Benjamin Kably (1), Ludwig Vincent (2), Laurence Proust (2), Julie Trasbot (2), Laurence Launay (2), Maud Beneton (2), Yannick Parmentier (2), Marylore Chenel (2)
(1) Pharmacology, Hôpital Européen Georges Pompidou, AP-HP, (2) Centre of Excellence in Pharmacokinetics, Servier, France
Objectives:Mechanism-based inhibition (MBI) is a subtype of enzymatic inhibition in which the enzyme is permanently inactivated by an intermediate reactive metabolite that irreversibly binds to the active site. It has been shown that for this type of inhibitors, the recommended methods to predict the risk of drug-drug interactions (DDI) used in early development stages tend to over-estimate the real interaction observed in clinical studies [1][2]. Several factors may impact these predictions and could explain this over-prediction: inhibitor concentration ([I]), inhibition parameters (KI/kinact) or enzyme degradation constants (kdeg) used in the calculations. The aim of this work was to evaluate the different DDI risk assessment methods for MBI and the impact of each parameter in order to identify the optimal conditions for a more accurate prediction.
Methods: We first identified 14 MBI drugs and collected the parameters needed to predict DDI (fu, Cmax, kdeg, t1/2(CYP), KI, kinact, Fg, fm(CYP)). We then compared the different calculation methods: the basic model, the mechanistic static model and the dynamic PBPK model. We evaluated the impact of each parameter. For PBPK models, we selected drugs for which static models mispredicted the risk, whatever the parameters used (Ritonavir, Saquinavir, Dasatinib and Rofecoxib). The PBPK models were developed and qualified with SimCYP®. As mentioned above, inhibition parameters (KI, kinact) could have a large impact on DDI risk prediction. Therefore, two experimental methods used for determining these parameters were also compared: i) a conventional method which includes a pre-incubation step with the MBI before the standard substrate co-incubation; ii) an alternative one-step substrate disappearance kinetic method that allows to take the inhibitor depletion and the competitive inhibition into account. The results obtained from the alternative method were computed using the SIVA® software.
Results: The comparison of the DDI risk predictions obtained for 14 MBI drugs with the static method showed that among the 4 possible inhibitor concentration values ??(Cmax, Cmax,u, Cin, Cin,u) the Cmax,u gave the largest number of acceptable predictions (respectively 12.2%, 39.3%, 0% and 8.2% of Rpredicted/Robserved ratios contained within the [0.5; 2] interval). The DDI risk assessment performed using PBPK modeling for the 4 selected molecules showed acceptable Rpredicted/Robserved ratios for all the interaction studies simulated and therefore greatly improved the predictions. Using the alternative method, the inhibition parameters KI and kinact could only be determined for one drug (Ritonavir) and the values were similar to the previously published values.
Conclusions: Our results showed that the basic model more accurately predicts the risk of DDI with MBIs using Cmax,u as the inhibitor concentration. The alternative method for KI/kinact determination, that can potentially overcome several hypotheses made by the two-step conventional method, was evaluated but did not show a clear benefit in the case of the Ritonavir. The simultaneous estimation of other parameters (Kcat, Kd, ??Km) on top of the MBI’s parameters (KI/kinact) may explain the shortcoming of this model. All the more since these parameters are often unknown in early stages of the development. Despite the need of many parameters to be qualified and used as a tool for clinical studies simulations, the use of PBPK modeling showed a real improvement regarding the accuracy of DDI risk prediction for MBIs.
[1]Fujioka Y, Kunze KL, Isoherranen N. Risk assessment of mechanism-based inactivation in drug-drug interactions. Drug Metab Dispos. 2012
[2]Obach RS, Walsky RL, Venkatakrishnan K. Mechanism-based inactivation of human cytochrome p450 enzymes and the prediction of drug-drug interactions. Drug Metab Dispos. 2007
Reference: PAGE 28 (2019) Abstr 8835 [www.page-meeting.org/?abstract=8835]
Poster: Drug/Disease Modelling - Absorption & PBPK