%20-%20https://sketch.com%20--%3e%3ctitle%3e3ECBBF50-FD08-412A-943D-2CFC5E025A16%3c/title%3e%3cdesc%3eCreated%20with%20sketchtool.%3c/desc%3e%3cg%20id='Stylesheet'%20stroke='none'%20stroke-width='1'%20fill='none'%20fill-rule='evenodd'%3e%3cg%20transform='translate(-926.000000,%20-668.000000)'%20id='BUTTONS'%3e%3cg%20transform='translate(100.000000,%20597.000000)'%3e%3cg%20id='BTN-/-RIGHTS'%20transform='translate(816.000000,%2065.000000)'%3e%3cg%20id='ICON-/-Rights'%20transform='translate(10.000000,%206.000000)'%3e%3cpath%20d='M13.5752841,0%20L1.34943182,0%20C0.59375,0%200,0.586666667%200,1.33333333%20L0,22.6666667%20C0,23.4133333%200.620738636,24%201.34943182,24%20L17.6505682,24%20C18.40625,24%2019,23.3866667%2019,22.6666667%20L19,5.36%20L13.5752841,0%20Z%20M13.5752841,1.89333333%20L17.0838068,5.36%20L13.5752841,5.36%20L13.5752841,1.89333333%20Z%20M1.34943182,22.6666667%20L1.34943182,1.33333333%20L12.1988636,1.33333333%20L12.1988636,5.36%20C12.1988636,6.10666667%2012.7926136,6.69333333%2013.5482955,6.69333333%20L17.6235795,6.69333333%20L17.6505682,22.6666667%20L1.34943182,22.6666667%20Z'%20id='Shape'%20fill='%23595959'%20fill-rule='nonzero'%3e%3c/path%3e%3ccircle%20id='Oval'%20stroke='%23595959'%20stroke-width='1.5'%20fill='%23595959'%20cx='9.5'%20cy='13.5355339'%20r='5.5'%3e%3c/circle%3e%3cpolyline%20id='Path-2'%20stroke='%23FFFFFF'%20stroke-width='1.5'%20transform='translate(9.500000,%2013.035534)%20rotate(45.000000)%20translate(-9.500000,%20-13.035534)%20'%20points='8%2016.0355339%2011%2016.0355339%2011%2010.0355339'%3e%3c/polyline%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/svg%3e)
Introduction
As a highly efficacious atypical antipsychotic, clozapine is recommended for treatment-resistant schizophrenia in most of the world (Crilly, Reference Crilly2007). Following a 1975 report in Finland of 16 cases of clozapine-related agranulocytosis resulting in eight deaths due to infection (Idanpaan-Heikkila et al., Reference Idanpaan-Heikkila, Alhava, Olkinuora and Palva1975), clozapine was withdrawn in many countries. However, since the landmark study in 1988 demonstrating its safety with strict blood monitoring (Kane et al., Reference Kane, Honigfeld, Singer and Meltzer1988), clozapine has been approved by the Food and Drug Administration of the USA followed by the majority of health authorities for patients with treatment-resistant schizophrenia (Alvir et al., Reference Alvir, Lieberman, Safferman, Schwimmer and Schaaf1993) because of its superior efficacy in improving psychotic symptoms (Wahlbeck et al., Reference Wahlbeck, Cheine, Essali and Adams1999; Volavka et al., Reference Volavka, Czobor, Sheitman, Lindenmayer, Citrome, McEvoy, Cooper, Chakos and Lieberman2002; Siskind et al., Reference Siskind, McCartney, Goldschlager and Kisely2016), aggressive, hostile or self-harming behaviors (Volavka et al., Reference Volavka, Czobor, Nolan, Sheitman, Lindenmayer, Citrome, McEvoy, Cooper and Lieberman2004; Faay et al., Reference Faay, Czobor and Sommer2018), substance abuse (Arranz et al., Reference Arranz, Garriga, Garcia-Rizo and San2018), and reducing the risk of hospitalization (Land et al., Reference Land, Siskind, McArdle, Kisely, Winckel and Hollingworth2017) and overall mortality (Vermeulen et al., Reference Vermeulen, van Rooijen, van de Kerkhof, Sutterland, Correll and de Haan2018).
Despite its better efficacy compared to all other antipsychotics for treatment-resistant schizophrenia, clozapine is still under-utilized in many countries because of its severe side effect profile, particularly the concern about the development of agranulocytosis (Oyesanmi et al., Reference Oyesanmi, Kunkel, Monti and Field1999). According to the Clozaril Patient Monitoring System (CPMS), the cumulative incidence of agranulocytosis related to clozapine was 0.8% at 1 year and 0.93% at 1.5 years; more than 80% of patients developed agranulocytosis in their first 3 months of treatment (Alvir et al., Reference Alvir, Lieberman, Safferman, Schwimmer and Schaaf1993). Between 1993 and 1996, the rate of agranulocytosis was 0.9% in Australia (Copolov et al., Reference Copolov, Bell, Benson, Keks, Strazzeri and Johnson1998). The general consensus is that patients receiving clozapine need regular white blood cell monitoring, particularly during the first 18 weeks of clozapine administration (Bastani et al., Reference Bastani, Alphs and Meltzer1989). Due to the stringent blood monitoring, the rate of agranulocytosis was significantly reduced to 0.38% in the 5 years after its re-introduction (Honigfeld, Reference Honigfeld1996).
In a recent meta-analysis, the incidence of severe clozapine-associated neutropenia, defined as an absolute neutrophil count (ANC) <500 ul, was 0.9%, while the fatality rate of severe neutropenia was 2.1% (Myles et al., Reference Myles, Myles, Xia, Large, Kisely, Galletly, Bird and Siskind2018). In recent years, under strict blood monitoring, both the risk of agranulocytosis and the clozapine-related death rate have been significantly reduced. A thorough search of the literature could not locate any meta-analysis on the prevalence of agranulocytosis and death related to clozapine.
Clozapine has been widely prescribed in China without interruption since 1976 even when it was withdrawn from the market throughout the world. A nationwide survey of psychotropic medications in China found that clozapine was prescribed for 39.7, 32.5, and 26.4% of schizophrenia patients in 2002, 2006, and 2012, respectively (Li et al., Reference Li, Xiang, Su, Shu, Yu, Chiu, Correll, Ungvari, Lai, Ma, Wang, Bai, Li, Sun, Shi, Chen, Mei, Li and Si2015). Many prevalence studies of clozapine-associated agranulocytosis have been published in Chinese language journals, which are not generally accessible to non-Chinese readership.
This meta-analysis examined the prevalence of agranulocytosis and related death and their associated factors in clozapine-treated patients.
Methods
Inclusion and exclusion criteria
Studies that fulfilled the following criteria were included: (a) cross-sectional or cohort studies (only baseline data of cohort studies were extracted) reporting accessible data on the prevalence of agranulocytosis or data that could generate a prevalence figure of agranulocytosis in clozapine-treated patients; (b) published in English or Chinese languages. Case reports and case series, prescription surveys, and reports with very small sample size were excluded.
Although the Council for International Organisations of Medical Sciences (CIOMS) (2001) defined agranulocytosis as neutrophil count of <0.1 × 10^9/L, currently the traditional criteria (neutrophil count of <1.0 × 10^9/L or <0.5 × 10^9/L) are still used to define agranulocytosis in clinical guidelines on blood monitoring for clozapine, particularly in China. The two traditional criteria were used in the studies included in the current meta-analysis. In order to reflect actual clinical practice and remain consistent with the included studies, in this meta-analysis, the term ‘agranulocytosis’ was defined according to the earlier diagnostic criteria of the CIOMS (Bankowski et al., Reprinted, Reference Bankowski, Bruppacher, Crusius, Gallagher, Kremer and Venulet2000).
Search strategy
Two reviewers (XHL and XMZ) systematically and independently searched major international (PubMed, PsycINFO, and EMBASE) and Chinese (WanFang, Chinese National Knowledge Infrastructure, and Sinomed) databases from their inception up to 8 June 2018 with the following search terms: (Clozapine OR Clozaril OR Leponex) AND (agranulocytosis OR ‘granulocyte deficiency’ OR agranulocytopenia OR aleucocytosis OR aleukocytosis OR hypoleucocytosis OR leucopenia OR leukopenia OR neutropenia OR ‘sudden death’ OR ‘unexpected death’ OR mortality OR death). In addition, a manual search was conducted by reviewing the reference lists of relevant meta-analyses and reviews.
Data extraction
Data extraction was independently performed by two reviewers (XHL and XMZ) who screened the titles and abstracts first, and later read the full texts. If more than one article were published using the same dataset, only the article with complete data was included. Any inconsistencies in data extraction were discussed or resolved by involving a third reviewer (YTX).
Quality assessment
Following the methodology of other studies (Cooper et al., Reference Cooper, Balamurali and Livingston2007; Pringsheim et al., Reference Pringsheim, Jette, Frolkis and Steeves2014), two researchers (XHL and XMZ) independently assessed the quality of studies with the methodological quality checklist that comprised eight items (Boyle, Reference Boyle1998). Each study was scored from 0 to 8. The score of 7–8 was considered as ‘high quality’, 4–6 as ‘moderate quality’, and 0–3 as ‘low quality’. Any disagreement in the assessment was discussed and resolved involving a third investigator (WWR). Table 3 shows the details of the quality assessment of the studies.
Statistical analyses
The Comprehensive Meta-Analysis Version 2 (http://www.meta-analysis.com) was used to analyze data. The pooled prevalence estimates of agranulocytosis and their 95% confidence intervals (95% CI) were calculated using the random-effects models. The I 2 statistic was calculated to measure heterogeneity (Higgins et al., Reference Higgins, Thompson, Deeks and Altman2003). When high heterogeneity was present (I 2 > 50%), sensitivity and subgroup analyses were performed to explore the sources of heterogeneity. Publication bias was assessed with the funnel plot, and the Begg's and Egger's tests (Egger et al., Reference Egger, Davey Smith, Schneider and Minder1997). The level of significance was set at 0.05 (two-sided).
Results
Study characteristics and quality assessment
Thirty-six (13 in English and 23 in Chinese) of the 5176 potentially eligible articles met the inclusion criteria and were included in the analyses (Figure 1). Table 1 shows the characteristics of these 36 studies on 260 948 subjects. They were conducted between 1984 and 2018 across 12 countries in five continents: Asia (25 studies: 23 in China; one each in Korea and Saudi Arabia), Europe (six studies: one each in UK and Ireland, Hungary, Iceland, Italy, Denmark, and the Slovak Republic), North America (three studies in the USA), South America (one study in Argentina) and Oceania (one study in Australia). Sample sizes ranged from 147 to 99 502; the mean age of patients varied between 15.0 and 51.2 years. Eight studies were based on clozapine monitoring systems.
PRISMA flow diagram.
Characteristics of studies included in the meta-analysis
CPMS, Clozaril Patient Management System; ANMAT, Argentine drug-regulatory agency; CAMP, Clozapine Authorization and Monitoring Program; CNR, The Clozaril National Registry; ICLOS, Italian Clozapine Monitoring System; CDARS, Clinical Data Analysis and Reporting System
CCMD-2-R, Chinese Classification of Mental Disorders, the Second Edition, Revised; CCMD-3, Chinese Classification of Mental Disorders, the Third Version; DSM, Diagnostic and Statistical Manual of Mental Disorders; ICD-9, International Classification of Disease, the Ninth version
a Having strict blood monitoring during the clozapine treatment
b Median age
c The mean age was derived from the number of 195 patients
One study reported on two samples: clozapine alone and clozapine plus lithium (Gan and Chen, Reference Gan and Chen1995). Clozapine-associated agranulocytosis occurred from 26.07 days to 4 years after exposure to clozapine. Thirteen studies (Fan et al., Reference Fan, Cui, Shen, Ju, Tang, Bai, Yang and Kong1992; Zhu et al., Reference Zhu, Wang and Wu1992; Peacock and Gerlach, Reference Peacock and Gerlach1994; Gan and Chen, Reference Gan and Chen1995; Lv et al., Reference Lv, Niu, Zhou, Zhou, Pan, Sun, Yu, Gu, Mei, Li, Wei, Feng, Yu, Wang, Wang, Jiang, Wang, Yu, Wu, Huang, Zhan, Zhang, Ji, Cui, Zhan, Sun, Zhuang, Fu and Guang1997; Copolov et al., Reference Copolov, Bell, Benson, Keks, Strazzeri and Johnson1998; Deng et al., Reference Deng, Wang, Ji and Zou1999; Zhao et al., Reference Zhao, Si and Zhang1999; Xu et al., Reference Xu, Wang, Wu and Chen2001; Gaszner et al., Reference Gaszner, Makkos and Kosza2002; Kang et al., Reference Kang, Cho, Oh, Lee, Chae and Ko2006; Pecenak et al., Reference Pecenak, Novotny and Janik2009; Balda et al., Reference Balda, Garay, Papale, Bignone, Bologna, Brandolini, Prokopez, Balasini, Baldessarini and Daray2015) reported the mean length of clozapine exposure when agranulocytosis occurred (9.8 weeks); of them, nine studies reported the mean length with standard deviations (s.d.) (5.4w ± 3.3w) (Fan et al., Reference Fan, Cui, Shen, Ju, Tang, Bai, Yang and Kong1992; Zhu et al., Reference Zhu, Wang and Wu1992; Gan and Chen, Reference Gan and Chen1995; Lv et al., Reference Lv, Niu, Zhou, Zhou, Pan, Sun, Yu, Gu, Mei, Li, Wei, Feng, Yu, Wang, Wang, Jiang, Wang, Yu, Wu, Huang, Zhan, Zhang, Ji, Cui, Zhan, Sun, Zhuang, Fu and Guang1997; Copolov et al., Reference Copolov, Bell, Benson, Keks, Strazzeri and Johnson1998; Deng et al., Reference Deng, Wang, Ji and Zou1999; Zhao et al., Reference Zhao, Si and Zhang1999; Xu et al., Reference Xu, Wang, Wu and Chen2001; Balda et al., Reference Balda, Garay, Papale, Bignone, Bologna, Brandolini, Prokopez, Balasini, Baldessarini and Daray2015). Four studies (Copolov et al., Reference Copolov, Bell, Benson, Keks, Strazzeri and Johnson1998; Munro et al., Reference Munro, O'Sullivan, Andrews, Arana, Mortimer and Kerwin1999; Kang et al., Reference Kang, Cho, Oh, Lee, Chae and Ko2006; Kelly et al., Reference Kelly, Kreyenbuhl, Dixon, Love, Medoff and Conley2007) reported that 53.7–87.5% of agranulocytosis occurred in the first 18 weeks of clozapine treatment. Insufficient data were available to calculate proportion of agranulocytosis in the first 12 months of clozapine exposure.
Several studies referred to the term ‘incidence’ (such as Copolov et al., Reference Copolov, Bell, Benson, Keks, Strazzeri and Johnson1998; Honigfeld et al., Reference Honigfeld, Arellano, Sethi, Bianchini and Schein1998; Munro et al., Reference Munro, O'Sullivan, Andrews, Arana, Mortimer and Kerwin1999; Lambertenghi Deliliers, Reference Lambertenghi Deliliers2000; Kang et al., Reference Kang, Cho, Oh, Lee, Chae and Ko2006; Abanmy et al., Reference Abanmy, Al-Jaloud, Al-Jabr, Al-Ruwaisan, Al-Saeed and Fatani2014; Balda et al., Reference Balda, Garay, Papale, Bignone, Bologna, Brandolini, Prokopez, Balasini, Baldessarini and Daray2015; Sing et al., Reference Sing, Wong, Cheung, Chan, Chu and Cheung2017), but, in fact, the figures reflected prevalence, i.e. the number of cases over a period, thus they were included in the analyses. The mean score of quality assessment was 5.1 (range: 4–8). One study (2.6%) was rated as ‘high quality’ (Gerbino-Rosen et al., Reference Gerbino-Rosen, Tompkins, Nusser, Napolitano, Henderson and Kumra2005) while the rest of 35 studies were rated as ‘moderate quality’ (Table 2).
The quality assessment
The pooled prevalence of clozapine-associated agranulocytosis and related death
The prevalence of agranulocytosis ranged from 0.1% to 2.7% in the 36 studies with a pooled prevalence of 0.4% (95% CI 0.3–0.6%, I 2 = 90.2%); i.e. one case of agranulocytosis in every 250 clozapine-treated patients.
In the 30 studies that reported deaths caused by clozapine-associated agranulocytosis (n = 33), the pooled prevalence of death was 0.05% (95% CI 0.03–0.09%, I 2 = 63.6%; range: 0–0.8%); i.e. one in every 2000 clozapine-treated patients. Among patients with agranulocytosis, the pooled prevalence of death was 10.0% (95% CI 6.1–15.8%, I 2 = 43.7%); i.e. one in every 10 patients.
Subgroup analysis
Table 3 shows the results of the subgroup analyses. The prevalence of agranulocytosis was not significantly associated with gender, Chinese studies, and geographic locations. The prevalence of agranulocytosis was 0.5% in men, 0.7% in women; 0.4% in Chinese and 0.5% in non-Chinese studies. Studies conducted in North and South America (0.2%) reported a lower prevalence than those in Asia (0.4%), Europe (0.5%), and Oceania (0.9%), without significant group difference. There were significant differences in the prevalence of agranulocytosis between studies with smaller sample size (0.8% in studies of sample size <1000) and larger sample size (0.3% in studies of sample size ⩾1000; Q = 11.41, p = 0.001). The trend of significant difference was found between studies with or without strict blood monitoring (0.5% v. 0.3%; Q = 2.887, p = 0.089). No significant differences were found between different periods of publication (before 1991 v. 1991 and after), and between clozapine alone v. clozapine + other psychotropic medications. Broad criteria of agranulocytosis (neutrophil<1.0 × 109/L) was associated with a slightly higher prevalence (0.5%) than those (0.4%) applying stringent criteria (neutrophil<0.5 × 109/L), but the difference was not significant (p = 0.50).
Subgroup analysis of pooled prevalence of agranulocytosis
a Gan and Chen (Reference Gan and Chen1995), had two samples (clozapine alone; clozapine + lithium)
Meta regression of the prevalence of agranulocytosis caused by clozapine
Meta-regression analysis found a significant negative association between the publication period (before or being/after 1991) and the prevalence of agranulocytosis based on 36 studies (β = −0.03, 95% CI −0.05 to −0.02, p < 0.001). Years of data collection were negatively associated with prevalence of agranulocytosis in 33 studies with available data (β = −0.03, 95% CI −0.05 to −0.01, p < 0.001). Age showed a significant trend with agranulocytosis in 16 studies with available data (β = 0.05, 95% CI −0.002 to 0.11, p = 0.06), while study quality assessment had significant association with the prevalence of agranulocytosis (β = −0.46, 95% CI −0.66 to −0.26, p = 0.00001).
Publication bias and sensitivity analysis
Online Supplementary Figure S1 shows the funnel plot of all studies. Neither the funnel plot, nor the Egger's (t = 0.049, 95% CI −1.49 to 1.56; p = 0.96) and Begg's tests (z = 1.35, p = 0.18) found any publication bias. After excluding each study sequentially, the recalculated pooled results did not significantly change, indicating that there was no outlying study that significantly influenced the overall results.
Discussion
The pooled prevalence of clozapine-associated agranulocytosis in this meta-analysis was 0.4%, similar to the US registry studies in the 5 years post 1990 (0.38%) (Honigfeld, Reference Honigfeld1996). The pooled prevalence of death caused by clozapine-associated agranulocytosis was 0.05%, which is higher than the previously reported figure (0.012%) based on the national registry database of the US manufacturer of clozapine (Honigfeld, Reference Honigfeld1996). Comparison across studies should be made with caution due to different study designs and definitions of agranulocytosis. This was the first meta-analysis of prevalence of clozapine-associated agranulocytosis, therefore direct comparisons with other meta-analyses could not be performed.
A higher risk of agranulocytosis in Asian patients receiving clozapine is found in some (Munro et al., Reference Munro, O'Sullivan, Andrews, Arana, Mortimer and Kerwin1999), but not in all studies (Shapiro et al., Reference Shapiro, Issaragrisil, Kaufman, Anderson, Chansung, Thamprasit, Sirijirachai, Piankijagum, Porapakkham, Vannasaeng, Leaverton and Young1999; Sing et al., Reference Sing, Wong, Cheung, Chan, Chu and Cheung2017). Gene sequence variations, such as HLA-DQB1 and HLA-A1, have been associated with an increased risk of clozapine-associated agranulocytosis (Lieberman et al., Reference Lieberman, Yunis, Egea, Canoso, Kane and Yunis1990; Amar et al., Reference Amar, Segman, Shtrussberg, Sherman, Safirman, Lerer and Brautbar1998; Valevski et al., Reference Valevski, Klein, Gazit, Meged, Stein, Elizur, Narinsky, Kutzuk and Weizman1998; Lahdelma et al., Reference Lahdelma, Ahokas, Andersson, Suvisaari, Hovatta, Huttunen and Koskimies2001; Dettling et al., Reference Dettling, Cascorbi, Opgen-Rhein and Schaub2007; Athanasiou et al., Reference Athanasiou, Dettling, Cascorbi, Mosyagin, Salisbury, Pierz, Zou, Whalen, Malhotra, Lencz, Gerson, Kane and Reed2011; Kadasah et al., Reference Kadasah, Arfin and Tariq2011; Goldstein et al., Reference Goldstein, Jarskog, Hilliard, Alfirevic, Duncan, Fourches, Huang, Lek, Neale, Ripke, Shianna, Szatkiewicz, Tropsha, van den Oord, Cascorbi, Dettling, Gazit, Goff, Holden, Kelly, Malhotra, Nielsen, Pirmohamed, Rujescu, Werge, Levy, Josiassen, Kennedy, Lieberman, Daly and Sullivan2014; Legge et al., Reference Legge, Hamshere, Ripke, Pardinas, Goldstein, Rees, Richards, Leonenko, Jorskog, Chambert, Collier, Genovese, Giegling, Holmans, Jonasdottir, Kirov, McCarroll, MacCabe, Mantripragada, Moran, Neale, Stefansson, Rujescu, Daly, Sullivan, Owen, O'Donovan and Walters2017). Genetic variations between ethnic groups may be associated with different prevalence of clozapine-associated agranulocytosis. However, in this meta-analysis, the prevalence of clozapine-associated agranulocytosis in Asia was not significantly higher than in America, Europe, and Oceania.
The prevalence of agranulocytosis in studies with smaller sample size (0.8%) was significantly higher than in those with larger sample size (0.3%; p = 0.001). There is no explanation for this observation; it may well be that findings of studies with smaller sample size are not sufficiently stable (Cao et al., Reference Cao, Wang, Zhong, Zhang, Ungvari, Ng, Li, Chiu, Lok, Lu, Jia and Xiang2017). No differences were found in the prevalence of agranulocytosis in studies published before and after 1991 (Crilly, Reference Crilly2007), when strict registry-based prescribing system for hematological monitoring was mandatory. This finding is consistent with that of a recent review (Myles et al., Reference Myles, Myles, Xia, Large, Kisely, Galletly, Bird and Siskind2018).
Studies using the broad diagnostic criteria of agranulocytosis (neutrophil < 1.0 × 109/L) had a higher pooled prevalence than those applying more stringent criteria (neutrophil < 0.5 × 109/L), although the difference did not reach significant level. Recent treatment guidelines recommended a flexible threshold for clozapine treatment, i.e. if the ANC ranges from 1.5 × 109/L to 0.5 × 109/L in patients with benign ethnic neutropenia, clozapine could still be continued since people with persistent or episodic neutropenia usually do not progress to agranulocytosis (Bastiampillai et al., Reference Bastiampillai, Gupta, Chan and Allison2016). Following the recommended flexible threshold of neutrophil count would allow more patients to benefit from clozapine treatment (Sultan et al., Reference Sultan, Olfson, Correll and Duncan2017).
Consistent with previous findings (Alvir et al., Reference Alvir, Lieberman, Safferman, Schwimmer and Schaaf1993; Atkin et al., Reference Atkin, Kendall, Gould, Freeman, Liberman and O'Sullivan1996; Lieberman and Alvir, Reference Lieberman and Alvir1992; Munro et al., Reference Munro, O'Sullivan, Andrews, Arana, Mortimer and Kerwin1999), age seemed to be significantly associated with the prevalence of agranulocytosis (β = 0.05, p = 0.06). This may be due to the higher vulnerability to hematopoietic toxicity associated with clozapine in older patients (Munro et al., Reference Munro, O'Sullivan, Andrews, Arana, Mortimer and Kerwin1999). Combining clozapine with other psychotropic medications increases the risk of agranulocytosis (Peacock and Gerlach, Reference Peacock and Gerlach1994; Lau and Yim, Reference Lau and Yim2015), which was not confirmed in this meta-analysis. This could be partly due to the small number of studies with concomitant medications with clozapine.
A negative association between publication period and the prevalence of agranulocytosis was found. In earlier times, regular blood monitoring for clozapine was not mandatory in certain parts of the world, such as China, due to insufficient training or economic reasons. Hence, mild-to-moderate clozapine-associated neutropenia was not identified early leading to higher risk of agranulocytosis. Unexpectedly, the time frame of data collection was negatively associated with the prevalence of agranulocytosis (β = −0.03, p < 0.001). The possible explanation is that most cases of agranulocytosis occur in the first 18 weeks of exposure to clozapine. Since many patients were only included in the clozapine monitoring system after they had been taking clozapine for a period of time, which is common in China, the most vulnerable period for agranulocytosis was not captured and the prevalence of agranulocytosis was consequently lower. Furthermore, a negative association between low study quality and the prevalence of agranulocytosis was found.
There are several limitations of this meta-analysis. First, several included studies were based on clozapine monitoring systems with various timespan, thus the prevalence of agranulocytosis in specific timeframes could not be calculated. Second, in order to reflect actual clinical practice, two diagnostic criteria of agranulocytosis (neutrophils of <1.0 × 10^9/L or <0.5 × 10^9/L) were applied. However, this did not significantly influence the prevalence of agranulocytosis according to the subgroup analyses. Third, similar to other recent meta-analyses (Winsper et al., Reference Winsper, Ganapathy, Marwaha, Large, Birchwood and Singh2013; Long et al., Reference Long, Huang, Liang, Liang, Chen, Xie, Jiang and Su2014; Mata et al., Reference Mata, Ramos, Bansal, Khan, Guille, Di Angelantonio and Sen2015; Li et al., Reference Li, Cao, Zhong, Ungvari, Chiu, Lai, Zheng, Correll and Xiang2016), high heterogeneity still remained in certain subgroup analyses which is difficult to avoid in meta-analysis of observational surveys. The heterogeneity was probably associated with factors that were not examined in most included studies, such as clozapine doses, treatment stage and duration, psychiatric comorbidities, and concurrent use of other psychotropic medications. Fourth, different periods of data collection and quality of studies could moderate the pooled prevalence of agranulocytosis. However, meta-regression analysis did not find any significant moderating effect of the period of data collection on the primary results.
In conclusion, this meta-analysis found that the prevalence of clozapine-associated agranulocytosis is low. Death due to agranulocytosis following exposure to clozapine appeared to be rare.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0033291719000369.
Author ORCIDs
Yu-Tao Xiang, 0000-0002-2906-0029
Acknowledgements
Not applicable.
Financial support
The study was supported by the University of Macau (MYRG2015-00230-FHS; MYRG2016-00005-FHS), the National Key Research & Development Program of China (No. 2016YFC1307200), the Beijing Municipal Administration of Hospitals Clinical Medicine Development of Special Funding Support (No.ZYLX201607), and the Beijing Municipal Administration of Hospitals’ Ascent Plan (No. DFL20151801).
Conflict of interest
None.
'/%3e%3cpath%20d='M17%207.971L15.711%206.677%2010.912%2011.495%2010.912%200%209.088%200%209.088%2011.495%204.289%206.677%203%207.971%2010%2015z'%20transform='translate(8%208)'/%3e%3c/g%3e%3c/g%3e%3c/svg%3e)
