Physiologically-based Optimization Of IVIVE Modeling

The processes of absorption(A),distribution(D),metabolism(M),and excretion(E)in the organism determine the therapeutic effects and side effects of the drug.In recent years,international pharmaceutical companies have used in vitro prediction methods to evaluate the early ADME properties of candidates to find the potential for their development.This method has greatly reduced the high failure rate caused by poor pharmacokinetic properties in the development of new drugs.Clearance is an important parameter of pharmacokinetics.Accurate prediction of drug clearance rate in vivo is of great significance to the accurate prediction of pharmacokinetic properties of drugs in humans.At present,there are many methods to predict the clearance rate by combination of in vitro model and mathematical model.However,the accuracy still needs to be improved.Therefore,on the basis of optimizing the existing mathematical model,the current study intends to introduce the physiological factors of human body to optimize the existing model,and investigate the accuracy of the model by in vitro experimental data from liver microsomes.Firstly,model optimization is carried out for the most commonly used model,well-stirred model,to predict liver clearance.The premise of the full stirring model is relatively ideal.It is assumed that the concentration of free drugs between plasma and liver tissues is equal,and the process instantly reaches equilibrium.However,experimental evidences have showed that the concentrations of free drugs and plasma in liver cells are not equal,which greatly affects the accuracy of the model.Considering the complexity of the physiological environment and the complex relationship between drugs and tissue components,in particular,most of drug molecules have a certain degree of lipid solubility.The non-specific binding of drugs to important components in tissues,such as proteins and lipids,will greatly affect the free concentration of drugs.Therefore,this study optimizes the model from two aspects.First,considering the difference between the in vitro experiment and liver cell pH and the distribution of drugs in liver tissue,a new model of well-stirred was established.Compared with the traditional well-stirred model and the Berezhkovskiy optimization method,the average fold error(AFE)values are 1.26,0.39,and 0.46,respectively;the root mean squared error(RSME)value is 0.38,0.65,0.54.The second method is to emphasize that clearance of the liver is the ratio of the total amount of liver scavengers per unit time to the immediate plasma concentration,and to establish a new model.In the multiples map,the main distribution is within the error range of three times,and the AFE value is 0.69,0.44;RMSE values 0.28,0.52.The prediction accuracy and accuracy of the model are improved.Secondly,we used the combination of in vitro experiments and mathematical models to predict the possibility of drug-drug interactions between the cancer drugs sunitinib and irinotecan.Clinical studies have shown that sunitinib and irinotecan have shown synergistic efficacy in the treatment of thyroid cancer.Previous studies have shown that the active metabolite SN-38 of irinotecan is metabolized by UDP-glucuronosyltransferase 1A1(UGT1A1),but the possible mechanism is unclear.In vitro experimental data showed that sunitinib showed weak inhibition of SN-38 glucoronidation in human liver microsomes and recombinant UGT1A1,with a K_i of 64.10?M and 74.49?M,respectively.This partly explains the possible mechanism for their synergistic efficacy in combined treatment.In short,the optimization of the traditional model makes the model prediction more accurate and can be used to predict the liver clearance.