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 of this study was to assess the relationships between ziftomenib exposure and relevant efficacy and safety endpoints in patients with R/R AML and to identify risk factors on exposure-response (E-R) relationships.
Methods: The E-R population included patients enrolled in Study KOMET-001 (NCT04067336) with available PK and response data. E-R analyses for efficacy and safety were performed for the overall patient population (patients with NPM1-m, lysine N-methyl transferase 2A rearrangements (KMT2A-r), and/or other mutations), the population of NPM1-m patients, and the population of KMT2A-r patients. Steady-state exposure levels of ziftomenib were computed based on individual Bayes parameters of population pharmacokinetic model and dose (50 mg, 100 mg, 200 mg, 400 mg, 600 mg, and 800 mg). Logistic regression models were constructed to assess the statistical relationships between ziftomenib exposure levels, risk factors, and the probability of responses based on various combinations of the following efficacy endpoints- i) complete response (CR), ii) partial hematologic response (CRh), iii) morphological CR without measurable residual disease (CRmrd-), iv) CR with incomplete count recovery (CRi,), v) CR with incomplete platelet recovery (CRp), or vi) morphologic leukemia free state (MLFS). Probability of adverse events (AE) with toxicity grade of ≥3 was evaluated, except for differentiation syndrome (DS), where probability was assessed across all toxicity grades. The best fitting model according to Akaike information criterion was selected. Covariate analysis was performed only on logistic models with significant contribution of exposure levels (p < 0.05). Stepwise approach was used with forward additive and backward elimination approach (p < 0.05 and p< 0.01).
Results: No exposure parameters of ziftomenib were statistically related to any efficacy endpoints for the NPM1-m population. In the overall efficacy population and in patients with KMT2A-r, significant positive relationships between maximum concentration of ziftomenib and probability of overall response (MLFS, CRp, CRi, CRh, CR, CRmrd-) and complete/partial response (CRp, CRi, CRh, CR, CRmrd-) were observed (p < 0.05). No significant safety E-R relationships were observed for i) any grade ≥3 AEs, ii) any grade ≥4 AEs, iii) any grade DS, or iv) any grade AEs that occurred in ≥10% of the safety population for the NPM1-m, KMT2A-r, or overall populations.
Conclusions: The absence of E-R relationship on efficacy and safety endpoints in the KOMET-001 R/R NPM1-m AML population suggests that a 600 mg QD dose of ziftomenib monotherapy demonstrates a positive benefit-risk profile and is the appropriate dose in this population.
