FormalPara Key Summary Points

RISPERDAL CONSTA (Consta) has significantly reduced hospitalization time and relapse rate by improving patient adherence to the treatment compared with oral medications. However, the 3-week lag in drug release requires oral risperidone supplementation for 3 weeks. Rykindo was developed to improve drug release for achieving an efficacious plasma concentration faster than Consta

Due to its different PK profiles resulting from the different drug release profile, Rykindo is not directly interchangeable with Consta. Instead of conducting a safety and efficacy clinical trial, a population PK analysis was performed to establish a strategy for the switch from Consta to Rykindo and to introduce Rykindo to patients on oral risperidone treatment

With the models developed on a single ascending dose study, two relative bioavailability studies (single injection and multiple injection), and a population PK model for oral Risperdal by Vermeulen et al., simulations provided core information for New Drug Application approval of Rykindo by the Food and Drug Administration via 505(b)(2) pathway

This established approach provides guidance to physicians to initiate Rykindo therapy in adult patients with schizophrenia. For switching from Consta to Rykindo, the first Rykindo injection is administered with 2 –3 additional weeks following the original Consta dosing time. For patients taking Risperdal, introducing Rykindo with a 1-week supplemental oral dose can achieve comparable or superior exposure to that of Consta with 3 weeks of oral supplements

Introduction

Schizophrenia is an overwhelming mental illness with a global prevalence of almost 1% [1, 2]. It is basically treated with long-term, consistent antipsychotic medication [3, 4]. Noncompliance with oral antipsychotics causes disease recurrence and may even lead to treatment resistance [5,6,7]. Risperdal® (Risperdal), an oral formulation of atypical antipsychotic drug risperidone, because of its fewer side effects is the the preferred first-line treatment for acute and chronic schizophrenia [8]. Risperidone long-acting injection microsphere (LAI) RISPERDAL® CONSTA® (Consta), the first LAI developed by Alkermes and Janssen, was approved by Food and Drug Administration (FDA) in 2003 and is administered through intramuscular (IM) route at doses of 25–50 mg every 2 weeks (Q2W). Consta has significantly improved patient adherence to the treatment compared with oral medications, which has resulted in reduced hospitalization time and relapse rates [9,10,11].

However, Consta is associated with a 3-week lag before most of the drug is released, thereby necessitating oral risperidone supplementation for the first 3 weeks to ensure adequate therapeutic plasma concentrations [12]. RYKINDO® (Rykindo, code name LY03004), another LAI of risperidone developed by Shandong Luye Pharmaceutical Co., Ltd., was approved for listing in the USA in January 2023. Compared to Consta, Rykindo is not associated with the 3-week delayed drug release. As such, efficacious plasma concentrations can be attained more rapidly than with Consta, as evident from Fig. 1.

Fig. 1
figure 1

Plasma concentration-time curves of active moiety after intramuscularly (IM) injected Rykindo and Consta in patients with schizophrenia. a Following a single IM injection of Rykindo and Consta at 25 mg and 50 mg; b following multiple-dose injection of either Rykindo or Consta at 25 mg every 2 weeks for five injections

Full size image

Further safety and efficacy studies might be required to prove the clinical equivalence of Rykindo and Consta because of the different drug release profile of Rykindo. In addition, a methodology for guiding the switch from Consta to Rykindo needs to be established given the different drug release profiles of Consta and Rykindo. The objectives of this population pharmacokinetic (PK) modeling and simulation are (1) demonstrating the comparable or superior exposure in active moiety of Rykindo for supporting equivalent clinical outcomes when switching from Consta to Rykindo, as well as initiation of Rykindo in patients on oral Risperdal treatment or Risperdal naïve patients; (2) to establish a dosing window for Rykindo Q2W dosing schedule.

Methods

Clinical Study Design and Data for Model Development

The population PK models for Rykindo and Consta were based on the data obtained from three phase I clinical trials (CT-1S01, CT-USA-104, and CT-USA-102) which involved 171 patients with schizophrenia and/or schizoaffective disorder. The study was performed in accordance with the Helsinki Declaration of 1964. Each study protocol was reviewed and approved by independent institutional review boards/ethics committees (Quorum Review IRB for CT-1S01 trial; Compass IRB, LLC for CT-USA-104 and CT-USA-102 trials). Written informed consent was obtained from all subjects prior to the initiation of study procedures. Brief information about the trials is provided below.

CT-1S01 (Protocol: LY03004/CT-1S01) was an open-label, single ascending dose trial to evaluate the PK and safety of Rykindo in stable patients with schizophrenia or schizoaffective disorder following IM injection. The tested doses were 12.5 mg, 25 mg, 37.5 mg, and 50 mg. Eight subjects were included in each dose level. CT-US-104 (Protocol: LY03004/CT-US-104) was a randomized, open-label, parallel study to assess the relative bioavailability of Rykindo in relation to Consta in stable patients with schizophrenia and/or schizoaffective disorder at two single-dose levels, 25 mg and 50 mg. Each group enrolled eight subjects, except in the group receiving 25 mg of Consta, which had seven subjects. CT-USA-102 (protocol: LY03004/CT-USA-102) was a randomized, open-label, parallel study to assess the relative bioavailability of Rykindo with respect to Consta in stable patients with schizophrenia and/or schizoaffective disorder at 25 mg Q2W (total 5 injections). A total of 108 patients were enrolled in the study, with 54 subjects in each treatment group. Rich blood samples were collected up to day 43 (Study CT-1S01), day 57 (Study CT-US-104), or day 113 (Study CT-USA-102) post-dose administration. Detailed information about the patient number, tested doses, and PK sampling times are provided in Supplementary Materials (Table S1).

Plasma concentration determination of the parent compound (risperidone) and its active metabolite (9-OH-risperidone) were conducted by Covance Pharmaceutical R&D (Shanghai Co., Ltd.) using a validated liquid chromatography-tandem mass spectrometry method (LC–MS/MS). The lower limit of quantification was 0.05 ng/ml for both compounds. The calibration curves for both compounds exhibited linearity over the concentration range of 0.05–50 ng/ml. The intra- and inter-day precisions [expressed as coefficient of variation (CV%)] were < 12.1% for risperidone and 6.4% for 9-OH-risperidone, and the accuracy (expressed as % relative error) was within ± 13.3% for risperidone and ± 7.2% for 9-OH-risperidone.

Population PK Model Development

The population PK models for both Rykindo and Consta were developed using the plasma concentrations of risperidone and 9-OH-risperidone from the studies mentioned above. First-order conditional estimation method with interaction (FOCE INTER in NONMEM, version 7.4, Icon Development Solutions, Ellicott City, MD) was used for the model development. R, version 3.5.2, was used for the data summary, plotting, and model diagnostics. A model-based simulation was also conducted using NONMEM.

Drug release of long-acting IM injection agents such as paliperidone palmitate or RBP-7000 (a new sustained-release formulation of risperidone) has been previously investigated and characterized in population PK models [13, 14]. Therefore, a one-compartment model with a series of multiple drug inputs, zero order, and/or first order was explored as a starting point.

A covariate analysis was conducted to explore the sources of variability. Subject demographics (age, gender, race, weight, body mass index, etc.) were the covariates tested. Risperidone is metabolized to 9-OH-risperidone via CYP2D6, with both having equivalent pharmacological activities [15]. Thus, the sum of risperidone and 9-OH-risperidone constitutes the active moiety in pharmacological response to risperidone. When the active moiety was modeled, the CYP2D6 genotype information [poor metabolizer, intermediate metabolizer, extensive metabolizer (EM), and ultra-metabolizer] was no longer relevant [16, 17]. The single-dose PK data were used for establishing the covariate model, which was finalized with data from all studies. Between-subject variability for a PK parameter was assumed to follow a log-normal distribution, and both proportional and additive error combinations were used to describe intra-individual variability.

The standard diagnostic tools such as goodness-of-fit criteria, diagnostic plots, and visual predictive check (VPC) were employed for model evaluation. Bootstrapping was conducted for evaluating the stability of the final model as well as for estimating the standard error (SE) for the model parameters when covariance steps failed.

Model-Based Simulations

Population PK models for Rykindo and Consta developed in the current study were applied for simulating plasma concentrations of the active moiety following the IM injection of either Rykindo or Consta. The population PK model for oral Risperdal developed by Vermeulen et al. [18] was used to simulate plasma concentrations of risperidone and 9-OH-risperidone following oral dosing. The active moiety with oral dosing was derived from the summation of risperidone and 9-OH-risperidone. For simplicity, only EM was considered for the simulation of oral dose, as different CYP2D6 status did not significantly impact the concentration of the active moiety.

PK parameters of the active moiety such as area under the concentration-time curve within the dosing interval (AUCτ), average concentration within one dosing interval (Cave), maximum observed concentration (Cmax), and minimum observed concentration (Cmin) (at steady state or prior to steady state after the switch) were derived using simulated concentration profiles. The log-transformed values were used for the calculation of AUCτ, Cmax, and Cmin (at steady state and prior to steady state) point estimates (LS means) for the test ratios of test (under different switching scenarios) to reference (Consta), as well as the 90% confidence interval (CI).

Simulation of Steady-state Exposures Following Injections of Rykindo or Consta

Concentration profiles of the active moiety following either Rykindo or Consta injections at 25 mg Q2W were simulated for 1000 virtual subjects in each arm. Consta arm was the reference arm. Steady-state PK parameters of the active moiety, AUCτss, Cavess, Cmaxss, Cminss, and degree of fluctuation (PTF, 100 × (Cmaxss − Cminss)/Cavess) were calculated. Steady-state exposures of Rykindo and Consta were compared.

Simulation of the Switch from Consta to Rykindo

Different dosing times for the first injection of Rykindo when switching from Consta to Rykindo were explored. For example, if the last Consta dose was administered on day 84 (seven consecutive doses, Q2W), several Rykindo dosing scenarios were considered, such as the first Rykindo injection administered on day 98 (same as the scheduled Consta dosing time if treatment had been continued with Consta), on day 112 (2 weeks following), or on day 119 (3 weeks following original scheduled Consta dosing time). The exposures of the active moiety during the transition period (before reaching a steady state after switching to Rykindo) were compared with the reference arm, in which the patients remained on Consta.

Introducing Rykindo to Patients on Oral Risperdal

The guideline for introducing Consta to patients on oral Risperdal indicates continuation of oral Risperdal for 3 weeks following the first injection of Consta (treated as reference arm here). Regarding introducing Rykindo to patients on oral Risperdal, continuation of oral Risperdal for a few additional days or weeks (different scenarios) following first injection of Rykindo was explored. The exposures of the active moiety of these different scenarios were compared with those of the reference arm.

Before commencement of treatment with Consta, risperidone-naïve patients need to be initiated on oral Risperdal to establish their drug tolerability. Introducing Rykindo in naïve patients follows the same approach, similar to that for patients on oral Risperdal.

Dosing Window

To explore the Rykindo administration window, two different scenarios were simulated: (1) If a dose was administered ± x days of the scheduled dosing time, the subsequent doses would be administered at the original dosing times. (2) If a dose was administered ± x days of the scheduled dosing time, the subsequent doses would be administered at Q2W with a new start date of the schedule. The exposure of the active moiety prior to the achievement of a new steady state was compared with the steady state exposure.

Results

PK of Rykindo from the Phase I Clinical Trials

PK profiles of Rykindo/LY03004 at single doses of 12.5 mg, 25 mg, 37.5 mg, and 50 mg (Study CT-1S01) demonstrated an immediate increase in plasma concentrations of the active moiety. Plasma concentrations of the active moiety continued to increase to reach Cmax between 14 and 17 days across all dose levels. Dose proportional increase in exposure (AUC0-t and Cmax) was observed over the dose range of 12.5–50 mg (data under submission).

In Study CT-USA-104 aimed at comparing the PK of Rykindo with Consta following a single IM injection, Cmax was achieved between days 14 and 17 for Rykindo and days 32 and 34 (Median) for Consta at 25 mg and 50 mg. The comparison of the PK profiles of Rykindo and Consta also indicated that, with Rykindo, risperidone was released without a 3-week delay (Fig. 1a). The geometric means of Cmax (18.5 and 30 ng/ml for Rykindo at 25 mg and 50 mg vs. 17.6 and 36.0 ng/ml for Consta at 25 mg and 50 mg) and the geometric means of AUC0-t (227 and 399 day*ng/ml for Rykindo vs. 193 and 490 day*ng/ml for Consta at 25 mg and 50 mg) of the active moieties from Rykindo and Consta were found to be comparable.

Following multiple-dose injections of either Rykindo or Consta 25 mg Q2W (Study CT-USA-102), the trough concentrations for Rykindo and Consta on day 14 (prior to the second injection) were 77.0% and 11.2% of the steady-state trough levels (post fifth dose), respectively. By day 28 prior to the third injection, the trough concentration of Rykindo had reached 94% of its steady-state trough level, while the trough concentration of Consta was at 91% of its steady-state trough level by day 42, suggesting that Rykindo achieved steady state ~ 2 weeks earlier than Consta. (Fig. 1b). PK parameters at steady state for active moiety (the last dosing interval after the fifth injection) were derived using non-compartmental analysis. The point estimate ratio (Rykindo to Consta) of Cmaxss and AUCτss of active moiety in steady state was 102.3% and 94.08%, respectively.

Population Pharmacokinetic Modeling of Rykindo

A total of 2216 PK concentration records for risperidone and 9-OH-risperidone following the administration of Rykindo were available from 97 subjects in three trials. Demographic characteristics of the subjects and CYP2D6 metabolizer status are summarized in Table 1.

Table 1 Demographics and CYP2D6 status of subjects used in the model development
Full size table

Different input functions such as first-order or zero-order, parallel or sequential release were used to explore risperidone release from the IM depot site. The PK profile of risperidone has been optimally represented by the drug release depicted in the structure model (Fig. 2a). Risperidone and 9-OH-risperidone were noted to share similar patterns in their PK profiles, indicating that the structure model for risperidone was applicable to the sum of risperidone and 9-OH-risperidone. The population PK model development and subsequent model-based simulation were changed for the active moiety (Fig. 2b), since the active moiety is responsible for clinical efficacy [15]. Initial results have shown that the estimated value of elimination rate constant (K) (0.379 1/d) approximately equaled the value of absorption rate constant (KA) (0.322 1/d), suggesting “flip–flop” PK. Therefore, in the subsequent analysis, K was set to KA to reduce the unknown number of parameters.

Fig. 2
figure 2

Pharmacokinetic structure model of Rykindo/Consta. a Risperidone; b active moiety. D2 0-order release duration; F2 fraction of the immediate release; ALAG1 lag time of the main release; F1 fraction of the main release; KA first-order absorption rate constant; ALAG3 lag time of the middle release; D3 0-order release duration; CLp clearance of the parent compound; Km Risperidone metabolism rate constant to 9-OH-risperidone; CLm clearance of the metabolite; CL clearance of the active moiety

Full size image

A one-compartment model with an immediate 0-order release followed by two delayed releases (an intermediate 0-order and a 1st-order main release) adequately described the PK characteristics of the active moiety following a single as well as multiple Rykindo administrations. Diagnostic plots of the final model are presented in the Supplementary Materials (Fig. S1). Figure 3 demonstrates the VPC conducted with 100 replicates on the original observed data sets. Derived model parameters are given in Table 2. Since the covariance step failed during optimization, the SE of the parameters in Table 2 was derived from 200 bootstrap replicates.

Fig. 3
figure 3

Visual predictive check of the final model for the active moiety of Rykindo/LY03004 (a single dose at 25 mg; b multiple dose at 25 mg every 2 weeks). Open circle: observed data; solid lines represent 2.5th, 50th, and 97.5th percentiles of observed concentrations at the nominal time points; dashed lines represent median of 2.5th, 50th, and 97.5th percentiles of predicted concentration; shaded area represents the 95% prediction interval of 2.5th, 50th, and 97.5th percentiles of predicted concentrations

Full size image
Table 2 Parameters of the final PK model of Rykindo/Consta active moiety
Full size table

Following an injection of Rykindo, 14.8% of the drug was immediately released with absorption duration of 0.764 days, followed by a slow intermediate release that occurred 3.47 days post-injection with an absorption duration of 2.18 days. The absorption rate constant from the release site to the blood circulation was 0.118/day. The intermediate stage release accounted for 42.1% of the total absorption. The final release occurred 13.3 days post-injection accounted for 43.1% of the total absorption. Cmax was observed approximately 2.5 weeks post administration of a single dose. The apparent clearance of the active moiety was 183 l/d for males and 156 l/d for females in the single dose studies and 124 l/d for males and 105 l/d for females in the multiple-dose study. Females demonstrated about 15% lower clearance compared with males, suggesting that the gender effect on PK was unlikely clinically relevant. Difference of the first-order absorption rate constant was observed in different clinical studies.

Population PK model of Consta active moiety

A total of 1766 PK concentration records for risperidone and 9-OH-risperidone were available from two trials involving treatment of Consta, with 66 subjects. Detailed information of the subjects has been given in Table 1. The same structural PK model adequately described the PK profiles of the active moieties following the injections of Consta. Diagnostic plots of the final model are included in the Supplemental Material (Fig. S2). VPC was conducted with 100 replicates on the original observed data sets (Fig. 4). The estimated PK parameters of the active moiety of Consta are provided in Table 2. The SE of the parameters in Table 2 were derived from 200 replicates of bootstrap.

Fig. 4
figure 4

Visual predictive check of the final model for the active moiety of Consta (a single dose at 25 mg; b multiple dose at 25 mg every 2 weeks). Open circle: observed data; solid lines represent 2.5th, 50th, and 97.5th percentiles of observed concentrations at the nominal time points; dashed lines represent median of 2.5th, 50th, and 97.5th percentiles of predicted concentration; shaded area represents the 95% prediction interval of 2.5th, 50th, and 97.5th percentiles of predicted concentrations

Full size image

Before the main release 27.3 days post injection, the active moiety was maintained at low and near constant levels by the first two releases (11.9% from immediate release and 12.6% from an intermediate release in less than half a day post injection). Around 75.5% of the drug was from the main release. From the single-dose PK profiles, this main release contributed to the Cmax that was observed approximately 4.5 weeks post injection, which was 2 weeks later than that of Rykindo. The apparent clearance of active moiety was 105.8 l/d. No gender effect on clearance was statistically significant. The only covariate that is statistically significant is found in the studies on the primary absorption rate constant.

Simulated PK Profiles of Active Moiety for Rykindo and Consta

As the PK characteristics of both Consta and Rykindo were linear, simulations for both Consta and Rykindo were conducted only at single dose levels of 25 mg Q2W. PK profiles for Rykindo and Consta were simulated for 1000 virtual subjects of each treatment following seven injections. The PK profiles of the active moiety of Rykindo and Consta were comparable except for the time to reach the first Cmax (28 days for Rykindo vs. 49 days for Consta).

Steady-state PK parameters, such as AUCτss, Cmaxss, and Cminss, as well as the PTF, following Consta or Rykindo injections were derived using simulated PK profiles. The point estimate ratio (Rykindo to Consta) of Cmaxss and AUCτss of active moiety in steady state was 93.02% and 94.43%, respectively. Cminss of Rykindo was greater than that of Consta, indicating that the efficacy and safety would not be compromised with Rykindo treatment (Table S2).

Simulation for Switching from Consta to Rykindo

Since Cmax for Rykindo was achieved approximately 2 weeks prior to that of Consta, switching from Consta to Rykindo necessitated a delayed administration of the first Rykindo injection unlike the original Consta dosing schedule. Simulated PK profiles (median, 5th, and 95th percentiles) of the active moiety are presented in Fig. 5. The reference arm consisted of patients continuing with the Consta treatment. For the scenario involving an additional 2 weeks added to the original Consta schedule for the administration of the first Rykindo injection, the bioequivalence (BE) test was conducted on AUCτ, Cmax, and Cmin of the transition period (from day 98 to 154) along with the reference arm for the same period, which represents the steady-state exposure of Consta. Both AUCτ and Cmax were within the BE range; Cmin during the transition period was either within the BE range (from day 98–112) or greater (day 112–154, 1.3–1.6 folds) than that of Consta at steady state, indicating that the efficacy would not be compromised if the first Rykindo injection was administered 4 weeks after the last injection of Consta, as shown in Fig. 5a and Table 3.

Fig. 5
figure 5

Simulated pharmacokinetic profiles of the active moiety for the switch to Rykindo (LY03004, 25 mg every 2 weeks) 4 weeks a and 5 weeks b after the last Consta (25 mg every 2 weeks) injection

Full size image
Table 3 Simulated exposure of the active moiety during the transition period for the switch to Rykindo 4 and 5 weeks after the last Consta injection at 25 mg every 2 weeks
Full size table

When three additional weeks were added to the scheduled Consta dosing time for the first Rykindo injection (Fig. 5b), AUCτ and Cmax during the transition period except from day 119 to 133 were within the BE range. Though AUCτ and Cmax were lower (with 77.03% ratio for Cmax, 79.92% ratio for AUCτ) during this period (day 119 to 133), efficacy was unlikely to be compromised, given that Cmin was either within the BE range or greater (about 1.3-fold) than that of Consta at steady state (Table 3).

When 1 or 4 additional weeks were added to the scheduled Consta dosing time, patients could be either substantially over-exposed or under-exposed to the active moiety during the transition period (Fig. S3a and b).

Introducing Rykindo to Patients on Oral Dose Risperdal

Consta requires a 3-week oral Risperdal supplemental treatment following the first injection. Since Rykindo showed faster drug release, three treatment arms were further evaluated: without oral Risperdal supplemental treatment (Arm 2), with 1-week oral Risperdal supplemental treatment (Arm 3), or 3-day oral Risperdal supplemental treatment (Arm 4). The reference arm was patients receiving Consta with 3 weeks of oral Risperdal supplemental dosing.

Figure 6a presents the PK profile (median, 5th, and 95th percentiles) for active moiety from 1000 virtual subjects for the reference arm. Figure 6b, c, and d represent the PK profiles for Arm 2, Arm 3, and Arm 4, respectively. The dashed line in Fig. 6b–d represents the median of the reference arm. Both Arm 2 and Arm 4 demonstrated lower exposure compared with that of the reference arm during the 2nd week following the injection of the first Rykindo dose. However, 1-week oral Risperdal supplemental dosing resulted in comparable or superior exposure to that of the reference arm (Fig. 6c) (Arm 3). For example, the LS ratios of Cmax, AUC, and Cmin (Rykindo to Consta) for the week after the end of the oral supplemental dosing were 127.7%, 159.5%, and 181.4%, respectively. The higher exposure of Arm 3 was considered clinically beneficial, since the exposure in the reference arm in that week was substantially low compared with its steady-state exposure. The exposure in Arm 3 on the 14th subsequent day was either equivalent to or higher than that of the reference arm. These results suggested that introducing Rykindo to patients on oral Risperdal only required 1-week oral supplemental dosing to achieve clinical outcomes that were equivalent or superior to those of Consta with 3-week oral supplemental treatment.

Fig. 6
figure 6

Simulated pharmacokinetic profiles of the active moiety for introducing Rykindo to patients on oral dose Risperdal with switching from Risperdal (2 mg/day) to Rykindo (25 mg Q2W) or Consta (25 mg Q2W). a The first injection of Consta co-administered oral Risperdal for 3 weeks; b the first injection of Rykindo without oral Risperdal; c the first injection of Rykindo co-administered oral Risperdal for 1- week; d the first injection of Rykindo with oral Risperdal for 3 days. Q2W, every 2 weeks

Full size image

Dosing Windows for Rykindo

Simulations to support the Rykindo administration window were conducted by assuming dosing windows were ± 1 day, ± 2 days, and ± 3 days. Results indicated that outside of the ± 3 days window, the change in exposure would be outside of the 80%–125% range (data not shown). The median, 5th, and 95th percentiles of simulated PK profiles of the active moiety when Rykindo dosing was given ± 3 days are shown in Fig. 7. For scenario 1 (Fig. 7a), on one occasion, a 3-day delayed dosing was given on day 73 instead of the scheduled dosing on day 70. On another occasion, a dose was administered 3 days earlier on day 137 instead of the scheduled day 140. When the dosing time was either 3 days behind or ahead of the scheduled dosing time, and resuming Rykindo following the original dosing schedule, the exposure of the active moiety compared with the steady state was within or close to 20% difference, suggesting that ± 3 days can be accepted as the safe window for Rykindo administration. For scenario 2 (Fig. 7b), when the dosing time was either 3 days behind or ahead of the scheduled dosing time, and resuming Rykindo with a new start day for the Q2W schedule, the changes in exposure due to dosing schedule interruption were less profound compared to scenario 1.

Fig. 7
figure 7

Simulation the dosing windows for Rykindo. a A dose was administered ± 3 days of the schedule dosing time, and the subsequent dose would be given at the original dosing time; b a dose was administered ± 3 days of the schedule dosing time, and the subsequent dose would still follow Q2W schedule with a new start date of the Q2W schedule. Q2W every 2 weeks

Full size image

Discussion

Treatment with Consta has significantly reduced hospitalization time and relapse rates by improving patient adherence to the treatment compared with oral medications. However, the 3-week lag during the drug release phase requires 3 weeks of oral risperidone supplementation. Rykindo was developed to modify the drug release profile to achieve faster efficacious plasma concentrations of active moiety while maintaining the Q2W dosing schedule. The overall objective of this analysis was to assess whether clinical outcomes equivalent to Consta can be achieved by treatment with Rykindo in patients with schizophrenia or schizoaffective disorder, given its different drug release profile compared to that of Consta.

Population PK modeling quantified the difference in drug release profiles following IM injection of Rykindo and Consta. Faster release of the active moiety by Rykindo resulted in Cmax being achieved 2–3 weeks earlier than Consta following a single dose injection and about 3 weeks with multiple injections. Since the drug release determines the disposition of Rykindo and Consta, Rykindo can also be cleared from the blood circulation 2–3 weeks earlier than Consta. Therefore, Rykindo possesses advantages of fast onset of action as well as fast offset pharmacological response compared with Consta.

Though different drug release profiles of Rykindo and Consta do not impact their steady-state exposure (AUCτss and Cmaxss) of the active moiety, Rykindo cannot be switched to Consta without considering its faster drug release. Clinical trials to demonstrate clinically equivalent outcomes of Rykindo and Consta, to support a safe switch from Consta to Rykindo, or to introduce Rykindo to patients on oral Risperdal, may be required. In the current analysis, population PK model simulations were applied to establish a dosing strategy that would ensure comparable or superior exposure of the active moietywhen switching from Consta to Rykindo, and when introducing Rykindo to patients on oral Risperdal. Furthermore, a dosing window for Rykindo was recommended through simulations. Results from the PK model-based simulations demonstrated that when the first Rykindo injection is administered with 2–3 additional weeks following the original Consta dosing time, the exposure of the active moiety was comparable or superior to that of patients continuing treatment with Consta. In addition, when introducing Rykindo to patients on oral Risperdal, with 1-week supplemental oral dose, the exposure of the active moiety was comparable or superior to that of Consta with a 3-week supplemental oral dose. These analyses supported the claim that Rykindo can achieve equivalent clinical outcomes to those of Consta when switching from Consta to Rykindo. The 2-week reduction in oral supplemental required with Rykindo could further improve patient compliance.

Although introducing Rykindo without a supplemental oral dose resulted in noticeably lower exposure in the 2nd week following the first injection of Rykindo, the lowest Cmin (4.36 ng/ml) observed in the 2nd week was still 25% higher than the lowest Cmin (3.49 ng/ml) when switching from oral to Consta. Pilla Reddy et al. established the exposure-response modeling of antipsychotic drugs in patients with schizophrenia using a positive and negative symptom scale for schizophrenia total score and clinical utility and found that the effective concentration of risperidone was 5.2 ng/ml (risperidone + 9-OH-risperidone) and EC50 was 3.72 ng/ml [19]. Therefore, the Cmin of 4.36 ng/ml without oral supplemental treatment for Rykindo could result in at least partial clinical response.

The population PK model development started with risperidone and its active metabolite and was changed to the active moiety. The benefit of this approach significantly simplified the model development process regarding the structure model and covariate analysis on CYP2D6 status, given the small sample size of the subset of CYP2D6 status in particular. The approach was corroborated by the work published by Riedel and colleagues [16], where the observed CYP2D6 polymorphisms did not contribute to altered clinical efficacy though they did affect the risperidone to 9-OH-risperidone ratios. Ma et al. reported that the CYP2D6 activity greatly influenced the pharmacokinetics of risperidone and the ratio of risperidone to 9-OH-risperidone, but no significant differences were found in the active moiety [17].

The multiphase drug release explained the double peaks and prolonged plasma concentration of active moiety. It was impractical to measure substantial PK sampling times to accurately capture the parameters related to drug release/absorption, since the drug release phase lasted for weeks. The estimated drug release/absorption parameters in our analysis were derived from limited PK samples. This might explain why the estimated absorption rate constant following multiple dosing was faster than that of the single dose for both Rykindo and Consta, as different PK sampling schemes were implemented in the single-dose study and the multiple dosing study. However, adequate trough samples indicating for efficacy were collected in the multiple dose study that enabled the model to accurately predict the trough concentrations for both Rykindo (Fig. 3b) and Consta (Fig. 4b). Clinical data from both the single dose studies and the multiple dose study demonstrated that Rykindo were well tolerated and exhibited a safety and tolerability profile aligned with the well-established safety profile presented in the Consta United States Prescribing Information (manuscript under submission). Given the limitation of PK modeling and simulation, when switching to Rykindo, special attention should be paid to patients’ tolerability/safety.

Conclusions

In summary, with the established dosing strategy when switching from Consta to Rykindo or introducing Rykindo to patients on oral Risperdal dosing or Risperdal naïve patients, Rykindo not only achieved equivalence in clinical outcome, but also provided additional benefits by improving compliance with reduced oral supplemental treatment.