Am J Psychiatry 154:1, January 1997 BUSATTO, PILOWSKY, COSTA, ET AL. BENZODIAZEPINE RECEPTOR BINDING AND SCHIZOPHRENIA Correlation Between Reduced in Vivo Benzodiazepine Receptor Binding and Severity of Psychotic Symptoms in Schizophrenia Geraldo F. Busatto, M.D., Ph.D., Lyn S. Pilowsky, Ph.D., M.R.C.Psych., Durval C. Costa, Ph.D., Peter J. Ell, M.D., F.R.C.P., Anthony S. David, M.D., M.R.C.Psych., James V. Lucey, M.D., M.R.C.Psych., and Robert W. Kerwin, Ph.D., M.R.C.Psych. Objective: Although there is evidence from postmortem studies suggestive of deficient in- hibitory neurotransmission of γ-aminobutyric acid (GABA) in schizophrenia, no direct in vivo evidence has been obtained to date. The authors used single photon emission computed to- mography (SPECT) with iodine-123-labeled iomazenil ([ 123 I]iomazenil), a radioligand that selectively binds with high affinity to the benzodiazepine subunit of the GABA A receptor complex in the human brain, to investigate the presence of benzodiazepine receptor abnor- malities in the cerebral cortex of living subjects with schizophrenia. Method: Dynamic [ 123 I]iomazenil SPECT was performed in 15 patients (14 patients with DSM-III-R schizophre- nia and one with schizophreniform disorder) and 12 healthy subjects over a period of 2 hours. The time-integral method was used to generate ratios of “specific” to “nonspecific” [ 123 I]iomazenil binding at equilibrium for several cortical regions. Results: N o overall between- group differences in benzodiazepine receptor binding were found, but significant correlations emerged between the severity of schizophrenic symptoms and [ 123 I]iomazenil binding in limbic cortical regions: positive symptom scores were negatively correlated with benzodiazepine re- ceptor binding in the left medial temporal region, and negative symptoms were inversely related to receptor binding in the medial frontal region. These correlations were not significant when a Bonferroni correction for multiple comparisons was applied. Conclusions: These preliminary results are consistent with previous research implicating limbic cortical regions in the patho- physiology of schizophrenia, suggesting that reduced inhibitory GABAergic tone in these areas may contribute to the appearance of schizophrenic symptoms. (Am J Psychiatry 1997; 154:56–63) T he efficacy of classical antipsychotic drugs is me- diated by dopamine D 2 receptor blockade (1). This has provided the main support to the “dopamine hypothesis” for schizophrenia, which postulates a role for excessive dopaminergic activity in the genesis of positive schizophrenic symptoms (2). However, many patients whose schizophrenia is refractory to typical antipsychotics may respond to atypical agents that only weakly block D 2 receptors (3). This has prompted in- vestigators to search for alternative neurochemical mechanisms that may also contribute to the appearance of schizophrenic symptoms. One such mechanism may be disturbed activity of the inhibitory neurotransmitter γ-aminobutyric acid (GABA). GABA is the major transmitter of interneurons in cortical areas thought to be structurally affected in schizophrenia (4), such as the medial temporal region (5) and the anterior cingulate gyrus (6). Therefore, reduced GABA transmission may be a neurochemical correlate of anatomical changes in schizophrenia (7). Moreover, bio- chemical (8) and pharmacological (9, 10) studies in ani- mals have suggested a modulatory role for GABA in do- paminergic transmission in the brain. Such interactions have led to a “GABA deficit hypothesis” for schizophre- nia, which proposes that failure of GABAergic inhibition Presented in part at the Vth International Congress on Schizophre- nia Research, Warmsprings, Va., April 6–12, 1995. Received June 21, 1995; revisions received Oct. 23, 1995, and April 19 and June 17, 1996; accepted July 1, 1996. From the Department of Psychological Medicine, Institute of Psychiatry. Address reprint requests to Dr. Ker- win, Section of Clinical Neuropharmacology, Department of Psycho- logical Medicine, Institute of Psychiatry, De Crespigny Park, Den- mark Hill, London SE5 8AF, U.K. Supported by a Wellcome Trust Research Grant; Dr. Busatto is funded by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Brazil; Dr. Pilowsky and Dr. Lucey were Well- come Trust Research Training Fellows. The authors thank Drs. G. Dunn and P. Sham of the Biometrics Department, Institute of Psychiatry, for statistical advice. 56 Am J Psychiatry 154:1, January 1997