To the Editors Clozapine is a second-generation antipsychotic that is known for its high effectiveness in treatment-resistant cases of psychosis not responding to other second-generation antipsychotics or first-generation antipsychotics (FGAs).1 However, because of a number of adverse effects such as cardiotoxicity,2 neutropenia, and binge eating associated with severe weight gain,3 it is not used as first-line treatment. It is recommended only after 2 failed adequate antipsychotic trials.4 Flupentixol is an FGA that is still actively used in various countries including Germany, the United Kingdom, and Sweden.5 It is valued for being as effective as other FGAs5 but less likely to cause extrapyramidal motor side effects.6 We present here a patient with schizophrenia who first received a combination of clozapine and flupentixol. After discontinuation of flupentixol, clozapine plasma concentration abruptly decreased, leading to a severe exacerbation of psychotic symptoms and a suicide attempt. This pattern indicates a strong and clinically relevant pharmacokinetic interaction between these 2 antipsychotics. Mr A. is a 30-year-old man with an 8-year history of schizophrenia. Except for tobacco dependence (25–30 cigarettes per day), he had no psychiatric comorbidities. At admission to the ward, he reported auditory hallucinations, delusions of guilt and reference, and deficits in concentration. He showed severe formal thought disorders, emotional distress, and psychomotor agitation. Because of treatment resistance under various combinations of 4 antipsychotics (quetiapine, amisulpride, olanzapine, flupentixol), the patient agreed to treatment with clozapine. He was receiving a combination of olanzapine, amisulpride, and flupentixol at the time. We first abruptly discontinued olanzapine and tapered out and stopped amisulpride while increasing flupentixol to 12 mg/d. Simultaneously, clozapine (25 mg/d) was started and then stepwise increased to 400 mg/d over a period of 4 weeks. Hypersalivation was treated with pirenzepine (150 mg/d). Blood tests controlling liver function, white blood cell counts, and CK and myoglobin, as well as electrocardiograms controlling QTc time, were performed once weekly. Also, plasma levels of clozapine (therapeutic range, 350–600 μg/L7) and of its active metabolite desmethylclozapine (therapeutic range, 100–300 μg/L) were regularly determined (Fig. 1). At day 30 after starting clozapine (Fig. 1), the patient showed a significant improvement of psychopathology: psychomotor agitation and emotional distress decreased, the formal thought disorders and delusions ceased, and the auditory hallucinations widely disappeared. We then started tapering out the flupentixol over a period of 3 weeks. With flupentixol 2 mg/d (days 48–51) and clozapine 450 mg/d, plasma levels of clozapine (331 μg/L) and desmethylclozapine (212 μg/L) were (almost) in the therapeutic range, and the psychopathological status was still stable. However, 3 days after the complete discontinuation of flupentixol (day 58), the plasma concentrations of clozapine abruptly sank to a subtherapeutic level of 173 μg/L. The number of cigarettes smoked by the patient as well as his caffeine intake had not changed in this period, and medication adherence was monitored by the nursing staff. This decrease of clozapine plasma concentration was associated with a relapse of psychotic symptoms in the form of severe delusions of guilt, auditory hallucinations, psychomotor agitation, and emotional distress. Symptoms had dramatically worsened within only 2 days leading up to the patient attempting suicide by trying to drown himself in a river while wearing a backpack filled with stones. Fortunately, Mr A. could be saved from drowning, and he suffered no injuries and only mild hypothermia. Not being aware of a drug interaction, we assumed a lacking adherence. We therefore continued the medication with 450 mg/d clozapine but monitored medication intake even more precisely.