We propose an eight-week, double-blind, placebo-controlled trial of rosiglitazone in schizophrenia subjects treated with clozapine using Bergman's Minimal Model (MINMOD) intravenous glucose tolerance test. Bergman's Minimal Model analysis with frequent sampled intravenous glucose tolerance test (FSIVGTT) provides a sensitive and reliable method to measure glucose effectiveness, insulin secretion and insulin sensitivity. The MINMOD determines the relationship between insulin sensitivity, insulin secretion and the degree of obesity and can be used to study drug effects upon these variables.

This study is an eight-week, double-blind, placebo-controlled trial of rosiglitazone in schizophrenia subjects treated with clozapine using Bergman's Minimal Model Analysis (MINMOD) frequent sampled intravenous glucose tolerance test (FSIVGTT) for examination of glucose metabolism. Rosiglitazone, used for the treatment of type 2 diabetes, improves glycemic control by improving insulin sensitivity in target tissues while increase glucose uptake and improving signaling between insulin and glucose transporters. The MINMOD determines the relationship between insulin sensitivity, insulin secretion and the degree of obesity and can be used to study drug effects upon these variables. The computer program estimates glucose effectiveness (SG), insulin sensitivity (SI), and the acute insulin response to glucose (AIRG) from glucose and insulin values measured during the FSIVGTT.

The site for subject enrollment, clinical assessment and collection of blood samples at weeks 2, 4, and 6 will be at the Freedom Trail Clinic of the Erich Lindemann Mental Health Center. The FSIVGTT and blood samples at baseline and week 8 will be conducted at the Mallinckrodt General Clinical Research Center at Massachusetts General Hospital. Subjects will include 50 clozapine treated patients with the diagnosis of schizophrenia or schizoaffective disorder. Subjects will be matched according to baseline fasting glucose.

Subjects will be maintained on their pre-study medication. During the baseline assessment, a member of the research team will conduct the PANSS (Positive and Negative Symptom Subscale) and HAM-D (Hamilton Rating Scale for Depression) which will be repeated at weeks 4 and 8 to determine baseline and subsequent changes in psychopathology. The neuropsychological battery will be administered at baseline and week 8 and will consist of: North America Adult Reading Test, Trail Making, Degraded Stimulus Continuous Performance Test, California Verbal Learning Test, Faces and Family Picture subtests from WMS-III, Wisconsin Card Sorting Test, Letter and Category Fluency, Letter-Number Sequencing, and Grooved Peg Board. Baseline medical evaluation will include weight, height, vital signs, AIMS score, a 12-lead EKG and a physical examination.

Subjects will be instructed to eat a diet high in carbohydrates for three days before the FSIVGTT. They are instructed to fast for 12 hours preceding the test. On the morning of the test, subjects will be admitted to the General Clinical Research Center at Massachusetts General Hospital. A nutritional assessment will be performed. Baseline blood samples for fasting glucose, basic chemistry profiles, liver enzymes, complete blood count (cbc), insulin, prolactin, lipid profile, glycohemoglobin, leptin, clozapine serum levels and cortisol will be drawn. Glucose 0.3 g/kg body weight will be infused. Approximately 35 2 cc blood samples will be withdrawn for measurement of glucose and insulin concentration. At minute 20, 0.05 units/kg of insulin will be administered as an intravenous bolus injection.

Following baseline FSIVGTT, subjects will be randomized, double-blind, to either rosiglitazone or placebo. Subjects randomized to receive medication will receive rosiglitazone 4mg po qd for eight weeks or until study end. Subjects will be monitored for symptoms of diabetes at weeks 2, 4, and 8. Subjects will be evaluated at weeks 2, 4, and 8 including measurements of blood pressure, pulse and weight. Adverse events will be monitored at weeks 2, 4, and 8 with the AMDP-5, a survey of side effects and somatic symptoms. The Medical Outcomes Study Short Form-12 (SF-12) will be used to measure general health status and function. The SF-12 will be performed at baseline, weeks 4, and 8 or study end.

A cbc will be performed every two weeks, consistent with the requirements for treatment with clozapine. At week four, a fasting glucose, insulin level, and liver function tests will be performed. At the end of the eighth week, the FSIVGTT will be repeated and blood samples will be obtained for fasting glucose, basic chemistry profiles, liver enzymes, complete blood count (cbc), insulin, prolactin, lipid profile, glycohemoglobin, leptin, clozapine serum levels and cortisol. All other aspects of the subjects care will be treatment as usual.

Potential risks associated with an intravenous line include bruising at the site, bleeding and the risk of infection. During the procedure patient may experience symptoms of low blood glucose including lightheadedness, dizziness, weakness, irritability, sweating, tremor, tachycardia, anxiety, and hunger. The amount of blood drawn with each FSIVGTT is approximately 150 cc. A low risk of anemia exists secondary to the FSIVGTT; anemia has also been a side effect of treatment with rosiglitazone. Treatment with rosiglitazone has been associated with side effects including accidental injury (cuts and abrasions), headache, back pain, anemia, edema, weight gain, decreased white blood cell count, increases in total cholesterol, LDL, and HDL and decreases in free fatty acids. Phlebotomy may produce discomfort, bruising, and rarely, infection. While the clinical assessment may be stressful, the questions contained in the SCID and PANSS do not differ from those asked in routine clinical evaluations.

Potential benefits to the subject include a comprehensive diagnostic re-evaluation, screening blood tests, and estimation of risk of diabetes. The results of the tests will be shared with patients and clinicians and may contribute to medical management by helping the patient's physician in deciding if an intervention is required to reduce the risk of diabetes. An increased understanding of the process by which clozapine induces glucose intolerance and the awareness that diabetes and its complications can be prevented with concomitant treatment with an insulin sensitizer are potential benefits to this population.

Inclusion Criteria:

• Male or female
• Age 18-65 years
• Diagnosis of schizophrenia, any subtype, schizoaffective disorder, any subtype or schizophreniform disorder
• Well established compliance with out-patient medications
• Current treatment with clozapine for a minimum of one year
• Evidence of insulin resistance: impaired fasting glucose (glucose ≥100 mg/dl) or hyperinsulinemia (fasting insulin ≥ 15 ng/dl) or a HOMA-IR (homeostasis model assessment for insulin resistance) (fasting glucose X fasting insulin/22.5) ≥2 or a SI (insulin sensitivity index)

Exclusion Criteria:

• Inability to provide informed consent
• Current substance abuse
• Significant medical illness, including congestive heart failure, severe cardiovascular disease, renal disease (serum creatinine > 1.5), anemia (Hemoglobin < 11.0 gm/dL) or psychiatrically unstable
• Severe hepatic impairment, active liver disease or increased serum transaminase levels (ALT>2.0X upper limit of normal) If at any time, ALT increases to 2X ULN, the subject's participation in the study will be terminated.
• Women of child bearing potential who are pregnant, breastfeeding, or who are unwilling or unable to use an effective form of birth control during the entire study
• Treatment with agents that induce weight loss
• History of diabetes mellitus or thyroid disease
• Current treatment with an oral hypoglycemic agent or insulin
• Known hypersensitivity to rosiglitazone or any of its components
• Fasting Glucose >126 mg/dL11. Treatment with other atypical antipsychotic agents thought to impair glucose metabolism (olanzapine) or low potency conventional agents (thioridazine, chlorpromazine)

Baseline to week 8 (entire study)

Our findings in this study were limited by the small sample size. It is possible that with a larger sample, significant improvements might be observed not only in glucose metabolism but also in lipid metabolism.