PII-183 - POPULATION PHARMACOKINETICS AND PHARMACODYNAMICS OF PF-07265803 AND ITS METABOLITE IN PATIENTS WITH LAMIN A/C RELATED DILATED CARDIOMYOPATHY.

H. AL Nebaihi¹, S. Riley², P. Jayachandran³; ¹Faculty of Pharmacy and Pharmaceutical Sciences, University of Alberta, ²Pfizer, Inc., ³Regeneron Pharmaceuticals, Inc..

Background: LMNA-related Dilated Cardiomyopathy (DCM) results from the loss of functional lamin proteins, which activates p38α MAPK. Activation is associated with progressive heart failure (HF) and sudden cardiac death. Despite conventional HF therapies, specific efficacy in this population is uncertain. Inhibiting p38 MAPK with the selective p38α inhibitor (PF-07265803) offers a potential novel therapy.

Methods: Data were pooled from double-blind, placebo-controlled Phase I studies in healthy volunteers (HV) and open label/double-blind phase II-III in LMNA-DCM patients. Doses of 25-1500 mg were administered to 126 HV while 46 LMNA-DCM patients received 100/400 mg BID. PK samples were collected pre-dose to 60 weeks. N-terminal pro-brain natriuretic peptide (NT-proBNP) was collected from baseline to 213 weeks. PopPK modeling was conducted using NONMEM 7.5 following a sequential modeling approach. Renal function, body weight (BW) and sex were assessed as covariates. Mrgsolve was utilized to simulate full PK profiles for patients. A longitudinal linear mixed effect model characterized the relationship between Area Under the Curve (AUCꞇ) and NT-proBNP level.

Results: Two-compartment model with first-order absorption and lag time, weight as a covariate on clearance (CL) and volume distribution, with separate CL for HV and LMNA-DCM subjects, was chosen as the final model. Mean clearance was lower in LMNA-DCM patients than HVs (1360 L/h vs 1650 L/h). The active metabolite exhibited CL < 8 L/h. PF-07265803 reduced NT-proBNP change rate by ~129% compared to placebo. Age was a significant covariate.

Conclusion: Two-compartment model with first-order absorption described PF-07265803 PK. Health status and weight accounted for ~5% of CL variability. Visual predictive checks and model diagnostics proven the adequacy of the PK model. A longitudinal NT-proBNP model demonstrated exposure-dependent reductions in NT-proBNP levels over time.