(S-076) Exposure-Response Characterization of the Safety Profile for a Novel KAT6 Inhibitor, PF-07248144, For Use In Dose Optimization During a Phase 1 First-in-Patient Study

Marlon Liyanage, PharmD – Pfizer Inc; Karey Kowalski – Pfizer Inc; Jason Williams – Exelixis; Dustin Huynh – Unaffiliated; Meng Li – Pfizer Inc; Joanna Masters – Pfizer Inc

Objectives: KAT6A, and its paralog, KAT6B, are histone lysine acetyltransferases (KATs) that regulate lineage-specific gene transcription via H3K23 acetylation (H3K23Ac). PF-07248144 is a novel, potent and selective catalytic KAT6A and KAT6B inhibitor currently being developed in ER+ HER2- metastatic breast cancer. A Phase 1 study (C4551001) to evaluate the safety, pharmacokinetics (PK), and early signs of efficacy of PF-07248144 is ongoing. The current analyses evaluated the potential exposure-response (E-R) relationships between PF-07248144 and safety endpoints of interest.

Methods: Exposure-response analyses for PF-07248144 were conducted using relevant clinical data from the ongoing Phase 1 study. Patients included in this analysis received PF-07248144 once daily as monotherapy or in combination with fulvestrant at doses ranging from 1 mg to 15 mg once daily, with expansion cohorts at 1 mg and 5 mg. Due to the dose modifications seen in study C4551001, longitudinal pharmacokinetic-pharmacodynamic (PKPD) models were pursued. Since ≥Grade 3 neutropenia was the main adverse event and driver of dose modifications, a nonlinear mixed effect semi-mechanistic model of myelosuppression was used to characterize the ER relationship and simulations were used to predict ≥Grade 3 neutropenia rates at various dosing regimens. Logistic regression analyses were also conducted for additional safety endpoints of interest including any grade dysgeusia, Grade 2 dysgeusia, ≥Grade 3 anemia, and ≥Grade 3 treatment emergent adverse events (TEAEs). PF-07248144 exposure metrics evaluated in logistic regression were generated from a population pharmacokinetic (popPK) model and included Cmax, Cavg, and Ctrough following single dose and at steady state (Day 15).

Results: The longitudinal PKPD model of neutrophil count over time used a typical semi-mechanistic myelosuppression model structure with an Emax drug concentration effect on cell proliferation rate and random effects on baseline neutrophil count and Emax. Model diagnostics indicated an adequate fit to the data and the model was considered appropriate for further PKPD simulations using the fixed and random effects from the associated popPK model and the PKPD neutropenia model. Simulations of ≥Grade 3 neutropenia rates showed a difference in central tendency between 1 mg and 5 mg that was in agreement with clinical observations from study C4551001.

A positive exposure-response relationship was found between PF-07248144 single-dose Ctrough and ≥Grade 3 TEAE. No significant exposure-response relationship was found for any grade dysgeusia, Grade 2 dysgeusia, or ≥Grade 3 anemia.

Conclusions: Based on clinical data from an ongoing Phase 1 study, PF-07248144 exhibits a positive exposure-response relationship for neutrophil count time course and ≥Grade 3 TEAEs, which are primarily influenced by neutropenia. Dysgeusia and ≥Grade 3 anemia do not appear dependent on exposure in the dose range tested. The E-R characterization across a variety of safety endpoints identified neutropenia as the primary safety endpoint to inform the benefit-risk ratio for differentiation of various dosing regimens of PF-07248144 and therefore aid in dose selection.

Citations: None

Keywords: KAT6, dose optimization, exposure-response