Objectives: Ponsegromab is a first-in-class, humanized anti-GDF15 monoclonal antibody which is being developed to treat cachexia in patients with solid tumors. The current study is to quantitatively characterize the relationships between ponsegromab PK, circulating GDF-15 levels (PD) and body weight (clinical response) over time in patients with cancer cachexia following ponsegromab subcutaneous injection up to 12 weeks.
Methods: Ponsegromab and GDF-15 concentration data from 5 phase 1/2 clinical studies in healthy volunteers or patients with cancer and cachexia or anorexia and elevated GDF-15 levels. and body weight data over 12 weeks collected in the phase 2 study in participants with cancer cachexia were included in the analysis. Model development used NONMEM Version 7.5.0 (ICON Dev. Soln, Ellicott City, MD, USA). Population parameter estimations used first-order conditional estimation method with interaction (FOCEI) algorithms and individual parameters were obtained from empirical Bayes estimates (EBE). Perl-speaks-NONMEM (PsN) Version 5.3.0 was used for performing sampling importance resampling (SIR). Exploratory analyses, diagnostic plots, and post-processing of NONMEM output were performed using R 4.2.1 with add-on packages (R Foundation for Statistical Computing, Vienna, Austria).
Results: To better understand the PK and GDF-15 effects on body weight, a joint population PK/PD/response model that simultaneously incorporated unbound and total ponsegromab PK concentrations, total GDF-15 concentrations, and body weight data up to 12 weeks was developed in this study. Unbound GDF-15 data were not used in the model estimation due to a high proportion of below limit of quantitation (BLQ) data however they were used in model validation. The quasi-steady-state (Qss)-TMDD approximation was implemented for ponsegromab PK and GDF-15 modeling. The response of body weight change over time was characterized using indirect response (IDR) model where the rate constant of body weight gain (Kin) was stimulated by unbound ponsegromab concentrations. Highly influential covariates including body weight on PK and population/patient cancer type on baseline GDF-15 levels were built in base model. Key diagnostic plots suggested there was no significant misspecification or bias with final model. Standard visual predictive check (VPC) showed that predicted values overly the observed data used for model development with good agreement. The model also adequately captured the unbound GDF-15 suppression following ponsegromab treatment over 12 weeks.
Conclusions: The final joint TMDD-IDR model adequately described the longitudinal ponsegromab PK/GDF-15/body weight data simultaneously and was considered appropriate for simulations to inform Phase 3 study design and dose selection.
