Safe and efficient methods for introducing clozapine to patients with treatment-resistant schizophrenia (TRS) are needed. We investigated risk factors for clozapine discontinuation in the early phase of its introduction.
We conducted a nested case-control study at 14 psychiatric hospitals in Chiba, Japan. Data from pre-registered TRS patients were collected at 7 time points within 12 weeks before and after the start of clozapine introduction. We examined the demographic data, prior and concomitant psychotropic drugs, strategies for switching from prior antipsychotics, and blood test and Global Assessment of Function results. The Clinical Global Impression-Severity Scale was retrospectively scored at 12 weeks before and after clozapine introduction.
Of 228 patients, clozapine treatment was continued in 213 (93.4 %) and discontinued in 15 (6.6 %) patients within 12 weeks. Clinical symptoms were improved to mild symptoms with a response rate of 14.9 %. Prior antipsychotics and concomitant psychotropic drugs except for mood stabilizers were significantly decreased. Histories of smoking (OR = 3.32, 95 %CI: 1.11-9.93) and antipsychotic treatment at chlorpromazine-equivalent doses <1200 mg within the past 5 years (OR = 3.93, 95 %CI: 1.24-12.50), but not antipsychotic switching strategy, were associated with clozapine discontinuation. Eosinophilia was the most frequent reason for discontinuation (n = 3, 20 %) and was associated with concomitant valproate at 4 weeks after the introduction.
Clozapine is an effective option for TRS patients (especially those treated with higher doses of prior antipsychotics) in Japan. Clinicians should be cautious about concomitant valproate in the early phase of clozapine introduction due to a high risk of eosinophilia.

Copyright © 2021 Elsevier B.V. All rights reserved.

Author