Table 7.

List of compounds for which Full PBPK models were used to address pharmacological and toxicological questions

Compound	Model purpose	A priori criteria?	Model Quality Assessment	Conclusions	Citation
Acetaminophen	Assessing various calibration strategies for linking PBPK models to toxicodynamic models of hepatotoxicity	No	Qualitative discussion of the agreement between simulated and observed Cmax, and metabolite ratios	Predicted liver toxicity in agreement with observed	(Péry et al., 2013)
Cyclosporine	Simulation of receptor occupancy in accute graft-versus-host organs and kidneys after intermittent or continuous infusion	No	Mann-Whitney test to compare means, chi-square test to compare proportions, bias and precision, number of simulations within 2-fold of the observed, weighted residuals	A greater therapeutic index was predicted following continuous infusion	(Gérard et al., 2010)
Cyclosporine	Establish a connection between the likelihood and severity of graft-versus-host disease and cyclosporine exposures in circulation, graft-versus-host target organs and lymphoid tissues	No	Student's t-test to compare means, chi-square test to compare proportions, Akaike information criterion for model selection	Blood cyclosporine levels can be used as an indicator of therapeutic efficacy	(Gérard et al., 2011)
Efalizumab	Develop a PD linked PBPK model to predict efficacy of efalizumab	Yes	Observed data within the predicted 5th and 95th centile	The model predicted the efficacy of efalizumab in treatment of psoriasis	(Chetty et al., 2015)
Formamide	Evaluate the relationship between dose and hepatic exposure	No	—	40mg/day dose was proposed bases on a safety index	(Yan et al., 2012)
Levofloxacin	Exploratory study to predict the extent of tissue exposure of levofloxacin in humans as a basis for future PK/PD work.	Yes	Fold error in PK parameters less than 2	Levofloxacin penetrated well into tissues, including the liver kidneys and spleen	(Zhu et al., 2015a)
Moxifloxacin	Simulate tissue concentrations versus time in patients with intra-abdominal infections	Yes	Fold error in PK parameters less than 2	Concentrations in intra-abdominal tissues were predicted to be higher than the minimum inhibitory concentration for common pathogens	(Zhu et al., 2015b)
Moxifoxacin	Evaluate the effect of macrophages on tissue concentrations of moxifloxacin to enhance understanding of the effects of disease on PK/PD	No	Simulated concentration versus time profiles were evaluated for bias and precision	Macrophage concentrations are predicted to effect tissue concentration of moxifloxacin	(Edginton et al., 2009)
Nicotine	Develop a PBPK model to describe nicotine exposure and receptor binding in the brain	No	Qualitative discussion of the agreement of the predicted and observed data	PK/PD modeling allowed for prediction of nicotine receptor occupancy in the brain	(Teeguarden et al., 2013)
“S1”	Predicting brain extracellular fluid concentrations as a starting point for PK-PD modeling	No	—	Unclear whether the PBPK model would accurately predict PD	(Ball et al., 2014)
Temozolomide	PD linked PBPK model for simulating the brain concentration of temozolomide and the levels DNA brain adducts	No	—	Predictions were in close agreement with observed data and parameter estimates had low coefficients of variation	(Ballesta et al., 2014)
Zidovudine	Model intra-cellular concentrations of zidovudine in peripheral blood mononuclear cells and establish efficacy and toxicity following various dosing regimens	No	PK features are well represented by the predictions	100mg 4 times daily is predicted to be the safest and most efficacious dosing scheme	(von Kleist and Huisinga, 2009)
Theoretical Compounds	Proof of concept study to evaluate mechanisms for differences in unbound plasma and tissue concentrations	No		This approach can be used to predict free tissue concentrations of various classes of drugs	(Poulin, 2015)