LB-027 - PHYSIOLOGICALLY-BASED PHARMACOKINETIC MODELING TO PREDICT HUMAN DOSE OF MITHRAMYCIN ANALOGUE FOR THE TREATMENT OF EWING SARCOMA

K. Niloy^1,2, M. Leggas^3,2; ¹St. Jude Children's Research Hospital, The University of Tennessee Health Science Center, Memphis, TN, United States, ²The University of Tennessee Health Science Center, Memphis, TN, United States, ³St. Jude Children's Research Hospital, Memphis, TN, United States.

Background: Mithramycin (MTM) was identified as an inhibitor of the EWS-FLI1 oncogenic transcription factor fusion, the causative agent of Ewing sarcoma (ES). Poor pharmacologic properties and severe toxicity limited its clinical use in children with ES. MTMSATrp, an MTM analogue, has improved pharmacologic properties. The objective of this work was to develop and qualify a PBPK model of MTM using preclinical and clinical pharmacokinetic (PK) data followed by MTMSATrp preclinical PBPK model development and qualification using same modeling workflow to predict human dose of MTMSATrp.

Methods: A PBPK model was developed for MTM in GastroPlus® using mouse PK data and qualified using rat and monkey PK data. A “middle-out” approach was followed where renal clearance (CL) was assumed across species and calculated using non-compartmental analysis. Physicochemical parameters were obtained from ADMET Predictor® or experimentally. Volume of distribution was predicted using the Poulin-Theil extracellular method. Specific permeability-surface area product (SpecPStc) was optimized using mouse PK data and assumed to be same across species. The average value of the % of additional CL not captured by renal CL equation (CLR = Fup x GFR) in preclinical species was added as additional CL to predict human CL. Following MTM human PBPK model qualification using clinical PK data, the same modeling workflow was followed to develop and qualify PBPK model for MTMSATrp and human dose that matches exposure required for efficacy in mouse xenograft was predicted.

Results: The PBPK models of MTM and MTMSATrp were able to capture the PK profiles in preclinical species and human (for MTM) and predicted AUCs were within 2-fold of the observed AUCs. To achieve similar plasma exposure observed in mouse at maximum tolerated doses of 0.9 mg/kg and 2.5 mg/kg on Q1Dx5 and Q3Dx6 schedules, MTMSATrp human PBPK model predicted 145 µg/kg and 400 µg/kg 6h IV infusion, respectively, for 13 years old and 50 kg American male physiology.

Conclusion: Upon PBPK model development and qualification of MTM in preclinical species and human, same modeling workflow led to the development and qualification of MTMSATrp preclinical PBPK model. Scaling the preclinical PBPK model to human led to prediction of human MTMSATrp dose required to achieve plasma exposure found efficacious in mouse xenograft.