Background: Clozapine is associated with the risk of serious neutropenia. However, this risk might decrease over time, meaning that indefinite absolute neutrophil count (ANC) monitoring could be unnecessary. We aimed to determine the epidemiology and timing of clozapine-associated neutropenia outcomes, to investigate variables that might contribute to the odds of neutropenia, and to determine risk of competing neutropenic events during clozapine treatment.
Methods: We performed a retrospective analysis of the Australian and New Zealand Viatris Pharmacovigilance system (one of two monitoring databases for these two countries) between June 6, 1990, and Oct 25, 2022. Patients were excluded from analysis if they commenced clozapine before 1990, did not have a haematology test within 2 weeks of commencement date, or had no follow-up. We measured minor neutropenia (ANC 1·0-1·5 × 109 per L) and serious neutropenia (ANC <1·0 × 109 per L) leading to cessation of clozapine within 6 weeks of the neutropenic event. We determined the rates of minor and serious neutropenia and calculated odds ratios (ORs) for the likelihood of neutropenia leading to cessation. For serious neutropenia leading to cessation, we used time-to-event to calculate rolling weekly averages and to perform competing risk analysis of outcomes using Cox proportional hazards models and a Fine-Gray subdistribution hazards regression model. For the subset of data where information on previous clozapine use was available, we did an analysis for participants who did and did not have previous clozapine exposure.
Findings: We included 26 630 people, with 2·6 million ANC values. Within the total cohort, 17 585 people (66%) were male, 9025 (33·9%) female, and 20 (0·1%) other gender, and the mean age was 36·1 years (SD 13·7). We did not have data on race or ethnicity. Of the 26 630 people taking clozapine, 1146 (4·3%) had minor neutropenia, 313 (1·2%) had serious neutropenia leading to cessation, and 223 (0·8%) had serious neutropenia unrelated to clozapine without cessation. In people with no previous exposure to clozapine (n=15 973), the cumulative incidence of serious neutropenia leading to cessation was 0·9% at 18 weeks and 1·4% at 2 years; the weekly incidence rate for serious neutropenia leading to cessation peaked at 9 weeks (0·128%) and fell to a rolling average weekly incidence of 0·001% by 2 years. For minor neutropenia, the cumulative incidence was 1·7% at 18 weeks and 3·5% at 2 years; the weekly incidence rate peaked at 9 weeks (0·218%) and fell to a stable rolling average of 0·01%. The median time to a serious neutropenic event leading to cessation was 17 weeks (IQR 9·96-102). Previous clozapine exposure reduced the risk of serious neutropenia leading to cessation (OR 0·19, 95% CI 0·12-0·31; p <0·0001).
Interpretation: Most serious neutropenia leading to clozapine cessation occurs within 18 weeks of treatment and becomes negligible after 2 years. Weekly haematological monitoring after the first 18 weeks could be safely reduced to once every 4 weeks and ceased after 2 years unless clinically indicated. Clozapine retrial after interruption with 2 cumulative years of unremarkable testing might not require further haematological monitoring. A serious neutropenia ANC threshold of ≤1·0 × 109 per L could be used in more jurisdictions.
Funding: None.
Copyright © 2024 Elsevier Ltd. All rights reserved.