PII-016 - PHARMACOKINETICS OF SHIP1 INHIBITORS IN MICE: COMPARISON OF IN SILICO AND IN VIVO PHARMACOKINETICS.

M. Kandasamy¹, T. Richardson¹, A. Masters¹, C. Bach¹, S. Sukoff Rizzo², S. Doolen², S. Quinney¹, C. Jesudason³; ¹Indiana University, Indianapolis, IN, USA, ²University of Pittsburgh, Pittsburgh, PA, USA, ³Lgenia, Inc., Indianapolis, IN, USA.

Background: INPP5D, expressed in microglia, has been identified as an Alzheimer’s Disease (AD) risk gene and encodes Src homology 2 (SH2) domain–containing phosphatase-1 (SHIP1). TAD025 and TAD032 inhibit SHIP1 in microglial assays by the IU-Purdue TREAT-AD Center. Physiologic based pharmacokinetic (PBPK) modeling using in silico and in vitro data was used to estimate mouse plasma and brain PK and compared with in vivo data.

Methods: PBPK models of TAD025 and TAD032 were developed using GastroPlus 9.8.2 based on in silico predictions (ADMET Predictor 10.3). Models were updated with in vitro fu,p, CLint, and MDCK2 Papp. Drugs were administered orally (100 mg/kg) to adult male C57BL/6J mice. Serial blood samples were analyzed by LC-MS/MS.

Results: Half-life predictions were improved by incorporating in vitro data, but bioavailability was overpredicted for both TAD025 and TAD032. Predicted to observed plasma AUC24 ratios were 10.7 and 16.4 based on in silico and in vitro parameters for TAD025 and 7.5 and 29 for TAD032. The predicted 1.5-fold accumulation in brain was confirmed by observed 24 hour brain to plasma ratios of 2.8 (TAD025) and 6.1 (TAD032).

Conclusion: In silico PBPK predictions benefited from the incorporation of in vitro data, but bioavailability appears to be underpredicted for both drugs.

Comparison of observed and predicted concentrations of SHIP1 inhibitors using in-silico and in-vivo models