The pharmacokinetics of a liposomal sub-unit antigen vaccine system composed of the cationic lipid dimethyldioctadecylammonium bromide (DDA) and the immunostimulatory agent trehalose 6,6-dibehenate (TDB) (8:1 molar ratio) combined with the Ag85B-ESAT-6 (H1) antigen were modelled using mouse in vivo data. Compartment modelling and physiologically-based pharmacokinetics (PBPK) were used to predict the administration site (muscle) and target site (lymph) temporal concentration profiles and factors governing these. Initial estimates using compartmental modelling established that quadricep pharmacokinetics for the liposome demonstrated a long half-life (22.6 days) compared to the associated antigen (2.62 days). A mouse minimal-PBPK model was developed and successfully predicted quadricep liposome and antigen pharmacokinetics. Predictions for the popliteal lymph node (PLN) aligned well at earlier time points. A local sensitivity analysis highlighted that the predicted AUCmuscle was sensitive to the antigen degradation constant kdeg (resulting in a 3-log change) more so than the fraction escaping the quadriceps (fe) (resulting in a 10-fold change), and the predicted AUCpln was highly sensitive to fe. A global sensitively analysis of the antigen in the muscle demonstrated that model predictions were within the 50th percentile for predictions and shown acceptable fits. To further translate in-vitro data previously generated by our group, the mouse minimal-PBPK model was extrapolated to humans and predictions made for antigen pharmacokinetics in muscle and PLN. Global analysis demonstrated that both kdeg and fe had a minimal impact on the resulting simulations in the muscle but a greater impact in the PLN. In summary, this study has predicted the in-vivo fate of DDA:TDB:H1 in humans and demonstrated the role of formulation degradation and fraction escaping the depot site can play in impacting upon the overall depot effect within the site of administration.
Details on the figures are available from the ReadMe file provided.