A Promising New Approach for In Silico Prediction of Drug Concentration
Profiles for Drug Candidates Lack of Experimental Pharmacokinetic Data
Abstract
Aim: The purpose of this study is to develop a novel protocol to predict
the concentration profiles of a target drug based on the PBPK model of a
structurally similar template drug by combining two software for PBPK
modeling, the SimCYP simulator and ADMET Predictor. Methods: The method
was evaluated by utilizing 13 drug pairs which come from 18 drugs in the
built-in database of the SimCYP software. All drug pairs have their
Tanimoto scores no less than 0.5. Three versions (V1, V2 and V3) of
models for the target drug were constructed by gradually replacing the
corresponding parameters of the template drug with those predicted by
ADME Predictor for the target drug. Normalized RMSE and Wilcoxon
rank-sum test were introduced for the evaluation of the model
performance. Results: Overall, V3 models demonstrated better performance
than the V1 and V2 models did. The relationship between the model
performance and structural similarity of drug pairs was also explored.
Three protocols have come out as guidance on how to build PBPK models
for target drugs: (1) V1 models are recommended when the structural
similarity is very high; (2) V2 models are recommended when the
similarity is below 0.65 or high than 0.85; (3) V3 models are
recommended when the similarity is below 0.85. Conclusion: By leveraging
the prediction accuracy and application practicality, this novel
approach has a great promise in predicting the preliminary PK profiles
for novel drugs, propelling the drug discovery process by suggesting
drug candidates with promising PK profiles.