Feed

Ziftomenib, a potent, highly selective, oral menin inhibitor, has demonstrated significant clinical benefit and deep responses in relapsed and/or refractory (R/R) nucleophosmin 1-mutated (NPM1-m) acute myeloid leukemia (AML) patients. Here we present data from analyses conducted to characterize ziftomenib pharmacokinetics (PK) in healthy volunteers and R/R AML patients and exposure-response (ER) relationships in R/R NPM1-m AML patients. A sequential two-stage modeling approach was employed to develop the popPK model for ziftomenib and its metabolites. A 2-compartment model with first-order elimination and first-order absorption with a lag time adequately described the PK data for ziftomenib with good precision. Thorough covariate analysis showed that the mutational status (NPM1-m vs. KMT2A-r) of patients, body weight, sex, race, age, mild or moderate renal or hepatic impairment, and P-gp inhibition did not impact ziftomenib PK. ER analyses were conducted using logistic regression models in a total of six combinations of efficacy endpoints and 12 safety endpoints. No steady-state exposure parameters of ziftomenib were statistically related to any of the efficacy or safety endpoints in R/R NPM1-m AML patients. In fact, flat ER profiles were observed between all ziftomenib exposure parameters and efficacy or safety endpoints over a wide range of ziftomenib exposure. The modeling results demonstrated a wide therapeutic margin for ziftomenib in adult R/R NPM1-m AML patients and supported a dose of 600 mg once daily in this patient population. Additionally, ER analyses demonstrated that antifungal azoles had no clinically meaningful impact on efficacy or safety profiles and thus could be co-administered with ziftomenib.