PK/PD simulations as a tool for rational design of clinical dosage regimens: an example with Fradafiban.

ABSTRACT

In clinical studies, pharmacodynamic effects should be achieved, e.g. maintenance of certain effects with reversibly acting drugs. Available data base for early phase II studies is frequently kinetics in healthy volunteers from phase I studies and pharmacodynamic effects from in vivo or ex vivo data. Using the fibrinogen receptor antagonist Fradafiban as an example, a procedure to achieve rational dosage regimens is described. Applied methods were curve fitting of plasma concentrations obtained in phase I studies, correlation of fibrinogen receptor occupancy (FRO) to plasma concentrations using a sigmoid Emax model with Hill coefficient for PK/PD correlation, simulation of time course of FRO for various dosage regimens, using estimates for variability from PK and PD data in order to estimate not only mean values but also expected range of FRO. In a phase II study with Fradafiban administered intravenously, therapeutically active plasma levels had to be achieved rapidly and maintained over 24 hours employing a simple infusion regimen in 20 patients and 3 dose groups. For the target dose, a FRO of 80% should be achieved in most patients. Using the tools mentioned above, a rapid initial infusion rate of 10 mg over 30 minutes, followed by a maintenance infusion of 30 mg for the remaining 23.5 hours for the target dose resulted in a median predicted FRO of 82%. For the lower dose, a 5/15 mg infusion over 0.5/23.5 hours, achieving a predicted FRO of 68% was used and the upper dose (15/45 mg) resulted in 87% FRO. Median experimental results were 84% FRO for the target dose and 73% and 88%, respectively, for the lower and upper dose. As these results fit reasonably well to the predictions, it can be concluded that kinetics in the mostly elderly patients is similar to kinetics in healthy young volunteers. Furthermore, FRO in patients is nearly identical to that in spiked human plasma. Taken together, it could be proven that PK/PD methods are a useful tool for design of clinical studies of Fradafiban.