H. Yang,1 S. Ramalingam,2 S. Kummar,3 R. Harvey,2 P. Ivy,4 J. Beumer1; 1University of Pittsburgh, Pittsburgh, PA, 2Winship Cancer Institute of Emory University, Atlanta, GA, 3National Institutes of Health Clinical Center Maryland, Bethesda, MD, 4Investigational Drug Branch, National Institutes of Health, Pittsburgh, PA

BACKGROUND: This study was undertaken to develop a population-PK (PPK) model for vorinostat in cancer patients with varying degrees of hepatic function.
METHODS: Vorinostat was quantitated in 563 plasma concentrations from 55 patients in the PK portion after administering 400 mg single dose vorinostat PO. Platelet counts from 50 patients after vorinostat dosing ranging from 100-400 mg QD for a 3-week cycle in the treatment portion were collected. PPK analysis was performed with NONMEM. Impact of baseline covariates on PK were examined: body weight, age, sex, CrCL, performance status, LDH, albumin, and hepatic function status (NCI-ODWG) through backward elimination (if <6.63 objective function decrease (χ2, df = 1, p < 0.01). Individual AUCss were estimated for the treatment part to evaluate linear or Emax models describing the PK-PD of the change of platelet counts (cycle 1 week 1 cycle 2 week 1), as well as the impact of covariates.
RESULTS: A one-compartment model with first order absorption and elimination with proportional residual error fit the PK data. None of the covariates was significant. Estimates of CL/F (geometric mean±CV% L/h) were comparable between normal (228±44.7), mild (259±62.0), moderate (227±58.9) and severe (260±125) hepatic impairment. PPK parameter estimates of CL/F; V/F and Ka were 240 L/h, 288 L and 0.317 h-1, respectively. PK-PD analysis suggested no apparent association between AUCss and change of platelet count. Platelet decrease was also independent of hepatic function status or any other covariates.
CONCLUSION: The PPK model adequately described vorinostat data in our study population. The analysis showed that PK appears to be independent of hepatic impairment.