Parametric and non-parametric approaches to quantitating pharmacodynamic response Public Deposited

Downloadable Content

Download PDF
Last Modified
  • March 21, 2019
  • Olson, Emily Renee
    • Affiliation: Eshelman School of Pharmacy
  • The goal of this project was to evaluate factors that impact pharmacokineticpharmacodynamic (PK-PD) relationships, and to develop novel approaches for analysis of those relationships utilizing opioids as a model pharmacologic class. The role of protein binding in in vitro-to-preclinical and preclinical-to-clinical estimates of opioid potency was investigated; the optimal metric for characterizing in vivo potency was unbound brain EC50. The role of P-glycoprotein (P-gp) efflux at the blood brain barrier (BBB) in attenuating central nervous system (CNS) concentrations and pharmacologic response was evaluated in mice receiving loperamide, methadone, alfentanil or fentanyl. An integrated PK-PD model fit to the time course of antinociception and serum/brain concentrations revealed the influence of P-gp on brain:plasma ratio and brain:plasma equilibration half-life. The influence of behavioral responses in the hotplate latency antinociception assay was investigated for loperamide, methadone and sufentanil. PK-PD modeling suggested that processed responses, jumping and hind-paw licking, reflected centrally-mediated processes, whereas reflexive responses, lifting or shaking of the hind paw, were mediated predominately in the periphery, with a slight additive central contribution. The influence of truncating the dynamic range of pharmacologic effect on recovery of PD parameters was investigated with fentanyl and morphine. Incorporating Emax as a parameter, as opposed to a fixed value, in PK-PD modeling improved model predictions when compared to assigning a fixed value at an artificial ceiling. The utility of area bounded by a hysteresis loop (ABH) as a nonparametric descriptor of dissociation between pharmacodynamics and pharmacokinetics was assessed with five opioids administered to P-gp-competent and P-gp-deficient mice. A Matlab routine was developed to calculate ABH, and this novel quantitative approach was evaluated relative to image analysis of the hysteresis loop. ABH correlated with area under the curve (AUC) and the model-derived parameters EC50 and ke0. Taken together, these approaches provide a framework for understanding the multiplicity of factors that influence drug concentrations and pharmacologic effect, and provide insight into how information derived from parametric and non-parametric analyses can be used to characterize PK-PD relationships and predict the behavior of PK-PD systems.
Date of publication
Resource type
Rights statement
  • In Copyright
  • Pollack, Gary M.
Degree granting institution
  • University of North Carolina at Chapel Hill
  • Open access

This work has no parents.