compared with chlorpromazine in treatment-resistant schizo- phrenia. Am. J. Psychiatry 155, 914±920. Weiden, P., Aquila, R., Standard, J., 1996. Atypical antipsychotic drugs and long-term outcome in schizophrenia. J. Clin. Psychiatry 57, 53±60. Goldstein, J., 1999. Atypical antipsychotic drugs: beyond acute psychosis, new directions. Emerging Drugs 4, 127±151. John O. Brooks III Ph.D., M.D.) * Veterans Affairs Health Care System, 3801 Miranda Avenue, Ward 4B2, Palo Alto, CA 94304, USA E-mail address: johnbrks@stanford.edu PII: S0920-996400)00045-1 p Tel.: 11-650-493-5000, ext. 65182; fax: 11-650-858-3948. Reduced prefrontal cognitive activation associated with aggression in schizophrenia The etiology of aggressive behavior in schizo- phrenia is a heterogeneous mixture of social, psycho- logical, and biological factors Wirshing et al., 1994). In a recent review, Davidson et al. 2000) hypothe- sized that one of the causes of impulsive aggression could be dysfunction of the prefrontal cortex. As a preliminary study, we assessed the relationship between prefrontal function and aggression in 15 schizophrenics by using Single Photon Emission Tomography SPECT) and by evaluating the patients both at rest and during activation with the Wisconsin Card Sorting Test WCST). All patients were right handed with SCID-P-DSM- IV schizophrenia. Additional inclusion criteria were age between 18 and 65 years, no major medical or neurological illness and no additional psychiatric disorders. All subjects gave written consent after a full explanation of the study procedure and were injected with 925 MBq 25 mCi) of 99mTc-Bicisate and scanned after 45 min using a GE rotating gamma camera with a parallel-hole low-energy high-resolution collimator. Sixty-four planar images were acquired over 3608, with 30 s time resolution, and a 64 £ 64 matrix with a pixel size of 4.5 mm. A zoom factor of two was used. Images were pre®ltered with a 0.5-1, order 10, Butterworth ®lter, 0.12 cm corrected attenuation. The patients were randomly assigned either to a resting SPECT scan or to a WCST prefrontal activation SPECT scan. Regions of interest ROIs) of 2 £ 2 pixel 9 £ 9 mm 2 ) were manually placed on axial planes in the left and right superior, middle, and inferior regions of the prefrontal gyri. Indices of relative perfusion were used a ratio between the average perfusion of the ROIs included within the individual prefrontal gyri and the mean cerebellar perfusion in a 2 £ 2 pixel ROI of the same hemisphere). Evaluation of aggression was based on the behavior of the patient during the ®rst week of hospitalization and was derived from the recordings of the entire clinical staff. Three patients were classi®ed with aggressive behavior and 12 were not. The three patients with aggression were compared to the others by measurements of prefrontal rCBF at rest and during WCST activation using two multivari- ate analyses of covariance MANCOVAs) with age, chlorpromazine equivalents and duration of illness as the covariates. The difference between aggressive and nonaggressive patients on the prefrontal rCBF at rest was not signi®cant. On the contrary, there was a difference between the two groups on the prefrontal rCBF scores during WCST activation Wilks' Lambda  0:056; F  14:04; df  6; 5; p  0:0054: In particular, subsequent univariate ANCOVAs showed signi®cant differences on the rCBF of the right middle indices of relative rCBFs: aggressive  0:765 ^ 0:042 vs. nonaggressive  0:880 ^ 0:040 and right inferior indices of relative rCBFs: aggressive  0:809 ^ 0:036 vs. nonaggressive  0:912 ^ 0:027 prefrontal gyri F  5:68; df  1; p  0:038 and F  12:75; df  1; p  0:005; respectively). The ®nding of reduced prefrontal rCBF in aggres- sive subjects is in line with the results of previous functional brain studies Raine et al., 1994; Amen et al., 1996; Volkow et al., 1995; Davidson et al., 2000). How does selective prefrontal dysfunction relate to an increased predisposition toward aggression? Raine and Buchsbaum 1996) have hypothesized that reduced prefrontal functioning could result in a loss of inhibition normally exerted by the frontal cortex on phylogenetically older subcortical structures which are believed to play a role in facilitating aggression. Letters to the Editors / Schizophrenia Research 50 2001) 131±135 134