Simulation of PET scan timings for receptor occupancy studies of CNS drugs: a simple fixed-time design performed as well as scattered time point designs.

PURPOSE

This study aimed to determine the effect of PET scan timings on the reliability of occupancy parameter estimates and to identify the scan timing design that gives the most reliable occupancy parameter estimates.

METHODS

We compared the performance of designs with various sets of sampling time points using the stochastic simulation and estimation method in Perl-speaks-NONMEM. Biases, relative standard errors, relative estimation errors, and root mean square errors were used to compare the performance of designs.

RESULTS

Unlike the results of a previous report, we found that rather complicated designs where each subject or group of subjects are allocated to different scan timings were not superior to the simple, conventional fixed-time designs regardless of whether effect compartment or receptor binding models were used.

CONCLUSIONS

We conclude that the conventional fixed-time designs that have been used so far may give robust PD parameter estimates for occupancy data obtained from human PET studies of CNS drugs.