E lementary sensory processing in schizophrenia has not been well studied with functional neuroimaging techniques, in spite of behavioral evidence suggesting abnormal function in the first 500 msec of stimulus input (1, 2). Using functional magnetic resonance imaging (MRI), Renshaw et al. (3) found an enhanced hemodynamic response to visual stimulation in patients with schizophrenia. A follow-up study from their laboratory (4) suggested that this enhanced perfusion change could occur because of increased cerebral blood volume in the vasculature of the occipital cortex, which they observed by using dynamic susceptibility contrast MRI.To cross-validate the initial functional MRI finding, we studied a similar group of medicated patients with the [ 15 O]H 2 O positron emission tomography (PET) technique and a graded photic stimulus, which yields a "dose-response" relationship in the visual cortex between stimulation rate and blood flow (5). We also infused subjects with acetazolamide, which increases cerebral blood flow (CBF) by means of a nonneuronal mechanism (inhibition of carbonic anhydrase). We reasoned that if patients with schizophrenia have greater vascular volume, as suggested by Cohen et al. (4), then we would see a greater CBF increase during the acetazolamide infusion in patients with schizophrenia than in normal comparison subjects.
METHODWe recruited 11 stable patients with schizophrenia. Four were inpatients who were to be discharged, and seven were outpatients. Four were women. The patients' diagnoses were established by structured interview (6) and were based on DSM-IV criteria. All of the patients were taking psychotropic medications (nine were taking clozapine). Their mean age was 36.1 years (SD=7.5), and their mean duration of illness was 13.4 years (SD=6.8). Ten comparison subjects were matched to the patients in age (mean=35.8 years, SD=6.2) and sex (four of the comparison subjects were women). The comparison subjects had no history of psychiatric illness, according to structured interview (6), and none reported any first-degree relatives with an axis I psychiatric disorder. After a complete description of the study, written informed consent was obtained from all subjects.Subjects wore light-tight goggles that delivered a flashing visual stimulus by means of inset LED arrays at rates of 1, 3, 8, and 30 Hz. Subjects received a total of six scans; the four stimulation conditions and the one "null" condition (goggles on, no stimulus) were pseudo-randomized in the first five scans. Fifteen minutes before the sixth scan (a null stimulus), subjects received an intravenous infusion of 1 g of acetazolamide. Subjects were instructed to keep their eyes open for all scans.PET scans were performed with a Siemens CTI 931/08-12 camera (CTI Inc, Knoxville, Tenn.). For each emission scan, subjects received an intravenous bolus injection of 66 mCi of [ 15 O]H 2 O. Data were collected in a single 60-second frame beginning 5 seconds after the arrival of radioactivity in the brain. In seven normal subjects and fi...