Exposure-Response Relationships for Axatilimab, a Humanized Monoclonal Antibody Targeting CSF-1R, in Patients with Chronic Graft-Versus-Host Disease

Introduction: Axatilimab, a high-affinity humanized monoclonal antibody against colony-stimulating factor 1 receptor (CSF-1R), is under investigation for the treatment of chronic graft-versus-host disease (cGVHD) and other diseases. A population pharmacokinetic/pharmacodynamic (PK/PD) model for axatilimab was previously developed using pooled data from 4 clinical studies. The structural model consisted of a 2-compartment axatilimab PK model with saturable clearance and turnover models for colony-stimulating factor 1 (CSF-1), nonclassical monocytes, aspartate aminotransferase, and creatine phosphokinase (CPK). Body weight was the only covariate identified that affected axatilimab steady-state exposure by >20%. The objective of this analysis is to describe the exposure-response relationships for efficacy and safety in patients with cGVHD who received axatilimab.

Methods: Exposure-efficacy relationships were assessed in patients treated in the phase 2 AGAVE-201 study (NCT04710576; n=239; axatilimab 0.3 mg/kg every 2 weeks [Q2W], 1.0 mg/kg Q2W, 3.0 mg/kg every 4 weeks [Q4W]). Binary efficacy assessments included the overall response rate (ORR) and ≥7-point improvement in modified Lee Symptom Scale (mLSS response). Duration of response (DOR), a time-to-event endpoint, was assessed among all patients in AGAVE-201 who achieved an overall response. Exposure-safety relationships were assessed in all treated patients with cGVHD (n=278) in AGAVE-201 and a phase 1/2 study (SNDX-6352-0503; NCT03604692; axatilimab 0.15 mg/kg Q2W, 0.5 mg/kg Q2W, 1.0 mg/kg Q2W, 3.0 mg/kg Q2W, 3.0 mg/kg Q4W). Evaluated safety endpoints included 5 general safety assessments (grade ≥3 treatment-emergent adverse events [TEAEs], TEAEs leading to dose modifications, serious TEAEs, treatment-related TEAEs, AEs of special interest) and 6 sets of grouped AE terms (amylase and lipase increases, CPK elevations, liver enzyme elevations, periorbital edema, infections of unspecified etiology [infections not otherwise specified as bacterial, viral, or fungal], infusion-related reactions). For binary or time-to-event endpoints, logistic or Cox regression analyses, respectively, were performed using predicted axatilimab exposure metrics that were derived from the population PK/PD model; axatilimab exposure metrics from the first dose, the first treatment cycle, and a steady-state treatment cycle were evaluated. To further evaluate the effects of body weight with the 0.3 mg/kg Q2W regimen, forward simulations were completed using percentiles of an observed body weight distribution (range, 18.1–151 kg) for each efficacy and safety outcome that was associated with axatilimab exposure in the final model.

Results: For the exposure-efficacy analysis, ORR and mLSS responses were associated with axatilimab exposure, with lower axatilimab exposure increasing the odds of response. Among the 153 patients with a response, DOR did not have a significant association with axatilimab exposure. For the exposure-safety analysis, all safety endpoints except infections of unspecified etiology were associated with axatilimab exposure, with higher axatilimab exposure increasing the odds of TEAEs. In forward simulations evaluating the effect of body weight on the axatilimab 0.3 mg/kg Q2W regimen across 2 efficacy and 10 safety endpoints, the maximum differences in median predicted probabilities between the 10th and 90th percentiles of with body weight were <1.4% and <1.7% for efficacy and safety, respectively.

Conclusions: These results support the 0.3 mg/kg Q2W regimen of axatilimab in patients with cGVHD.