Kim Stuyckens (1) Italo Poggesi (1) Nadia Terranova (7) Dominique Swerts (1) Vijay Peddareddigari (2), Feng Roger Luo (3) Suso J. Platero (2) Moitreyee Kishore (1), Rastislav Bahleda (4), Josep Tabernero (5) Jeffrey R. Infante (6) Johan W. Smit (1)
(1) Janssen Research and Development, Beerse, Belgium; (2) Janssen Research and Development, Spring House, PE; (3) Janssen Research and Development, Raritan, NJ; (4) Drug Development Department (DITEP), Gustave Roussy Institute, Villejuif, France; (5) Vall d'Hebron University Hospital, Barcelona, Spain; (6) Sarah Cannon Research Institute/Tennessee Oncology, PLLC, Nashville, TN, USA (7) University of Pavia, Italy
Objectives: JNJ-42756493, an oral potent and selective FGFR1,2,3 and 4 inhibitor, is currently under development for patients with cancer with FGFR aberrations. A pharmacokinetic/pharmacodynamics (PK/PD) model approach to support the dosing strategy was applied for managing safety by controlling serum phosphate and calcium homeostasis and safeguarding efficacy.
Methods: After each dose level of the dose escalation portion of a first in human study, biomarkers indicative of efficacy and safety, plasma PK as well as AE information were evaluated. A population PK model was developed using non-linear mixed-effects modelling (NONMEM 7.2) to assess exposure profiles across the different dose levels (0.5, 2, 4, 6, 9, 12 mg QD). JNJ-42756493 human plasma exposure profiles were projected on the window of threshold concentrations for efficacy based on a mouse xenograft PK/PD model [1]. In addition, exposure profiles were linked to mode of action-related safety and efficacy biomarkers: FGF23, serum phosphate, parathyroid hormone, calcium, and vitamin D.
Results: A two-compartment model with first-order absorption was shown to accommodate JNJ-42756493 PK data. Body weight on Vc/F and alpha-1-acid glycoprotein (AGP) on CL/F and Vc/F were identified as the most significant covariates affecting the absorption and distribution characteristics of the drug. Likely due to the variable AGP levels in patients (45-254 mg/dL) and the extensive binding to AGP, the unbound fraction ranged between individuals (0.1-0.8%), influencing the inter-patient variability of PK. PK were dose-proportional and time independent over the tested dose range. Apparent clearance and distribution volumes were relatively low; the compound had a half-life of approximately 3 days. Since hyperphosphatemia was observed at the higher doses, PK-PD models linking phosphate and exposure were developed to allow simulating various dosing strategies, thereby maintaining exposures within the xenograft based efficacy window controlling phosphate levels. Simulated dosing strategies are currently tested in the clinic.
Conclusions: Modelling and simulation of JNJ-42756493 pharmacokinetics and pharmacodynamics helped the conduct of the 1 study, safeguarding efficacy while remaining below predefined safety thresholds.
References:
[1] Simeoni, M et al. Clinical Cancer Research 2004; 64, 1094-1101.
The authors also acknowledge: University of Pavia and DDMoRe (Nadia Terranova, Paolo Magni, Giuseppe De Nicolao) for the analysis of the xenograft model data.
Reference: PAGE 23 (2014) Abstr 3215 [www.page-meeting.org/?abstract=3215]
Poster: Drug/Disease modeling - Oncology