To the Editor: Mirjam Talvik, M.D., and colleagues (1) concluded that neither haloperidol nor clozapine preferentially occupies extrastriatal dopamine D₂ receptors. In their study, haloperidol (N=3) produced a mean occupancy of 80.7% in the striatum (with the method of simplified reference regions), 63.0% in the temporal cortex, and 60.0% in the frontal cortex. Clozapine (N=4) had a mean occupancy of 43.3% in the striatum, 40.5% in the temporal cortex, and 44.5% in the frontal cortex.

Although the use of statistical tests in studies with such small groups is problematic, we used a median test and the authors’ data to test whether the null hypothesis is supported in comparing D₂ receptor occupancy in cortical regions relative to the striatum for both drugs. In comparing temporal and frontal cortical to striatal occupancy in the haloperidol-treated subjects, a p value of 0.025 for each cortical region was obtained, i.e., significantly lower levels of occupancy by haloperidol of cortical versus striatal D₂ receptors, indicating preferential cortical control occupancy by haloperidol. With clozapine, no significant regional differences were seen (p=1.00). Thus, the null hypothesis with regard to haloperidol is not supported by the published results.

In contrast to other studies (2, 3), the study by Dr. Talvik et al. (1) suggested lower occupancy of extrastriatal than striatal D₂ receptors by haloperidol, rather than greater occupancy of extrastriatal than striatal D₂ receptors by atypical antipsychotic drugs. This discrepancy may be related to the use by Dr. Talvik et al. of the reference-region method with [¹¹C]FLB 457 to determine occupancy. Modeling studies of [¹¹C]FLB 457 in nonhuman primates (4) have shown that 77% of cerebellar uptake is due to specific binding. Xiberas et al. (3) reported a mean decrease of 12.5% in the cerebellar uptake of [⁷⁶Br]FLB 457 after a broad range of amisulpride doses, indicating the presence of D₂-type receptors in the cerebellum. These cerebellar D₂ receptors would have biased the results obtained by Dr. Talvick et al. (1), who used the reference-region method with [¹¹C]FLB 457 as a ligand to measure extrastriatal D₂ receptors.

A second possible reason for the discrepancy between the findings of Dr. Talvick et al. (1) and other studies (2, 3) is the use by Dr. Talvick et al. of two tracers, [¹¹C]raclopride and [¹¹C]FLB 457, to measure striatal and extrastriatal receptor occupancy, respectively. Xiberas et al. (3), on the other hand, used the preferred method of a single radio ligand, [⁷⁶Br]FLB 457, to measure both striatal and extrastriatal receptor occupancy. Dr. Talvik et al. (1) needed to demonstrate that [¹¹C]FLB 457 and [¹¹C]raclopride would produce nonsignificantly different receptor occupancies in the same region to validate their method. Further studies of D₂ receptor occupancy (5) in striatal and extrastriatal regions are in progress in our laboratory with a single radioligand to validate that method. Our preliminary results do not support the finding of Dr. Talvick et al. with regard to clozapine producing uniform occupancy across striatal and extrastriatal D₂ receptors.