Pharmacokinetics-Guided Dose Allocation in Comparison with Binary Dose Spacing

In designing dose-ranging studies at early Phase II, one of the major challenges is dose allocation or dose spacing. Binary dose spacing (BDS) has been established over the years as a practical and useful way of allocating doses. In this manuscript, the pharmacokinetics (PK)-guided dose allocation, pharmacokinetics/pharmacodynamics (PK/PD)-guided dose allocations are compared with BDS. Additionally, a mixed dose allocation strategy is proposed and also compared. Simulations are performed to study various properties of these four dose allocation methods. The comparisons include 1. power of the PoC test; 2. bias on target dose estimation; 3. root mean square error (RMSE); 4. probability that the estimated dose delivers > 75% of true effect; and 5. slopes and confidence intervals of slopes. Results indicating that BDS performs reasonably well under most scenarios. The PK-guided approach performs similar to BDS, the PK/PD-guided approach performs better, if the PD information is reliable. However, if the PD information is not highly reliable, the mixed procedure could be considered.