Many individuals with schizophrenia abuse cannabis at the onset of their illness, portending a poorer course of illness and poorer treatment response. Preliminary evidence suggests that clozapine may uniquely reduce substance use in patients with schizophrenia. The purpose of this study is to establish an effective methodology for studying early treatment with clozapine in patients with co-occurring schizophrenia and cannabis use disorder, while generating pilot data comparing clozapine vs. risperidone on substance use, psychiatric symptoms, side effects, and treatment discontinuation.

Schizophrenia is a relapsing disorder that produces profound effects on those afflicted once it becomes chronic, often leading to a severe and long-term disability. However, during the initial years of illness many patients are more treatment responsive and may achieve substantial periods of remission. Comorbid substance use disorder, a common concomitant of schizophrenia, is associated with increases in morbidity and mortality. Within the early phases of schizophrenia, comorbid cannabis use disorder (CUD), the most common comorbid disorder (approximately 50%), appears to be linked to a poor outcome in these patients, and may be a factor in their long-term deterioration. While data indicate that first episode patients presenting with comorbid CUD have an earlier onset of illness and a poorer outcome than those without CUD, continued cannabis use after antipsychotic treatment, which occurs in approximately 50% of these patients in the initial months after hospitalization, is associated with an even worse outcome. The apparent "toxicity" of cannabis use in first episode patients is paralleled by data from chronic patients, in whom CUD is associated with clinical exacerbations, non-compliance with treatment, poor global functioning, and increased relapse. A growing body of data suggests that a critical period exists in patients with schizophrenia during the early phases of psychosis in which symptoms and functioning continue to worsen, and that treatment with antipsychotic medications during this period may improve the natural course of the disorder. While the availability of novel antipsychotic medications has sparked further research in the early phases of schizophrenia, there have been few studies including clozapine (CLOZ). We hypothesize that CLOZ may be more effective than other novel agents in controlling cannabis use in patients with first episode schizophrenia who are comorbid for CUD. We postulate that standard antipsychotic medications do not decrease substance use in this population largely because they do not restore normal functioning of the dysfunctional dopamine (DA) pathways. CLOZ, by contrast, through its varied actions on dopaminergic, serotonergic and, particularly, noradrenergic neurons, coupled with its weak D2 and potent noradrenergic alpha2 blocking ability, may tend to have a "normalizing" effect on the signal detection capability of these dysfunctional DA systems. CLOZ is rarely used in first episode patients even though it is generally considered a highly effective medication, and more likely to prevent relapses than typical antipsychotics in both treatment refractory and non-refractory populations. While side effects of CLOZ are a clinical concern and need to be taken seriously, experience over the past 2 decades has increased the comfort level for its use. For example, agranulocytosis has actually occurred in 0.37% with implementation of the required white count monitoring system, and granulocyte stimulating therapy has provided an effective treatment for those few patients who do develop agranulocytosis. Gradual dose titration schedules appear to reduce risk for seizure and myocarditis, and experience in first episode patients suggesting that the clinically effective dose of CLOZ is lower than in chronic patients indicates that these risk should be even lower. If our hypothesis is correct, CLOZ, despite its side effect profile, may have a key therapeutic role in these patients, a role with important public health implications. Yet without a clear demonstration of the benefit/risk and benefit/burden profiles of its use, the naturalistic experiment of assessing the overall effectiveness of CLOZ in first episode patients who are comorbid for CUD is unlikely to be undertaken.

This study is enrolling individuals who are in their first episode of schizophrenia or schizoaffective disorder and who are currently using cannabis. Study participants undergo a screening visit including the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) for diagnostic evaluation, and a physical examination, blood tests, and review of medical, psychiatric, and substance use history. Following the screening visit, eligible participants are randomly assigned to single-blind treatment with clozapine or risperidone for 24 weeks. Participants assigned to clozapine initially receive a daily dose of 12.5 mg, which is carefully titrated to the lowest dose necessary to manage psychotic symptoms and well tolerated. Participants assigned to risperidone initially receive 0.5 mg daily and are carefully titrated in the same fashion. Cross titration off of the previous antipsychotic (if any) and onto study medication is completed within 2-4 weeks. Participants on clozapine have weekly blood tests. Concomitant medications are minimized and kept stable throughout the protocol except as needed to manage side effects or urgent clinical symptoms. Study medication, psychiatric visits, assessments and labwork are all provided without charge to participants. Study visits take place once a week. At study visits, medication side effects, physical and psychiatric symptoms, substance use, and treatment services received are assessed. A Lifestyle Intervention is also provided to help prevent metabolic side-effects in this vulnerable population, and assist with recovery. The investigators meet weekly to review clinical care and manage any variations in study protocol.

Inclusion Criteria:

• Age 17 - 45
• Meets Diagnostic and Statistical Manual for Mental Disorders (DSM-IV) criteria for schizophrenia or schizoaffective disorder
• Currently within the first episode of schizophrenia ("First Episode" is defined as having onset with the first evidence of psychotic symptoms by history, and the first episode will have ended if the Clinical Global Impressions Scale (CGI) has been < 2 and there has been no rating > 2 on any one of the Brief Psychiatric Rating Scale (BPRS) psychotic items for 6 weeks or longer)
• Meets DSM-IV criteria for cannabis use disorder
• Cannabis use within the five weeks prior to recruitment (screening visit or hospital admission) by self-report (TLFB), collateral report, or drug screen.
• Requires treatment with an antipsychotic medication
• Patients (or guardians) must provide informed consent prior to entry into the study

Exclusion Criteria:

• Medical contraindications to treatment with clozapine or risperidone, including previous paralytic ileus.
• Cumulative treatment with antipsychotic medication in excess of 16 weeks prior to hospital admission (or case identification if an outpatient), unless waived by the Medication Adjustment Group (MAG)
• History of allergic reaction to clozapine or risperidone
• History of seizure disorder or blood dyscrasia. Note: If patients have a history of seizures, but not a diagnosed seizure disorder, they may be admitted to the study if approved by the MAG.
• Current treatment with clozapine
• Currently pregnant, planning to become pregnant, or unwilling to use an acceptable form of birth control.
• Currently residing in a residential program designed to treat substance use disorders.
• Treatment at baseline with a psychotropic agent proposed to curtail substance use (e.g. disulfiram, naltrexone, valproic acid, topiramate, acamprosate or benzodiazepines) will require a review by the medication adjustment group before entering the client into the study
• Patients who, in the opinion of the investigator, are judged unsuitable to participate in the study (For example, patients who are actively homicidal or have a pending incarceration that would prevent them from participating in the study)
• History of, or current breast cancer
• People who are doing well on current therapy
• Lack of an identifiable primary family/support person, and unable to come to a study site for weekly visits
• Treatment with serotonin re-uptake inhibitors will not be excluded but requires a review by the MAG prior to randomization.
• Patients with current cocaine dependence will require review by the MAG to determine stability for the study.
• Treatment with multiple antipsychotics or long acting injectable antipsychotic at baseline is not excluded, but will be reviewed by the MAG to assess appropriateness for the study.

For the 24 weeks of the study plus 30 days after the final visit.