Clozapine is a second generation antipsychotic drug used in psychiatry to treat schizophrenia, affective disorders or certain symptoms of dementia. In neurology, clozapine is frequently used and recommended to manage symptoms of psychosis associated with Parkinson's disease (PD). The risk of neutropenia or agranulocytosis associated with clozapine estimated at 1.3% is well known to doctors around the world with a peak at one month and a decrease in risk after more than a year of treatment. This risk has led to the policy of no blood, no drugs and monitoring of the complete blood count (CBC) weekly for 18 weeks and then monthly for the duration of treatment. Some studies suggest an increased risk of infections related to immunodeficiency induced by clozapine itself. This clozapine-induced immunodeficiency would be comparable to that encountered in patients with common variable immunodeficiency or under immunosuppressive treatment. In addition, this immunosuppressive effect linked to clozapine would not be dose dependent but time dependent. However, the only studies currently performed have been in psychiatric patients treated for schizophrenia. It seems important to specifically explore clozapine-related immunodeficiency in PD patients treated with clozapine for PD-related psychosis. In this study, the investigators propose to evaluate the variations in serum immunoglobulin levels and lymphocyte subpopulations (B, T, NK) in parkinsonian patients treated with Clozapine at 6 months and 1 year after initiation of treatment.