Model-Informed Drug Development of the Masked Anti-PD-L1 Antibody CX-072
Mark Stroh1, Michelle Green2, Bjorn L Millard3
1CytomX Therapeutics, Inc, South San Francisco, California, USA.
Abstract:
CX-072 is an anti-PD-L1 (programmed death ligand 1) Probody therapeutic (Pb-Tx) designed to be preferentially activated by proteases in the tumor microenvironment and not in healthy tissue. Here, we report the model-informed drug development of CX-072. A quantitative systems pharmacology (QSP) model that captured known mechanisms of Pb-Tx activation, biodistribution, elimination, and target engagement was used to inform clinical translation. The QSP model predicted that a trough level of masked CX-072 (intact CX-072) of 13-99 nM would correspond to a targeted, 95% receptor occupancy in the tumor. The QSP model predictions appeared consistent with preliminary human single-dose pharmacokinetic (PK) data following CX-072 0.03-30.0 mg/kg as monotherapy: CX-072 circulated predominantly as intact CX-072 with minimal evidence of target-mediated drug disposition. A preliminary population PK (POPPK) analysis based upon 130 subjects receiving 0.03-30.0 mg/kg as monotherapy included a provision for a putative time-dependent and dose-dependent antidrug antibody (ADA) effect on clearance (CL) with a mixture model. Preliminary POPPK estimates for intact CX-072 time-invariant CL and volume of distribution were 0.306 L/day and 4.84 L, respectively. Exposure-response analyses did not identify statistically significant relationships with best change from baseline sum of measurements and either adverse events of grade ≥ 3 or of special interest. Simulations suggested that > 95% of patients receiving CX-072 10 mg/kg every two weeks would exceed the targeted trough level regardless of ADA, and that dose adjustment by body weight was not necessary, supporting a fixed 800 mg dose for evaluation in phase II.
Insights
CX-072, an anti-PD-L1 Probody therapeutic, demonstrated predictable pharmacokinetics in early trials. Model-informed development supports a fixed 800 mg dose for Phase II, showing efficacy regardless of antidrug antibodies.
Area of Science:
- Immunology
- Pharmacology
- Biotechnology
Background:
- CX-072 is a novel anti-PD-L1 Probody therapeutic designed for tumor-specific activation.
- Tumor microenvironment proteases activate CX-072, minimizing off-target effects in healthy tissues.
Purpose of the Study:
- To report the model-informed drug development of CX-072.
- To establish pharmacokinetic and exposure-response relationships for CX-072.
- To support dose selection for further clinical evaluation.
Main Methods:
- Quantitative systems pharmacology (QSP) modeling to predict Probody therapeutic (Pb-Tx) behavior.
- Analysis of preliminary human pharmacokinetic (PK) data from single-dose studies.
- Population PK (POPPK) analysis incorporating antidrug antibody (ADA) effects.
- Exposure-response analyses to assess safety and efficacy relationships.
Main Results:
- QSP model predicted target receptor occupancy based on masked CX-072 trough levels.
- Preliminary PK data aligned with QSP predictions, showing CX-072 circulating as intact drug.
- POPPK analysis estimated intact CX-072 clearance and volume of distribution.
- No statistically significant exposure-response relationships were found for adverse events or efficacy measures.
- Simulations indicated a fixed 800 mg dose every two weeks would achieve target trough levels in most patients, irrespective of ADA status.
Conclusions:
- Model-informed drug development successfully guided the clinical translation of CX-072.
- CX-072 exhibits predictable pharmacokinetics, with a fixed dose regimen being suitable for Phase II studies.
- The fixed 800 mg dose is supported for further clinical evaluation, demonstrating potential efficacy independent of ADA presence.
