(M-063) Population Pharmacokinetic/Pharmacodynamic Analysis of YH35995 for the First-in-Human Dose Prediction

YuKyung Kim, MD, PhD – Clinical Research Physician, Yuhan Corporation; Dong Kyun Kim, PhD – Senior Principal Scientist, Yuhan Corporation; Seong Bok Jang, NA – Clinical Pharmacology Director, Yuhan Corporation

Objectives: YH35995 is a novel selective inhibitor of the glucosylceramide synthase (GCS) with a high blood-brain barrier penetration for a treatment of neurological symptoms of Gaucher disease type 3. In this study, a population PK/PD model for YH35995 was developed using animal data to predict human PK/PD profile and enable first-in-human (FIH) dose selections.

Methods: The plasma concentration data following single and multiple doses in mice, rats, and dogs were utilized to develop the PK models. The plasma glucosylceramide (GL1) level data following single and multiple doses in the mice were used to establish a simultaneous PK/PD model where the relationship between drug concentration and effect of YH35995 showed a counter-clockwise hysteresis profile. The effect of the drug exposure was described by an indirect response model with inhibition on Kin and parameterized using an Imax function. Visual predictive checks and other diagnostic methods were performed to evaluate the predictive performance of models. A simple allometric scaling was employed to obtain the human PK parameters. Human PK/PD profiles were simulated under various dose regimens to establish an appropriate dose range considering the results of in vitro and in vivo studies. All modeling and simulations were performed using NONMEM® version 7.5.1.

Results: A two-compartment model with the first order absorption and elimination best described the animal PK data. The indirect response model with the Imax function adequately described the observed time course of the plasma GL1 inhibition in mice. In the final PK/PD model for mice, Ka, CL/F, V2/F, V3/F, and Q/F were well estimated, and inter-individual variabilities (IIVs) of CL/F and Ka were estimated to be 20.8% and 30.1%, respectively. Relative baseline of the plasma GL1 level (%, BASE) and Imax were fixed to 100% and 1, respectively, since these parameters were estimated closely to 100% and 1. IC50 was estimated to be 57.4 ng/mL and its IIV was estimated as 16.9%. Residual variabilities of PK and PD were described by proportional error models. Based on the human PK/PD simulations, the proposed starting dose is 3 mg, since simulated human Cmax (1.3 ng/mL) after single dose of 3 mg was comparable to the IC30 (1.36 ng/mL) of in vitro GCS enzyme activity. Efficacious doses for humans are suggested to be 60 to 120 mg QD compared to the expected exposures at the sufficient efficacious doses (5-10 mg/kg QD) in mice and target levels of -70% to -75% for the plasma GL1 inhibition. With the projected human half-life of 125 hr, the maximum plasma GL1 inhibition after administration of 180 mg QW was predicted to be approximately -68.3% at steady state, which nearly reached the target levels [1].

Conclusions: The population PK/PD analyses of YH35995 were successfully conducted with preclinical observations. The modeling and simulation could guide the FIH study of YH35995 by projecting various doses, PK, and GL1 inhibitions for humans.

Citations: [1] Peterschmitt et al. Pharmacokinetics, Pharmacodynamics, Safety, and Tolerability of Oral Venglustat in Healthy Volunteers. Clinical pharmacology in drug development. 2021;10(1):86-98.