PBPK Modelling of Antisense Oligonucleotide Therapeutics: Application for Predicting Plasma and Tissue...

Antisense oligonucleotide (ASO) therapeutics present new opportunities for treating challenging-to-treat diseases. Prediction of concentration-time course in systemic circulation and tissues within the context of large molecule physiologically based pharmacokinetic (PBPK) modelling has proven to be useful in animal-to-human extrapolation and first-in-human dose selection. A human PBPK model was developed and verified using bepirovirsen clinical data to predict plasma pharmacokinetics (PK) and tissue concentrations. Liver and kidney partition coefficient ratios from monkey studies, corrected for plasma unbound fraction in monkeys and humans, informed bepirovirsen concentration predictions in the human liver and kidney. Liver and kidney partition coefficients were calculated to be 2147 and 2822, respectively. All predicted PK parameters in healthy volunteers (except t_max) were within two-folds of observed data. Predicted vs. observed clearance (L/h), AUC_0-inf (µg.h/mL) and C_max (µg/mL) for the 300 mg single dose in healthy volunteers were 2.94, 114.43, and 10.56 vs. 2.25, 136.6, and 6.5, respectively. The majority of observed plasma concentrations for all doses were within the 5th and 95th percentiles of the predictions. The evaluated model was used to predict the impact of moderate hepatic impairment on bepirovirsen PK in virtual patients. The predicted vs. observed bepirovirsen exposure in moderate hepatic impairment was 0.9 and 0.7-fold lower, respectively, compared with healthy volunteers. The PBPK model predicted the liver and kidney tissue C_max values to be 172 and 132 µg/mL, respectively. In conclusion, a PBPK modelling approach for bepirovirsen, an ASO, is presented in this article and offers opportunities for future applications to other oligonucleotide therapeutics.