Dr. Talvik and Colleagues Reply

To the Editor: Drs. Pilowsky and Ell claim that clozapine treatment induces regionally selective occupancy of D₂ receptors in vivo. In our study we did not find support for regional selectivity. We agree with Drs. Pilowsky and Ell that this issue needs to be clarified.

Using the high-affinity PET radioligand [¹¹C]FLB 457, a close analog of the SPECT ligand [¹²³I]epidepride, we were initially puzzled by images that gave the impression of higher occupancy in extrastriatal regions. This unexpected observation was made for patients treated with clozapine and patients treated with low doses of haloperidol and seemed to be confirmed when ratio analyses were applied for the calculation of D₂ occupancy. However, this finding was not confirmed when we used [¹¹C]raclopride for striatal binding. We recently conducted a study that suggests methods limitations as a possible reason for the discrepant results (1).

When ratio approaches are used to obtain reliable estimates for binding potential, the radioactivity must reach equilibrium during the acquisition time. The problem is that the time to reach equilibrium is highly dependent on the density of D₂ receptors, which varies about 100-fold in the human brain. A consequence is that regions with low receptor density (i.e., the temporal cortex) will reach equilibrium long before regions with high receptor density (i.e., the striatum). As shown in simulations based on experimental data, the end-time method, a ratio method used in SPECT, overestimates the binding potential if the time to equilibrium is within the time of data acquisition but underestimates the binding potential if the time to equilibrium falls beyond the acquisition time (1). In SPECT studies using this method, low striatal and high extrastriatal binding potential have been reported (2, 3). In the reports of the PET studies referred to in the letter, the methods were not reported in detail, which makes it difficult to appraise the results (Meltzer et al., 1999; Xiberas et al., 1999). It cannot be discounted that these results were also influenced by the methods limitations of applying a ratio analysis on striatal binding at pre-equilibrium conditions.