Population pharmacokinetic modeling of total active moiety exposure following doses of TV 46000, a long-acting subcutaneous risperidone injection
Itay Perlstein (1), Avia Merenlender Wagner (2), Anthe Zandvliet (3), Farina Hellmann (3), Eline van Maanen (3), Nele Plock (3), Floris Fauchet (3), Yu-Wei Lin (3), Anna Elgart (2), Rajendra Singh (4)
(1) Magic Wand Research LLC, Philadelphia, PA USA; (2) Teva Pharmaceutical Industries Ltd Netanya, Israel; (3) Certara, Princeton, NJ, USA; (4) Teva Branded Pharmaceuticals West Chester, PA, USA
Background: TV‑46000 is a long-acting subcutaneous (sc) risperidone injection in development for treatment of schizophrenia with flexible dosing either once monthly (q1m) or once every two months (q2m). Population pharmacokinetic (PopPK) modeling was used to characterize the pharmacokinetics (PK) of the parent drug, risperidone and active metabolite, 9-OH risperidone. Additionally, the impact of potential covariates was evaluated on the total active moiety (risperidone + 9-OH risperidone, TAM) PK following TV-46000 administration in patients with schizophrenia.
Methods: A sequential parent-metabolite PopPK model was developed using nonlinear mixed effect modeling. To describe the complex absorption from the sc depot two and three-fraction absorption models with first-order absorption with or without transit compartments were considered. Covariate selection was performed using stepwise covariate modelling (SCM) with significance levels of 0.01 and 0.001 (likelihood ratio test) for forward addition and backward elimination, respectively. Potential effects of intrinsic and extrinsic factors were evaluated on apparent clearance (CL) and rate constants for fast (KA1) and slow (KA2) absorption. The covariates tested included demographics, concomitant medications, and disease status. In addition, impact of injection site, rubbing post injection, product presentation (vial or prefilled syringe presentation) and injection volume were evaluated. Published model was used to simulate oral risperidone and long acting injectable (LAI) administered intramuscular (IM) of risperidone every 2 weeks.
Results: PK of the parent compound, risperidone, was adequately described by a 1-compartment model, with double first-order absorption route (with a fast absorption process and a slow process, including transit-compartments) to describe the absorption from the sc depot and first-order elimination. KA1 decreased with increased injection volume (more pronounced with subtherapeutic doses, 12.5mg and 25mg). The effect of body mass index (BMI) shifted between the 2 absorption processes: as lower BMI had higher KA1 values, indicating fast absorption, whereas the slow absorption route process (KA2) increased with higher BMI. Administration to the arm was associated with a higher KA1 than injection in the abdomen. Although injection site was a significant covariate in the parent model, no impact was observed on the overall steady-state TV-46000 exposure, as CL was similar for both sites. PK of 9-OH risperidone was adequately described by a 1-compartment model with first-order input from the risperidone compartment and first-order elimination. Simulated TV‑46000 provided comparable exposure to simulated oral risperidone exposure of 2 to 5 mg daily doses. Additional simulations of TV-46000 to other risperidone LAI formulations was conducted: once monthly (q1m) doses of TV-46000 appeared to have comparable exposure to risperidone administered as IM injections every 2 weeks. There were no statistically significant or clinically relevant covariate effects on the PK of 9-OH risperidone. Overall, TAM exposure (AUCss) was not affected to a clinically meaningful extend by covariates (injection volume, BMI, and injection site).
Conclusions: The sequential parent-metabolite model with double first-order absorption route for the parent adequately described the PK of risperidone and 9-OH risperidone. While injection volume, BMI and injection site were identified as statistically significant covariates affecting the double first-order absorption route of risperidone, overall TAM exposure (AUCss) was not affected by the identified covariates and there were no clinically relevant effects.