable (LAI) formulations of atypical antipsychotics are valuable treatment alternatives to oral agents, and provide the opportunity to achieve sustained plasma therapeutic concentrations in the therapeutic range and, consequently, improvement in long-term outcomes^9–11.

Aripiprazole once-monthly (AOM) 400 mg (AOM 400), an extended-release intramuscular (IM) suspension of aripiprazole, is the 1st dopamine D₂-receptor partial agonist available as a LAI formulation for the treatment of schizophrenia. AOM 400 is indicated for the treatment of schizophrenia in adult patients in the United States and Canada, and in Europe for maintenance treatment of schizophrenia in adult patients stabilized with oral aripiprazole. AOM 400 is also approved for maintenance treatment of manic or mixed episodes of bipolar-I disorder in adult patients in the United States and Canada^12–14.

Although the mechanism of action of aripiprazole remains unknown, it is believed to exert its effects through partial agonist activity at dopamine D2 and serotonin 1 A receptors, and antagonistic activity at serotonin 2 A receptors^14,15. In pivotal trials, AOM 400 reduced the rate of, and delayed the time to, impending relapse relative to placebo or a subtherapeutic dose of aripiprazole once-monthly in patients with schizophrenia^16,17.

The approved starting and maintenance dose of AOM is 400 mg, with the option to reduce to 300 mg for tolerability reasons^13,14. With the single-injection start regimen, the 1st dose of AOM 400 is recommended to be administered as a single IM injection in the gluteal or deltoid site with concomitant oral aripiprazole (10–20 mg) for 14 consecutive days to maintain therapeutic drug concentrations (94.0–534.0 ng/mL) during initiation of therapy^13,14,18. The US and Canadian prescribing information further state that patients not taking oral aripiprazole may continue taking their current oral antipsychotic for the 1st 14 days of treatment with AOM 400 as aripiprazole concentrations increase to therapeutic levels^12,14.

In patient populations considered to be at potential risk for adherence-related relapse or suboptimal treatment outcomes, initiation of LAIs from Day 1 and attaining therapeutic plasma concentrations without reliance on an additional 14 days of oral tablet dosing during initiation may offer numerous treatment advantages, including: reduction of adherence-related undertreatment and consequent relapse during the initiation phase of treatment; lessening the need for extended length of stay in an acute hospital setting; minimizing the burden on healthcare providers, families, and other caregivers to supervise oral medications; and enhancing patient/provider options regarding choice of antipsychotic medications/formulations.

In the present investigation, a previously developed population-pharmacokinetic (popPK) model (unpublished data [submitted for publication]) was modified to simulate the plasma concentration of aripiprazole after various initiation scenarios, leading to a simplified, alternative start regimen consisting of two AOM 400 injections at separate gluteal and/or deltoid injection sites, together with a single 20-mg dose of oral aripiprazole on the 1st day of AOM 400 treatment. Postmarketing data on patients receiving higher-than-recommended doses of AOM (specifically, 2 simultaneous injections, which mimics the alternative initiation regimen) were used to assess the overall safety and tolerability of the two-injection start regimen.

PopPK analysis

Population pharmacokinetic (popPK) analysis is a well-established method that can describe the time course of drug exposure and quantify and explain variability in drug concentrations among individuals^19–21. PopPK analysis integrates all relevant PK information across a range of doses to identify factors that can affect a drug’s exposure. Such information can come from studies with rich PK sampling or sparse PK sampling, after a single dose or at steady state, and from healthy individuals or the patient population. Once a fit-for-purpose population PK model has been developed, it can be used for simulations of clinical scenarios that have not been studied, which can in turn inform strategies for managing dosing and administration in specific subpopulations, assist in planning of subsequent studies, and support drug labeling²². Examples of use of popPK data in regulatory decisions include clevidipine, a short-acting dihydropyridine calcium-channel antagonist, which is indicated for short-term intravenous management of blood pressure (popPK data were used to secure approval of an alternate dosing regimen) and lamotrigine, an anticonvulsant indicated for adjunctive treatment of partial epileptic seizures (popPK data were used to secure approval of an extended-release formulation as an alternative to immediate-release)²³. The present work applies the principles of popPK analysis to support an alternative treatment initiation regimen for aripiprazole once-monthly 400 mg (AOM) in lieu of a clinical study.

Materials and methods

Data sources

The present analysis utilized aripiprazole PK data from 7 clinical trials: data from a previously developed and validated popPK model (comprising data from 5 trials that included oral aripiprazole administration and/or IM depot injections in gluteal muscles) (unpublished data [submitted for publication]), plus data from two additional trials conducted to support the addition of the deltoid muscle as the site of administration. In all, 8214 aripiprazole concentrations (16% oral, 65% gluteal, 16% deltoid, and 3% triceps or thigh administration) from 817 healthy adult subjects (n = 52) or adult patients with schizophrenia (n = 765) were included in the final combined analysis data set.

Model development

A previously developed and validated popPK model for characterizing aripiprazole PK following oral administration and gluteal IM depot injections (unpublished data [submitted for publication]) was expanded to include the deltoid site of injection. The previously validated model was a 3-compartment model with sigmoid absorption (zero-order absorption to the oral compartment followed by 1st-order absorption [Ka]) for oral administration and 1st-order absorption for IM (mainly gluteal) administration (IMKa); the present model incorporated the deltoid injection site into the previously developed model by adding a deltoid depot compartment with a separate 1st-order absorption rate constant (DKa; Figure 1).

Figure 1. Diagram of the structural model describing aripiprazole pharmacokinetics following oral administration and intramuscular (IM) injections in gluteal and deltoid muscles. Update of the previously developed model to include the deltoid site of injection is denoted in red. Abbreviations. CL, apparent clearance; CMT, compartment; DKa, 1st-order absorption rate constant for deltoid administration; F_relative, bioavailability for IM depot relative to oral administration; IMKa, 1st-order absorption rate constant for IM administration (mainly gluteal); Ka, 1st-order oral absorption rate constant; Q1, intercompartmental clearance 1; Q2, intercompartmental clearance 2; R1, rate of dose into oral compartment; Vc, apparent central volume of distribution; Vp1, volume of distribution in peripheral compartment 1; Vp2, volume of distribution in peripheral compartment 2.

The population mean of each PK parameter was fixed to the value estimated in the previously developed model for oral and gluteal administration, except for DKa, which was estimated using data following deltoid administration. The interindividual variability for Ka was fixed to the value estimated in the previously developed model, while that for clearance (CL), central volume of distribution (Vc), and the 1st-order absorption rate constants following IM injections were estimated (DKa) or re-estimated (IMKa) using the final combined analysis data set. The previous model also included the following covariate effects: (i) the effect of cytochrome P450 2D6 (CYP2D6) metabolism status (extensive [EM; also known as “normal metabolizer”] vs poor metabolizer [PM]) on CL, (ii) the effects of co-administration of strong inhibitors of CYP2D6 or CYP3A4 on CL, and (iii) the effects of gender and body mass index on IMKa. Parameter estimation for the previously developed model is presented in Table 1. It was assumed that the covariate effect remained the same as for the previously developed model, and that the gender and body mass index effects on IMKa, estimated from data following IM injection mainly in the gluteus maximus, were also present for the deltoid injection. No further structural change or covariate analysis was conducted. The goodness of fit of the final combined model was assessed by standard diagnostic plots (observed vs population prediction [PRED], observed vs individual prediction, conditional weighted residual error [CWRES] vs PRED, and CWRES vs time).

Table 1. Parameter estimation for previously developed model without deltoid data.

Parameter	Final parameter estimate		Magnitude of interindividual variability (% CV)
	Population mean	% SEM	Final estimate	% SEM
Oral Ka, 1/h	0.540	Fixed	65.88	Fixed
CL for EM, L/h	3.71	4.0	38.34	6.9
CL for PM, L/h	1.88	6.9
Proportional change in CL for CYP2D6 inhibitor	−0.511	Fixed
Proportional change in CL for CYP3A4 inhibitor	−0.237	Fixed
Vc, L	93.4	8.8	124.50	15.2
Q1, L/h	0.591	Fixed	NE	NA
Vp1, L	118	Fixed	NE	NA
Q2, L/h	28.8	Fixed	NE	NA
Vp2, L	134	Fixed	NE	NA
R1, mg/h	9.33	Fixed	NE	NA
IMKa, 1/h	0.000904	5.3	55.59	8.2
Frelative for IM depot	1.48	4.9	NE	NA
IMKa_BMI: power for (BMI/28)	−0.975	11.5	NE	NA
IMKa_male: proportional shift for males	0.346	28.9	NE	NA
Phase 1 RV (% CV)^†	24.23	8.4	NA	NA
Phase 3 RV (% CV)^‡	28.11	4.7	NA	NA
Minimum value of objective function: 48,892.907

^†Proportional residual error variability for phase 1 data from aripiprazole once-monthly 400-mg development program. ^‡Proportional residual error variability for phase 3 data from aripiprazole once-monthly 400-mg development program. Note: estimate of apparent clearance (CL) for extensive metabolizers (EMs) was correlated with estimate of relative bioavailability (F_relative) for intramuscular (IM) depot (r = 0.901). Abbreviations. BMI, body mass index; CV, coefficient of variation; CYP, cytochrome P450; IMKa, IM depot 1st-order absorption rate constant; Ka, 1st-order absorption rate constant; NA, not applicable; NE, not estimated; PM, poor metabolizer; Q1, intercompartmental clearance 1; Q2, intercompartmental clearance 2; RV, residual variability; R1, rate of dose into depot; SEM, standard error of mean; Vc, apparent central volume of distribution; Vp1, volume of distribution in peripheral compartment 1; Vp2, volume of distribution in peripheral compartment 2.

Pharmacokinetic simulations

The final combined popPK model was utilized to simulate aripiprazole plasma concentration-time profiles following oral (single dose), gluteal, and/or deltoid administration of aripiprazole with the goal of identifying a regimen that would: (1) reduce reliance on extended oral tablet dosing during initiation; (2) maintain aripiprazole concentrations within a previously established therapeutic window¹⁸ with upper and lower bounds corresponding to simulated median minimum aripiprazole concentrations at steady state (C_min, ss) following daily administration of oral aripiprazole 10 mg (94.0 ng/mL) and the 75th percentile of maximum aripiprazole concentrations at steady state (C_max, ss) following daily administration of oral aripiprazole 30 mg (534 ng/mL), respectively; and 3) result in plasma concentrations similar (i.e. median, and 25th–75th and 5th–95th percentiles of concentration) to that of the single-injection start regimen (one AOM 400 injection together with 14 days of oral aripiprazole 10–20 mg).

Several single-day initiation regimens were simulated before the alternative (two-injection) start regimen was ultimately selected (based on a comparative analysis with the single-injection regimen). Simulations were then performed to assess the difference in time to achieve concentrations within the therapeutic window. These comprised: (1) simulation of aripiprazole PK for subjects with prior stabilization on oral aripiprazole; (2) simulation of aripiprazole PK for subjects without prior stabilization on oral aripiprazole; (3) simulation of aripiprazole PK following a missed 2nd, 3rd, 4th, or steady-state dose of AOM 400; and (4) simulation of aripiprazole PK for CYP2D6 EM vs PM subjects. In CYP2D6 PM subjects, the currently approved IM depot initiation dose for AOM is recommended to be reduced from 400 to 300 mg due to an ∼50% lower apparent clearance of aripiprazole)¹³. This dose reduction is consistent with recommendations from the Dutch Pharmacogenetics Working Group²⁴; the Clinical Pharmacogenetics Implementation Consortium has not issued guidance specific to aripiprazole.

Software and tools

Population PK modeling and simulation was performed using NONMEM (version 7.4.3, Icon Development Solutions, Ellicott City, MD) in RStudio (version 1.0.153) through the R package metrumrg. The 1st-order conditional estimation method with interaction was used for all stages of the model development process. Post-modeling analysis was conducted using R version 3.3.3 or above.

High-dose safety analysis (postmarketing data)

The postmarketing assessment of higher-than-recommended aripiprazole doses comprised safety/tolerability data derived from a cumulative review (through 24 June 2019) of cases from the AOM safety database (spontaneous reports, clinical trial reports, and literature reports) using the following search terms: “prescribed overdose”; “intentional overdose”; “accidental overdose”; “overdose”; and “extra dose administered.” Overdose cases included the following categories: doses of AOM >400 mg/administration; administration of AOM 300 or 400 mg more frequently than every 26 days; initial dose of AOM 400 and treatment with oral aripiprazole 10–20 mg for >14 consecutive days; and doses of oral aripiprazole >30 mg and/or AOM doses >400 mg in cases where both oral aripiprazole and AOM were considered as co-suspects.

Ethics statement

For the portion of the study that involved non-simulated patient data (the postmarketing safety analysis), data were extracted from the sponsor (Otsuka’s) pharmacovigilance database, with permission to use these data for the present study. Results were obtained from a database in a manner that preserved patient confidentiality; thus, no institutional review board/ethics committee approval was required.

Results

Final combined population-pharmacokinetic model

The diagnostic plots and visual predictive checks presented in Figures 2 and 3, respectively, demonstrate that the final combined popPK model adequately described aripiprazole concentrations following oral administration and gluteal and deltoid IM injections. Parameter estimation for the final model is presented in Table 2.

Figure 2. Diagnostic plots of the final combined population-pharmacokinetic model. Circles represents observed data (red representing data following deltoid intramuscular injection); black lines are lines of unity (upper panels) and horizontal lines with intercept zero (lower panels); and blue lines represent trend lines.

Figure 3. Prediction-corrected visual predictive checks for deltoid and gluteal sites of administration for the final combined population-pharmacokinetic model. The visual predictive check following the 1st gluteal injection included pharmacokinetic data from subjects coadministered oral aripiprazole. Red lines correspond to observed data; black lines correspond to predicted data. Solid and dashed lines represent median, and 5th and 95th percentiles, respectively, and shades define the 90% prediction intervals. Open circles represent observed data. CI, confidence interval.

Table 2. Parameter estimation for final combined model.

Parameter	Unit	Definition	Estimate^†	RSE %	Shrinkage %
Fixed effects
R1	mg/h	Rate of dose into oral absorption compartment	9.33 fixed	—	—
Ka	1/h	Oral 1st-order absorption rate constant	0.540 fixed	—	—
CL	L/h	Apparent clearance for subjects who were not poor CYP2D6 metabolizers	3.71 fixed	—	—
CLpm	L/h	CL for subjects who were poor CYP2D6 metabolizers	1.88 fixed	—	—
CL_INH2D6	—	Proportional change in CL in presence of strong CYP2D6 inhibitor	−0.511 Fixed	—	—
CL_INH3A4	—	Proportional change in CL in presence of strong CYP3A4 inhibitor	−0.237 Fixed	—	—
Vc	L	Apparent central volume of distribution	93.4 fixed	—	—
Q1	L/h	Intercompartmental clearance 1	0.591 fixed	—	—
Vp1	L	Volume of distribution in peripheral compartment 1	118 fixed	—	—
Q2	L/h	Intercompartmental clearance 2	28.8 fixed	—	—
Vp2	L	Volume of distribution in peripheral compartment 2	134 fixed	—	—
IMKa	1/h	Gluteal IMKa	0.000904 fixed	—	—
DKa	1/h	Deltoid IMKa	0.000776	6.2%	—
IMKa_BMI	—	Effect of BMI on IMKa and DKa: power for (BMI/28)	−0.975 fixed	—	—
IMKa_male	—	Proportional shift of IMKa and DKa for males	0.346 fixed	—	—
Frelative	—	Relative bioavailability for IM depot	1.48 fixed	—	—
Random effect: IIV (% CV)
Ka	—	IIV on oral Ka	0.434 (65.9 %) fixed	—	77%
CL or CLpm	—	IIV on CL for all subjects	0.153 (39.1%)	6.1%	8%
Vc	—	IIV on Vc	2.18 (148%)	2.5%	44%
IMKa	—	IIV on gluteal IMKa	0.359 (60.0%)	7.7%	24%
DKa	—	IIV on deltoid IMKa	0.237 (48.9%)	24.4%	74%
Residual variability (% CV)
Phase 1^‡	—	Proportional residual error for phase 1 data	0.0693 (26.3%)	—	—
Phase 3^§	—	Proportional residual error for phase 3 data	0.0600 (24.5%)	—	—

^†Fixed indicates parameters fixed to values estimated in previously developed model. ^‡Phase 1 data from aripiprazole once-monthly 400-mg development program. ^§Phase 3 data from aripiprazole once-monthly 400-mg development program. Abbreviations. BMI, body mass index; CYP, cytochrome P450; IIV, interindividual variability; IM, intramuscular; RES, standard error relative to mean; CV, coefficient of variation.

Simulations

For comparison of simulated PK profiles, a virtual population of 817 subjects with similar demographic characteristics to those in the final combined analysis data set was used to ensure that only the dosing regimen changed across simulations. Individual PK parameters for the subjects (all assigned as CYP2D6 EM except for simulation of CYP2D6 PM subjects) were generated from the final combined popPK model and its final parameter estimates. Individual PK profiles following oral, gluteal, and deltoid administration of aripiprazole were simulated using a 2-h sampling interval for 24 h after the preceding oral dose and a 24-h sampling interval for 28 days after the preceding IM depot dose.

Alternative initiation regimens with a single oral dose and two IM depot doses administered on Day 1 at different dose strengths were simulated. Aripiprazole PK following two injections of IM aripiprazole at separate gluteal and/or deltoid injection sites, together with a single 20-mg dose of oral aripiprazole on the 1st day, was comparable to that following the single-injection start regimen.

Two-injection start without prior oral aripiprazole stabilization

The median and 5th − 95th percentile of simulated aripiprazole PK profiles for the single-injection start regimen (AOM 400 with 14 days of 10–20-mg/day oral dosing) and the two-injection start (2 × AOM 400 with 1 day of 20-mg oral dosing) administered as two separate injections in the gluteal and/or deltoid sites are shown in Figure 4.

Figure 4. Simulated median and 5th and 95th percentile aripiprazole concentration-time profiles following the single-injection start regimen or the two-injection start regimen, followed by 400-mg intramuscular depot dose every 28 days. The reference lines represent median steady-state minimum concentration (C_min, ss) following daily dose of 10 mg oral aripiprazole (94.0 ng/mL), 75th percentile of steady-state maximum concentration (C_max, ss) following daily dose of 30 mg oral aripiprazole (534 ng/mL), and 95th percentile of C_max, ss following daily dose of 30 mg oral aripiprazole (741 ng/mL). Shades define 5th − 95th percentiles. IM, intramuscular.

Based on the simulations, the median and 5th − 95th percentiles of the plasma concentrations of aripiprazole following administration of the two-injection start were comparable to the one-injection start regimen of AOM 400 on Day 1 plus oral aripiprazole 10–20 mg/d for 14 days. Median aripiprazole PK profiles following the two-injection start reached therapeutic levels on the 1st day and remained above the lower bound of the therapeutic window (94.0 ng/mL) thereafter. The 95th percentiles of simulated concentrations following the two-injection start were comparable to or lower than that of the single-injection start regimen and below that of the simulated 30-mg oral C_max, ss (741 ng/mL), the latter representing the highest approved oral dose of aripiprazole. Furthermore, the two-injection start had no apparent impact on steady-state maintenance concentration.

Administration of the two-injection start with prior oral aripiprazole stabilization

The median, and 5th, 25th–75th, and 95th percentiles of simulated aripiprazole PK profiles for the single-injection start regimen and the two-injection start regimen administered as two separate injections in the gluteal or deltoid sites of subjects with prior stabilization on an oral aripiprazole dose of 20 mg (the highest typical oral dose a patient would be receiving prior to initiating treatment with AOM 400) or 10 mg are shown in Figure 5.

Figure 5. Simulated median, and 5th, 25th–75th, and 95th percentiles of pharmacokinetic profiles following administration of the single-injection start regimen or the two-injection start regimen for aripiprazole once-monthly initiation to subjects already stabilized on oral aripiprazole 20 mg (top panels) or 10 mg (bottom panels). The reference lines represent median steady-state minimum concentration (C_min, ss) following a daily dose of 10 mg oral aripiprazole (94.0 ng/mL), 75th percentile of steady-state maximum concentration (C_max, ss) following a daily dose of 30 mg oral aripiprazole (534 ng/mL), and 95th percentile of C_max, ss following a daily dose of 30 mg oral aripiprazole (741 ng/mL). Dashed lines represent 5th − 95th percentiles, and shades define 25th − 75th percentiles. IM, intramuscular.

As with the simulations in patients not previously stabilized on oral aripiprazole, the simulated median and 5th–95th percentiles of the aripiprazole PK profiles following administration of the two-injection start were comparable to the single-injection start regimen. Median aripiprazole plasma concentrations remained above the lower threshold of the therapeutic window. The 95th percentiles of simulated concentrations following the two-injection start were comparable to that of the single-injection regimen, or within the upper bound of the therapeutic window.

Subjects who are CYP2D6 poor metabolizers

In simulations of administration of the two-injection start to CYP2D6 EM and PM subjects, CYP2D6 PM subjects exhibited higher aripiprazole C_max and exposure, as expected (Figure 6). Thus, a dose reduction of the two-injection start regimen from two AOM 400 administrations to two 300-mg administrations, along with a single dose of oral aripiprazole 20 mg is recommended for known CYP2D6 PM subjects to ensure concentrations following the two-injection start regimen are comparable to the single-injection start regimen and remain within or slightly above the therapeutic window.

Figure 6. Simulated median aripiprazole concentration-time profiles and boxplots of maximum concentration (C_max) following administration of aripiprazole once-monthly 400-mg initiation regimens to extensive and poor cytochrome P450 2D6 (CYP2D6) metabolizers (top row) and extensive and poor CYP2D6 metabolizers already stabilized on oral aripiprazole (bottom row). Reference lines represent median steady-state minimum concentration (C_min, ss) following a daily dose of 10 mg oral aripiprazole (94.0 ng/mL), 75th percentile of steady-state maximum concentration (C_max, ss) following a daily dose of 30 mg oral aripiprazole (534 ng/mL), and 95th percentile of C_max, ss following a daily dose of 30 mg oral aripiprazole (741 ng/mL). Single-injection initiation regimen: oral aripiprazole 10–20 mg (14 days) plus 400 mg intramuscular (IM) depot (Day 1). For lower 2 graphs, starting concentration at time 0 is average concentration at steady state for CYP2D6 extensive metabolizers stabilized on oral aripiprazole 20 mg and poor metabolizers stabilized on oral aripiprazole 10 mg; for poor metabolizers, prior oral dose was reduced by half (10 instead of 20 mg). Boxplots show the 5th, 25th, 50th, 75th, and 95th percentiles of C_max. EM, extensive metabolizer; PM, poor metabolizer.

Subjects with missed maintenance intramuscular depot dose

In all simulation scenarios with missed maintenance doses (missed 2nd, 3rd, 4th, or steady-state dose of AOM 400), administration of the two-injection start following a missed maintenance IM depot (5 weeks following the last injection for missed 2nd and 3rd doses; 6 weeks for missed 4th and steady-state doses) resulted in median aripiprazole concentrations above the lower threshold of the therapeutic window and similar to those following the single-injection start regimen (Figures 7 and 8).

Figure 7. Simulated median pharmacokinetic profiles following initiation and re-initiation with single-injection or two-injection aripiprazole once-monthly 400 mg initiation regimen 5 weeks after the 2nd intramuscular (IM) depot dose (top 2 panels) or the 3rd IM depot dose (bottom 2 panels). Reference lines represent median steady-state minimum concentration (C_min, ss) following a daily dose of 10 mg oral aripiprazole (94.0 ng/mL), 75th percentile of steady-state maximum concentration (C_max, ss) following a daily dose of 30 mg oral aripiprazole (534 ng/mL), and 95th percentile of C_max, ss following a daily dose of 30 mg oral aripiprazole (741 ng/mL). Single-injection initiation regimen: oral aripiprazole 10–20 mg (14 days) plus 400 mg IM depot (Day 1). ss, steady-state.

Figure 8. Simulated median pharmacokinetic profiles following initiation and re-initiation with single-injection or two-injection aripiprazole once-monthly 400 mg initiation regimen 6 weeks after the 4th intramuscular (IM) depot dose (top 2 panels) or the 5th (steady-state) IM depot dose (bottom 2 panels). Reference lines represent median steady-state minimum concentration (C_min, ss) following a daily dose of 10 mg oral aripiprazole (94.0 ng/mL), 75th percentile of steady-state maximum concentration (C_max, ss) following a daily dose of 30 mg oral aripiprazole (534 ng/mL), and 95th percentile of C_max, ss following a daily dose of 30 mg oral aripiprazole (741 ng/mL). Single-injection initiation regimen: oral aripiprazole 10–20 mg (14 days) plus 400 mg IM depot (Day 1). ss, steady-state.

High-dose safety analysis (postmarketing data)

As of 24 June 2019, there were approximately 13,000 AOM 400 postmarketing cases in the company’s safety database, of which 2% (228 of 13,000 cases) met the predefined criteria for high-dose administration (i.e. doses higher than the recommended dose for AOM). Review of these 228 cases showed that 233 adverse events (AEs) were reported. Of these 233 AEs, 81% (188/233) were categorized as non-serious and 19% (45/233) as serious. Among the 228 cases with high-dose administration AOM 400, no AEs were reported in 102 (45%). The highest aripiprazole dose reported for this group of 228 cases was 1600 mg/week, reported in one case for an unknown duration, and there were no AEs reported with this case. Among the 228 cases, there were only 12 in which a patient received 800 mg of AOM 400 in 1 day (similar to the two-injection start regimen), and 75% of these cases (9/12) had no reported AEs.

Twenty percent, or 46 of the 228 high-dose cases reported an AE that was associated with aripiprazole. The events reported included expected events such as drug ineffective, psychotic disorder, extrapyramidal disorder, respiratory arrest, urinary tract obstruction, akathisia, muscle rigidity, vomiting, abnormal thinking, altered mood, drug interaction, tremor, sedation, swollen tongue, aggression, insomnia, hallucination, bradyphrenia, irritability, dizziness, anaphylactic shock, excessive masturbation, relapse, maniac symptoms, creatine phosphokinase increased, urinary retention, dyspnea, agitation, esophageal dysmotility, confusion, delirium, dyskinesia, sluggishness, fatigue, malaise, lethargy, anxiety, muscle contracture, neuroleptic malignant syndrome, weight gain, bradycardia, tardive dyskinesia, and schizophrenia symptoms.

To determine if higher doses of AOM 400 were associated with higher event frequencies of extrapyramidal symptoms (EPS)-related events and urinary dysfunction, event frequencies reported for the single-injection AOM 400 regimen were compared to those reported in association with high-dose AOM 400 administration (Table 3). From review of the postmarketing data, the frequencies of AEs and serious AEs related to EPS (akathisia, dystonia, dyskinesia, and Parkinson-like events) were lower in the AOM 400 high-dose group vs the single-injection group. Specifically, the frequency of EPS-related events for the AOM 400 high-dose group was 4.6% (43 out if 926 total reported events) compared with 7.3% (2222 out of 30,522 events) for the single-injection group (∼0.37-fold lower), and the frequency of serious EPS-related events for the AOM 400 high-dose group was 5.1% (13 out of 257 events) compared with 7.1% (392 out of 5523 events) for the single-injection group (∼0.28-fold lower). Urinary dysfunction AEs and serious AEs occurred at a higher frequency in the high-dose AOM 400 vs the single-injection group (Table 3); however, this was considered to be non-significant due to the overall low number of cases reported in the high-dose AOM 400 group (3, compared with 56 cases reported in the single-injection group.

Table 3. Comparison of EPS-related and urinary dysfunction event frequencies between AOM 400 single-injection group and AOM 400 high-dose group.

Events	AOM 400 single-injection group				AOM 400 high-dose group^†
	All adverse events	Serious adverse events	Event frequency (%)	Event frequency of serious events (%)	All adverse events	Serious adverse events	Event frequency (%)	Event frequency of serious events (%)
Akathisia SMQ	1042	151	3.4^‡	2.7^§	24	7	2.6^¶	2.7^‡‡
Dystonia SMQ	357	95	0.7^‡	1.7^§	4	3	0.4^¶	1.2^‡‡
Dyskinesia SMQ	215	42	1.2^‡	0.8^§	4	0	0.4^¶	0 ^‡‡
Parkinson-like events SMQ	608	104	2.0^‡	1.9^§	11	3	1.2^¶	1.2 ^‡‡
Total EPS-related events	2222	392	7.3^‡	7.1^§	43	13	4.6^¶	5.1^‡‡
Urinary dysfunction	56	16	0.18^‡	0.29^§	3	3	0.32^††	1.17^††

^†The high-dose group events are the total events from postmarketing cases (spontaneous and solicited [non-interventional] cases) in which the case reported 1or more events coded as either accidental overdose, extra dose administered, intentional overdose, overdose, or prescribed overdose, and AOM 400 was identified as a suspect product. These reported overdose preferred terms could be due to aripiprazole, AOM 400, and/or due to other co-suspect medications. ^‡Total number of adverse events for AOM 400 single-injection group was 30,552 (as per 24 June 2019). ^§Total number of serious adverse events for AOM single-injection group was 5523 (as per 24 June 2019). ^¶Total number of adverse events for AOM high-dose group was 926 (as per 24 June 2019). ^††Value higher for the AOM high-dose group vs the AOM single-injection group. ^‡‡Total number of serious adverse events for AOM high-dose group was 257 (as per 24 June 2019). Abbreviations. AOM 400, aripiprazole once-monthly 400 mg; EPS, extrapyramidal symptoms; SMQ, standardized MedDRA query.

Overall, no new safety signals were observed with higher aripiprazole doses in the postmarketing database. Based on these findings, the two-injection start regimen appears to offer no incremental safety risk to the intended patient population beyond that which would be expected as a consequence of following the single-injection start regimen.

Discussion

PopPK modeling and simulation is an approach that has been increasingly encouraged by regulatory authorities in recent years as a means to streamline the drug development process²³. The current investigation used popPK simulations to assess a simplified two-injection start regimen designed to address some of the limitations of the single-injection start regimen. The alternative regimen consists of two AOM 400 injections at separate gluteal and/or deltoid injection sites, together with a single 20-mg dose of oral aripiprazole on the 1st day of AOM 400 treatment.

The simulations performed in this investigation indicate that the two-injection plus single oral dose start would: (1) achieve therapeutic aripiprazole plasma concentrations on the 1st day of treatment; (2) support consistent clinical effectiveness over the entire dosing interval; (3) result in comparable aripiprazole plasma concentrations, and thus a comparable safety and tolerability profile, vs the single-injection start regimen (for which safety and tolerability are well established); and (4) provide an additional initiation option that obviates the need for 14-day oral tablet supplementation, potentially reducing adherence-related undertreatment during the initiation phase of treatment. Further, the simulations indicate that these benefits would pertain regardless of whether the subject had previously been stabilized on oral aripiprazole, and across all missed-dose scenarios assessed (missed 2nd, 3rd, 4th, or steady-state dose of AOM 400).

Simulations for extensive vs poor CYP2D6 metabolizers showed, as expected, higher aripiprazole C_max and exposure for the PM group with the two-injection AOM 400 start; thus, the recommendation is to start the two-injection regimen with two 300-mg administrations, along with a single dose of oral aripiprazole 20 mg in known CYP2D6 PM patients to ensure concentrations are comparable to the single-injection start regimen and remain within, or not far above the therapeutic window.

The findings from this popPK analysis—namely, that the two-injection start regimen provides comparable efficacy and no incremental safety risk compared with the single-injection start regimen—would be expected to be applicable to all approved indications for AOM.

With respect to safety and tolerability, analysis of patients receiving higher-than-recommended doses of aripiprazole in the postmarketing setting was not associated with any safety concerns beyond what would be anticipated as a consequence of following the single-injection start regimen.

The present investigation is subject to the same inherent limitations as any analysis based on popPK modeling compared with clinical data: i.e. for simulated data to be exchangeable with actual patient data, it is imperative not only that model parameters be unbiased, but that estimates of variability are also accurate²⁰. It should, however, be noted that the simulations were based on a previously validated model (unpublished data [submitted for publication]) that was adapted for the specific focus of the present investigation. In general, we believe that the development of such models is beneficial to drug development, as well as to patient care, as they can be performed in a more time-efficient and cost-effective manner, and offer the potential to spare patients from unneeded trial participation or burden. The extensive clinical experience with aripiprazole, which suggests that no additional safety signals would be anticipated at the doses assessed in the present study, also provides reassurance with respect to potential safety concerns; however, the postmarketing safety analysis included in our study is also subject to limitations. Postmarketing cases may not provide complete case information, and follow-up information cannot be obtained without the consent of the reporter. In addition, the accuracy of event frequencies is dependent on sample size. Reporting rates may be more accurate, or have more specificity to identify differences in event reporting between the single-injection regimen vs. higher-than-recommended doses; however, due to limitations of data collection, it is not possible to calculate the patient exposure for the higher-than-recommended dose group. Finally, with respect to our popPK analysis, which focused on the first dose of AOM administered (i.e. treatment initiation), modeling and simulation of metabolite concentrations (e.g. the major metabolite dehydro-aripiprazole) was not considered to be essential. In our modeling and simulations that incorporated CYP2D6 metabolism, because the number of ultrarapid metabolizers was very low (<1% of the overall population), we chose to compare 2 categories: 1) extensive/normal metabolizers (a category that included data for ultrarapid and intermediate metabolizers) and 2) poor metabolizers. Had patient numbers in the ultrarapid metabolizer group been sufficient to permit a meaningful analysis, an assessment based on all of the potential metabolic categories might have yielded additional insights into the PK profiles for these various subpopulations.

Conclusions

The two-injection start regimen for AOM 400 with a single oral dose of aripiprazole displays a comparable PK profile to the single-injection start regimen with 14 days of concurrent oral aripiprazole in simulations and may offer the potential for a number of advantages, including reduced hospital stays; reduced burden on patients, providers, and caregivers; reduced relapse risk; and expanded treatment options. The PK modeling data, as well the postmarketing safety analysis, support use of the two-injection start regimen with a single oral aripiprazole dose in clinical practice to optimize the therapeutic benefits of AOM 400 in patients with schizophrenia.

Transparency

Declaration of funding

This investigation was sponsored by Otsuka Pharmaceutical Development & Commercialization, Inc. and H. Lundbeck A/S. The resultant publication was funded by Otsuka Pharmaceutical Development & Commercialization, Inc.

Declaration of financial/other relationships

YW, XW, MH, JM, AF, NM, IR-W, and AR are employees of Otsuka Pharmaceutical Development & Commercialization, Inc. MP is an employee of Otsuka Pharmaceutical Europe Ltd. FL, MY, and LA are employees of H. Lundbeck A/S.

Author contributions

YW, XW, MH, FL, and AR made substantial contributions to the conception or design of the work. MP, MY, JM, LA, AF, NM, and IR-W were involved in the acquisition, analysis, or interpretation of data for the work. All authors contributed to drafting the work or revising it critically for important intellectual content; approved the final version for publication; and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Acknowledgements

Editorial support for this publication was provided by BioScience Communications, New York, NY, USA (funded by Otsuka Pharmaceutical Development & Commercialization, Inc.) These findings were presented, in part, at the 33rd ECNP (European College of Neuropsychopharmacology) Congress, 12-15 September 2020, Virtual.

Data availability statement

Otsuka and Lundbeck are committed to sharing data in accordance with the EFPIA/PhRMA principles for responsible sharing of clinical trial data guidelines and as required by applicable legislation. Legitimate research requests will be considered. Research proposals requesting patient-level data are reviewed by an Independent Review Panel at WIRB Copernicus Group (https://drc.irbnet.org/release/images/WCG-DRC-Bio.pdf). For inquiries on availability of data of interest, researchers should contact Otsuka ([email protected]). Please visit https://clinical-trials.otsuka.com/For-Researchers.aspx for further details. Data may be made available following review of a research proposal by the Independent Review Panel at WIRB Copernicus Group for researchers who meet the criteria for access to confidential data.

Correction Statement

This article was originally published with errors, which have now been corrected in the online version. Please see Correction (http://dx.doi.org/10.1080/03007995.2022.2034315)