(S-003) Population PK Modeling of Ziftomenib and its Metabolites in Healthy Subjects and in Patients with Relapsed or Refractory Acute Myeloid Leukemia
Sunday, October 19, 2025
7:00 AM - 5:00 PM MDT
Location: Colorado A
Amitava Mitra – Kura Oncology, Inc; Xiaoyan Yang – Kura Oncology, Inc; Roberto Ortiz – Certara; Claudia Jomphe – Certara; Mollie Leoni – Kura Oncology, Inc; Nathalie Gosselin – Certara
Executive Director Kura Oncology, Inc Boston, Massachusetts, United States
Disclosure(s):
Amitava Mitra, PhD: No financial relationships to disclose
Background: Ziftomenib is an investigational, small molecule menin inhibitor for oral administration. Ziftomenib binds to the site of the interaction between menin and the histone methyl transferase mixed-lineage leukemia (MLL), which plays a critical role in acute leukemias. Ziftomenib is in clinical development for the treatment of adult patients with relapsed or refractory (R/R) acute myeloid leukemia (AML) with a nucleophosmin 1 mutation (NPM1-m). The objective is to develop a population PK model of ziftomenib and its metabolites in adult patients with R/R AML and to evaluate covariates that could potentially affect the exposure of ziftomenib and its metabolites.
Methods: A non-linear mixed effect model (NLME) in NONMEM VII (level 7.5) was developed using the pooled concentration-time data for ziftomenib and its metabolites (KO-739 and KO-516) from healthy adult subjects (N = 14) and adult subjects with R/R AML (N = 174). PK analysis included 2436, 2376, and 2299 samples with measurable plasma concentrations of ziftomenib, KO-739, and KO-516, respectively. A sequential 2-stage approach was selected for model fitting of ziftomenib and its metabolites. In a first step, a population PK model for ziftomenib was developed. In a second step, individual Bayes parameters derived from the final model were used to link the PK of ziftomenib to the PK of both metabolites. At each step, relationships between random effects of PK parameters vs. covariates were explored to identify potential sources of variability. The selection of covariates was done formally using a univariate forward addition (with significance level of p< 0.01). All significant covariates were then included in a full model and then a backward elimination was conducted (with significance level of p< 0.001) to obtain the final model.
Results: The observed plasma concentration-time profiles of ziftomenib were adequately described by a 2-compartment, linear-elimination model with first-order absorption and lag time. A thorough covariate analysis using the population PK model showed that the mutational status of patients did not impact ziftomenib clearance. Additionally, no clinically meaningful effects on the PK of ziftomenib were identified for body weight, sex, race, age or moderate hepatic impairment, or P-gp inhibition. Food, proton pump inhibitors (PPI), CYP3A inhibitors, and health status were identified as covariates that contributed to the variability of ziftomenib PK in the studied population. For each metabolite, a 2-compartment, linear-elimination model was retained for the structural model. The disease status covariate was implemented on the biotransformation fraction from ziftomenib into KO-516 and KO-739 and the CYP3A4 inhibitor covariate on clearance of each metabolite. Finally, covariate analysis resulted in inclusion of disease status in the evaluation of central volume of distribution for each metabolite.
Conclusions: Population PK analyses successfully described the disposition of ziftomenib and its active metabolites. No additional covariates, except food and PPI, were identified that could significantly affect ziftomenib PK and thus dose modification is not warranted in patients with R/R AML.
