Agranulocytosis is a life-threatening side-effect of clozapine, the only approved drug for treatment-resistant schizophrenia. The long-term profile of this complication has not yet been well established. Here we aim to describe the risk of clozapine-induced agranulocytosis over the long term. We used the entire population of Finland to identify people diagnosed with schizophrenia or schizoaffective disorder between 1972 and 2014 and developed a Kaplan–Meier model of time to diagnosis of agranulocytosis during clozapine versus non-clozapine treatment over a 22-year observation period (1996 to 2017). Next, we developed a nested case–control model for agranulocytosis matching by sex, age, time since diagnosis, and being in the incident cohort on a 1 to 5 ratio. Various durations of use for clozapine and non-clozapine antipsychotic treatment were compared to the modal antipsychotic use duration, deriving adjusted odds ratios (aORs) in a multivariable regression model. Recurrence and lethality rates for clozapine-induced agranulocytosis were described. These data reflect on all individuals with lived experience of schizophrenia in Finland during the study time, although individuals with lived experience were not included in the design of the study. We identified 61 769 people with schizophrenia or schizoaffective disorder (14 037 individuals treated with clozapine and 47 732 individuals treated with non-clozapine antipsychotics), with a mean age of 46·67 years (IQR 34·44–57·61), of whom 30 721 (49·7%) were female and 31 048 (50·3%) were male (data on ethnicity not available). Among those, 398 individuals were diagnosed with agranulocytosis (231 individuals treated with clozapine and 167 individuals treated with non-clozapine antipsychotics), representing a cumulative incidence of agranulocytosis for 1·37% (95% CI 0·58–3·16) on clozapine and 0·13% (0·04–0·23) on non-clozapine antipsychotics. In the case (n=398) versus control (n=1987) model, the risk of clozapine-induced agranulocytosis decreased steeply over time from an aOR of 36·01 (95% CI 16·79–77·22) for less than 6 months on clozapine to 4·38 (1·86–10·34) for clozapine use of 54 months or more. Only one of 3559 individuals starting clozapine died because of clozapine-induced agranulocytosis. The risk of clozapine-induced agranulocytosis decreases steeply over time but might be persistently greater than that of non-clozapine antipsychotics. This long-term risk excess seems small in absolute terms compared with the known magnitude of the advantages of clozapine in relevant outcomes, including life expectancy. Given the widespread underuse of clozapine, relaxing the long-term neutrophil monitoring could favour the advantages of long-term clozapine use, including greater life expectancy, without incurring the intolerable risk of clozapine-induced agranulocytosis. Northwell Health and Sigrid Jusèlius Foundation.

中文翻译：

---

与其他抗精神病药物相比，氯氮平导致粒细胞缺乏症风险的长期持续性：芬兰的一项全国性队列和病例对照研究


粒细胞缺乏症是氯氮平的一种危及生命的副作用，氯氮平是唯一被批准用于治疗难治性精神分裂症的药物。这种并发症的长期概况尚未明确。在这里，我们的目的是描述氯氮平引起的粒细胞缺乏症的长期风险。我们利用芬兰全体人口来识别 1972 年至 2014 年间被诊断患有精神分裂症或分裂情感性障碍的患者，并开发了一个 Kaplan-Meier 模型，该模型显示了在 22 年的观察期内（1996 年至 2014 年）氯氮平与非氯氮平治疗期间诊断粒细胞缺乏症的时间。 2017）。接下来，我们开发了一个针对粒细胞缺乏症的巢式病例对照模型，该模型按性别、年龄、诊断后时间以及事件队列中的比例进行 1 比 5 匹配。将氯氮平和非氯氮平抗精神病药物的不同使用持续时间与模式抗精神病药物使用持续时间进行比较，得出多变量回归模型中的调整优势比（aOR）。描述了氯氮平诱导的粒细胞缺乏症的复发率和死亡率。这些数据反映了研究期间在芬兰有精神分裂症生活经历的所有个人，尽管有生活经历的个人并未包括在研究设计中。我们确定了 61 769 名精神分裂症或分裂情感性障碍患者（14 037 人接受氯氮平治疗，47 732 人接受非氯氮平抗精神病药治疗），平均年龄为 46·67 岁（IQR 34·44–57·61），其中其中 30 721 (49·7%) 为女性，31 048 (50·3%) 为男性（种族数据不可用）。 其中，398 人被诊断为粒细胞缺乏症（231 人接受氯氮平治疗，167 人接受非氯氮平抗精神病药治疗），粒细胞缺乏症的累积发生率为 1·37% (95% CI 0·58–3·16)。氯氮平和 0·13% (0·04–0·23) 的非氯氮平抗精神病药。在病例 (n=398) 与对照 (n=1987) 模型中，氯氮平诱导的粒细胞缺乏症的风险随着时间的推移从 aOR 36·01 (95% CI 16·79–77·22) 急剧下降，持续时间小于使用氯氮平 6 个月至 4·38 (1·86–10·34)，使用氯氮平 54 个月或以上。开始服用氯氮平的 3559 名个体中，只有 1 人因氯氮平引起的粒细胞缺乏症死亡。随着时间的推移，氯氮平引起的粒细胞缺乏症的风险急剧下降，但可能持续高于非氯氮平抗精神病药。与氯氮平在相关结果（包括预期寿命）方面已知的优势程度相比，这种长期超额风险从绝对值来看似乎很小。鉴于氯氮平普遍使用不足，放松长期中性粒细胞监测可能有利于长期使用氯氮平的优势，包括更长的预期寿命，而不会产生氯氮平引起的粒细胞缺乏症的无法忍受的风险。 Northwell Health 和 Sigrid Jusèlius 基金会。