(T-082) Population PK and PK/PD Modeling of Apitegromab in Spinal Muscular Atrophy Patients

Giridhar Tirucherai – Scholar Rock, Inc.; Yang Xu – Scholar Rock, Inc.; Jing Marantz – Scholar Rock, Inc.; Nathalie Gosselin – Certara, Certara Drug Development Solutions; Kathleen Koeck – Certara, Certara Drug Development Solutions; Amira Ghoneim – Certara, Certara Drug Development Solutions; Samer Mouksassi – Certara, Certara Drug Development Solutions

Objectives: Apitegromab (SRK-015), a fully human immunoglobulin G4 monoclonal antibody, is being developed as the first potential muscle targeted treatment for pediatric and adult patients 2 years of age and older with spinal muscular atrophy (SMA). Population pharmacokinetic (PK) and PK-total latent myostatin (pharmacodynamic [PD]) analyses used data from patients with SMA (2-21 years) and healthy individuals receiving intravenous apitegromab. Dosing regimens included single doses (apitegromab 1 mg/kg to 30 mg/kg) or repeat doses (apitegromab 10 mg/kg to 30 mg/kg every 2 weeks [Q2W]; apitegromab 2 mg/kg to 20 mg/kg Q4W) in studies SRK-015-001, TOPAZ, and SAPPHIRE.

Methods: A non-linear mixed-effects (NLME) approach was used to develop the population PK and PK-PD models. The dataset included 3792 evaluable PK observations from 236 healthy adults and patients with SMA who were treated with single or multiple doses of apitegromab and 3465 evaluable PD observations from 246 patients with SMA receiving either apitegromab or placebo. Covariate selection was performed by univariate forward addition (P < 0.01) and backward elimination (P < 0.001).

Results: The final population PK model was a 2-compartment model with linear elimination, incorporating a maturation function on clearance (CL), time-varying weight effect on CL and volume of distribution, and time-varying age effect on central volume of distribution (Vc). Weight effects on clearance and volume parameters were parameterized as power functions with estimated exponents of 0.583 and 0.628, respectively. Description of clearance as a function of maturation process with allometric scaling, better predicted exposures of apitegromab in pediatric patient populations than allometric scaling alone. The estimated fraction of maturation of CL of apitegromab in a typical full-term neonate at birth (β) was 0.173 and maturation half-life for CL was 90 months. Although a maturation function was applied to CL, similar exposures were observed with weight-based dosing across the age range of patients with SMA (2-21 years). Age was a significant covariate on Vc with an estimated exponent of 0.24.

The final population PK/PD was an indirect response model with drug effect on the output (i.e., Kout; rate constant of loss of total latent myostatin) parametrized with maximum inhibition (Imax) and concentration to achieve 50% of Imax (IC50). Baseline total latent myostatin and ambulatory status were identified as covariates on Imax, though changes in Imax due to these covariates were minimal ( < 0.2%).

Goodness-of-fit and prediction-corrected visual predictive checks confirm model adequacy to describe observed data and reliably derive patient-level exposure (PK and PD) parameters.

Conclusions: Following weight-based dosing of apitegromab, covariates such as weight, age, age group, concomitant survival motor neuron-targeted therapy, race, ambulatory status, antidrug antibody status, and sex did not result in clinically meaningful differences in PK or PD levels, indicating no need for dose adjustments based on these variables. The developed PK and PK-PD models provide support for model-informed dose selection of apitegromab for neonates and infants with SMA aged < 2 years.

Citations: N/A

Keywords: Pediatrics, PK-PD model, Maturation function