European Psychiatry 29 (2014) 1–10

Review

Systematic review of the efficacy and tolerability of Clozapine in thetreatment of youth with early onset schizophrenia

C. Schneider a, R. Corrigall a, D. Hayes a, M. Kyriakopoulos a, S. Frangou b,*a Child and Adolescent Mental Health Services, South London and Maudsley NHS Foundation Trust, London, UKb Department of Psychiatry, Icahn School of Medicine at Mount Sinai, Box 1230, 1425, Madison Avenue, New York, NY 10029, USA

A R T I C L E I N F O

Article history:

Received 20 May 2013

Received in revised form 20 July 2013

Accepted 11 August 2013

Available online 9 October 2013

Keywords:

Schizophrenia

Early onset

Paediatric

Efficacy

Tolerability

Clozapine

A B S T R A C T

Background: The use of clozapine (CLZ) for treatment-resistant schizophrenia is well established in

adults. However, it is seldom used in youth with early onset schizophrenia (EOS) largely because of lack

of clarity about its risk benefit ratio. This review synthesises and evaluates available evidence regarding

the efficacy and tolerability of CLZ in EOS with the aim to assist clinical decision-making.

Methods: We conducted a systematic review of the primary literature on the clinical efficacy and adverse

drug reactions (ADRs) observed during CLZ treatment in EOS. We also identified relevant practice

guidelines and summarised current guidance.

Results: CLZ showed superior efficacy than other antipsychotics in treating refractory EOS patients; short-

term clinical trials suggest an average improvement of 69% on the Brief Psychiatric Rating Scale that was

sustained during long-term follow-up (up to 9 years). No fatalities linked to CLZ treatment were reported.

Sedation and hypersalivation were the most common complaints, reported by over 90% of patients. Other

common ADRs (reported in 10-60% of patients) were enuresis, constipation, weight gain, and non-specific

EEG changes. Less common ADRs (reported in 10-30% of patients) were akathisia, tachycardia and changes

in blood pressure. Neutropenia was reported in 6–15% of cases but was usually transient while

agranulocytosis was rare (< 0.1%). Seizures were also uncommon (< 3%). Metabolic changes were

relatively common (8–22%) but emergent diabetes was not frequently observed (< 6%). Overall the rate of

discontinuation was low (3–6%). Current guidelines recommend the use of CLZ in EOS patients who have

failed to respond to two adequate trials with different antipsychotics and provide detailed schedules of

assessments to evaluate and assess potential ADRs both prior to initiation and throughout CLZ treatment.

Conclusion: Available data although limited in terms of number of studies are consistent in

demonstrating that CLZ is effective and generally safe in the treatment of refractory EOS provided

patients are regularly monitored

� 2013 Elsevier Masson SAS. All rights reserved.

Available online at

www.sciencedirect.com

1. Background

Schizophrenia commonly begins in adulthood, however asubstantial number of individuals experience the onset of thedisorder while they are children or adolescents [4]. The prevalenceof schizophrenia with onset before 13 years of age (childhoodonset schizophrenia; COS) is low (approximately 1 in 40,000children) [19] but the incidence of schizophrenia rises sharply atabout 12–14 years of age [21]. Approximately 5% of patientsdevelop schizophrenia, during their adolescent years before theage of 18 (adolescent onset schizophrenia; AOS) [19]. In thismanuscript we will use the term early onset schizophrenia (EOS)when we collectively refer to COS and AOS groups.

* Corresponding author.

E-mail address: sophia.frangou@mssm.edu (S. Frangou).

0924-9338/$ – see front matter � 2013 Elsevier Masson SAS. All rights reserved.

http://dx.doi.org/10.1016/j.eurpsy.2013.08.001

When schizophrenia manifests in childhood and adolescence,the onset of clinical symptoms occurs during a critical period forcognitive developmental thus leading to greater difficulties inacquiring basic academic skills [5]. EOS is also associated withgreater chronicity and clinical morbidity with the majority ofpatients (72–74%) requiring long-term psychiatric treatment[29,11,30,45]. Psychosocial outcome is commonly worse thanin adult onset schizophrenia [39]. Patients are often unable tosustain close relationships outside their immediate family andremain financially dependent either on their parents or on publicassistance [29,30]. Therefore, it is vital that the treatment andmanagement of EOS be based on the highest quality evidencepossible because of the long- term implications for engagementwith support services, adherence to treatment and quality of life.

Antipsychotic medication is the mainstay of pharmacologicaltreatment of schizophrenia regardless of age of onset. The efficacyof antipsychotic medication in EOS patients was subjected to

C. Schneider et al. / European Psychiatry 29 (2014) 1–102

meta-analysis by Armenteros and Davies in 2006 [3]. Thecombined sample comprised of 294 EOS patients, 209 of whomhad been treated with first generation antipsychotics (FGA), 83with a second generation antipsychotic (SGA), and 36 withplacebo. The median response rates across antipsychotic classeswere between 62–70%. The ‘‘Treatment of Early Onset Schizo-phrenia Spectrum Disorders Study’’ (TEOSS) is the largest study todate to directly compare the efficacy and safety of a FGA(molindone) to that of two SGAs (olanzapine and risperidone)in the acute treatment (8 weeks) of 116 young patients with EOSor schizoaffective disorder [42]. According to this study, treat-ment with any antipsychotic resulted in symptomatic improve-ment but response rates (50% with molindone, 34% witholanzapine, 46% with risperidone) were uniformly low. Theseresults suggest that the prevalence of poor treatment respondersamongst young patients is consistently high.

EOS patients are therefore likely to benefit from clozapine (CLZ)given its documented superior efficacy in adult schizophrenia [7].However, CLZ is also associated with metabolic, neurological andhaematological adverse drug reactions (ADRs) [12] and with theemergence of obsessive-compulsive symptoms [32]. In the UK lessthat 0.4% of all CLZ prescriptions are for EOS patients; nearly 40% ofchild and adolescent psychiatrists working in inpatient settingsreport that they have never prescribed CLZ [6]. The most commonreasons were unfamiliarity with CLZ, fear of potential ADRs andlack of specific information and guidance [6].

This article aims to assist clinicians in optimising treatment forEOS using three complementary approaches:

(a) we synthesise and comment on the available evidenceregarding the efficacy and tolerability of CLZ in EOS basedon a systematic literature review;

(b) we discuss guidelines developed by national bodies on the useof CLZ in EOS, and;

(c) we present a practical algorithm for the safe use andmonitoring of CLZ treatment in EOS.

2. Methods

We conducted a comprehensive search of English-languagestudies (clinical trials, naturalistic observational trials and casereports) published up to August 31st 2013 in electronic databases(PubMed, MEDLINE) using both free-text and MeSH searchkeywords: ‘‘clozapine’’, ‘‘schizophrenia’’, ‘‘psychosis’’, ‘‘childhoodOR adolescence’’, ‘‘early onset’’, ‘‘paediatric’’, ‘‘antipsychotic’’ andtheir differing terminations and combinations. The search wassupplemented by a manual review of reference lists from theidentified publications. We also reviewed guidelines published byAugust 31st 2013 from recognised international organisations,societies or colleges to identify those offering specific guidanceregarding the use of CLZ in EOS.

3. Results

Our search identified 16 clinical studies, detailed in Table 1, andfour case series published since 1994. There were only twoguidelines that focused specifically on the use of CLZ in youngpatients. Below we present in detail the available evidence.

3.1. Efficacy

3.1.1. Short-term studies

We identified five studies [16,26,41,43,27] that examined theefficacy of CLZ for periods up to 12 weeks (Table 1). All studiescame from USA centres and four [16,26,41,27] reported on

overlapping patient samples derived from the COS cohort of theChild Branch of the National Institutes of Mental Health (NIMH).One study had an open label design [16], three were double-blindrandomised controlled trials (DBRCT) [26,41,27] and one was a re-analysis of previous open label and double-blind studies [43]. Inthe DBRCTs the comparator antipsychotics were haloperidol orolanzapine. All studies focused on patients with refractoryschizophrenia, defined as failure to respond to adequate treat-ment with at least two antipsychotic drugs. They collectivelysuggest that CLZ has superior efficacy compared to otherantipsychotic agents when used in treatment refractory COS.After 6-8 weeks of CLZ treatment, there was an averageimprovement of 69% in the Brief Psychiatric Rating Scale (BPRS),which was the most commonly used outcome measure[16,26,41,43,27]. Improvement was noted for all symptomsespecially negative symptoms [16,27]. As four of the five studieswere conducted in a single centre with significant overlapbetween study samples the total population of patients fromwhich these observations derive is small (n = 93).

3.1.2. Long-term studies

We identified 5 studies [41,43,40,44,24] that presented datafrom follow-up periods ranging for 12 weeks to 9 years (Table 1).All long-term studies were open label trials and collectivelyincluded a relatively small sample (n = 110). Two studies consistedof open-label follow-up of the NIMH COS cohort mentioned above[41,43]. The remainder were COS and/or AOS studies conducted inthree different specialist centres in Germany, Israel and SouthKorea [40,44,24]. Evidence from these studies suggests that CLZ isassociated with sustained clinical improvement; compared toprevious treatment, the number and duration of hospitalizationreduced with CLZ treatment. The rate of discontinuation was verylow, averaging 6% [40,24], and was jointly accounted for by poorefficacy and tolerability.

3.2. Tolerability

Treatment with CLZ in patients with EOS has a complextolerability profile. Up to 40% of patients experience more than oneADR [24]. In Table 2, we summarise all ADRs that have beenreported during CLZ treatment in EOS and we draw attention to thehighest period of risk for each class of ADRs. Below, we discussADRs in more detail and comment on the level of evidenceavailable regarding tolerability to CLZ in youth with EOS.

3.2.1. CNS effects

3.2.1.1. Sedation. Sedation is an almost universal complaint duringCLZ treatment. Although worse at the start of treatment, it oftenpersists, especially at high doses [16,13].

3.2.1.2. Seizures. The risk of seizures during CLZ treatment in EOSis dose dependent and varies in different studies, ranging from0.2%, at an average daily CLZ dose of 200 mg, to 4% for daily CLZdoses of 500 mg or above [40]. EEG abnormalities were frequentduring CLZ treatment of EOS ranging from 10 to 60% [43,13] butthey were not predictive of seizure risk unless excessively severe(i.e. EEG abnormalities indicating spike discharges or spike-and-wave activity). The reported average incidence of seizures was 3%,with most of the cases appearing after the first year of treatment[41,43,24,17].

3.2.1.3. Akathisia. Akathisia in patients with EOS occurred at a rateof 15–31% both at treatment initiation and over the longer term[40,24]. This rate is significantly higher than the 3% commonlyreported in adults [12]; the mechanism for this is unknown but isthought to reflect greater sensitivity to dopaminergic blockade.

Table 1Clinical studies assessing the efficacy and tolerability of clozapine in early onset schizophrenia (in chronological order).

Study Sample Design Setting Duration Medication mean

dosage

In mg/day

Efficacy

measures

Adverse drugs reaction

(ADR)

Results summary

Short-term (up to 12 weeks)

Frazier et al., 1994 [16] n = 11a

Age: 14 (1.5)

OLRT Inpatient 6 weeks CLZ: 370.5

(range: 125–825)

BPRS, GAS,

SAPS, SANS

Hypersalivation 88%

Sedation 77%

Weight gain 77%

More than half of the

sample showed marked

improvement

Kumra et al., 1996 [26] n = 21a

Age: 14 (2.3)

DBRCT Inpatient 6 weeks CLZ: 176 (149)

HAL: 16 (8)

BPRS, CGAS

SAPS, SANS

Bunney-

Hamburg

Rating Scale

ADR profile of CLZ and

HAL were similar

except for insomnia

(more common with

HAL); drowsiness,

hypersalivation,

neutropenia more

common with CLZ

CLZ superior to HAL on all

measures of psychosis

Wudarsky et al., 1999 [47] n = 35a

Age: 14.1 (2.3)

Retrospective

observational

Inpatient 6 weeks CLZ: 325.4 (211)

OLZ: 17 (3.5)

HAL: 15.3 (8.23)

Not assessed Mean increase in

prolactin levels

HAL: 39 ng/ml

CLZ: 2 ng/ml

OLZ: 13.7 ng/ml

Prolactin levels remained

within the normal level

during CLZ treatment but

were elevated in 90% of HAL

and in 70% of OLZ treated

patients

Alfaro et al., 2002 [1] n = 40a

Age

HAL: 13.8 (1.5)

CLZ: 14.2 (2.3)

OLZ: 14.5 (3.2)

Retrospective

observational

Inpatient/

outpatient

6 weeks CLZ: 269.9 (173.3)

OLZ: 17.5 (2.8)

HAL: 15.4 (8.1)

Not assessed Prolactin levels (rates

of hyperprolactinaemia

not reported)

Significant correlations

between antipsychotic

plasma concentration and

prolactin levels only for the

OLZ treatment group; OLZ-

treated girls showed the

highest increase in

prolactin levels

Gerbino-Rosen et al.,

2005 [18]

n = 172

Age: 15.03 (2.13)

Retrospective

observational

Inpatient 8 weeks 431.4 (146.9) Not assessed Neutropenia: 13%

Agranulocytosis: 0.6%

half were successfully

rechallenged with CLZ

Incidence of

agranulocytosis similar to

that reported in the adult

literature

Shaw et al., 2006 [41] n = 25a

Age

CLZ: 11.7 (2.3)

OLZ 12.8 (2.4)

DBRCT Inpatient 8 weeks CLZ: 327 (113)

OLZ: 18.1 (4.3)

CGI-S

BPRS-24

SAPS, SANS

Weight gain (kg)

CLZ: 3.8

OLZ: 3.6

Hypertension,

tachycardia

(>100 beats/min),

enuresis

CLZ superior in efficacy

particularly for negative

symptoms; CLZ was

associated with higher

rates of hypertension,

tachycardia and enuresis

Sporn et al., 2007 [43] n = 54a

Age: 13.5 (2.5)

Re-analyses of

data from

previous CLZ-

treated

patients

assessed either

BD (n = 22) or

OL (n = 32)

studies

Inpatient 6 weeks CLZ: 298.2 (144.8) BPRS At 6 weeks

Tachycardia: 28%

Hypersalivation: 24%

Akathisia: 15%

Enuresis: 15%

70% of patients had

more than 1 ADR

Severity of illness at

baseline and improvement

during the first weeks of

CLZ treatment predicted

long-term response

C.

Schn

eider

et a

l. /

Eu

rop

ean

Psy

chia

try 2

9 (2

01

4)

1–

10

3

Table 1 (Continued )

Study Sample Design Setting Duration Medication mean

dosage

In mg/day

Efficacy

measures

Adverse drugs reaction

(ADR)

Results summary

Fleischhaker et al.,

2008 [14]

n = 45a

Age

CLZ: 17.4 (1.7)

OLZ: 15.7 (1.3)

RIS: 15.2 (2.8)

Prospective

observational

Inpatient 6 weeks CLZ: 294.9 (133.9)

OLZ: 16.1 (6.9)

RIS: 2.9 (1.5)

Not assessed Weight gain (kg)

CLZ: 2.5 (2.9)

OLZ: 4.6 (1.9)

RIS: 2.8 (1.3)

BMI increase (kg/m2)

CLZ: 0.8 (1)

OLZ: 1.6 (0.7)

RIS: 1.0 (0.5)

Weight gain observed for

all antipsychotics but

higher with OLZ

Kumra et al., 2008 [27] n = 39

Age

CLZ: 15.8 (2.2)

OLZ: 15.5 (2.1)

DBRCT Inpatient 12 weeks CLZ: 403.1 (201.8)

OLZ: 26.2 (6.5)

CGI, BPRS,

SANS, CGAS

Weight gain

Total serum cholesterol

Fasting triglycerides

Fasting blood glucose

Reduction in positive

symptoms was similar with

CLZ and OLZ but CLZ also

improved negative

symptoms; weight gain

higher with OLZ otherwise

no difference in ADRs

Hrdlicka et al., 2009 [22] n = 109

Age: 15.8 (1.6)

Retrospective

observational

Outpatient 6 weeks CLZ: 47.5 (118)

OLZ: 15 (6.1)

RIS: 2.7 (1.3)

HAL: 6.8 (1.1)

ZIP: 80 (0)

PER: 12 (6.9)

SUL: 50 (409.3)

Not assessed Weight gain (kg)

SGAs: 3.4 (3.2)

FGAs 2.0 (3.9)

No difference in weight

gain between the SGAs and

FGAs groups

Long-term (more than 12 weeks)

Schultz et al., 1996 [40] n = 40

Age

CLZ: 19.1 (2.2)

FGAs: 18.8 (2.3)b

OLRT Inpatient 36 weeks CLZ: 324

FGAs: 465b

BPRS, SAPS, SANS Prolactin levels Prolactin levels increased

with FGA but not with CLZ;

CLZ was superior in

improving positive and

negative symptoms

Turetz et al., 1997 [44] n = 11

Age: 11.3 (1.7)

OLRT Inpatient 16 weeks CLZ: 227.3 (34.4) CGI, BPRS, PANSS Sedation: 90%

Hypersalivation: 90%

Non-specific EEG

changes: 85%

Improvement in all

symptom scale scores; no

agranulocytosis

Wehmeier et al., 2004 [46] n = 36

Age range at

baseline: 9–21

Retrospective

observational

Not reported 2.5 to 79 months CLZ: 219.7

Range: 12.5–600

Not assessed Eosinophilia 66.7%

Elevated AST 58.3%

Elevated CK 52.7%

Elevated LDH 44.5%

Abnormal ECG 25%

No cases of myocarditis,

pericarditis, or

cardiomyopathy

Shaw et al., 2006 [41] n = 15a

Age at baseline

CLZ: 11.7 (2.3)

OLZ: 12.8 (2.4)

OL follow-up of

previous

randomised

trial of OLZ

andr CLZ

Outpatient 2–6 years Not reported CGI-S, BPRS-24,

SAPS, SAPS

Patients on CLZ showed

additional ADRs during

the follow-up period;

Lipid abnormalities

(n = 6) and seizure

(n = 1)

Clinical improvement was

sustained

Fleischhaker et al., 2006 [13] n = 51

Age at baseline:

16.1 (2.1)

Prospective

observational

Inpatient 26 months Mean dose at study

entry

CLZ: 321.9

(156.5)

OLZ: 16.6 (7.1)

RIS: 3.9 (1.7)

Not assessed Hypersalivation: 62.5%

Sedation: 56%

Weight gain: 56.3%

Constipation: 31.5%

OLZ associated with greater

weight gain; CLZ associated

with more sedation,

hypersalivation,

constipation and

hypotension

C.

Schn

eider

et a

l. /

Eu

rop

ean

Psy

chia

try 2

9 (2

01

4)

1–

10

4

Table 1 (Continued )

Study Sample Design Setting Duration Medication mean

dosage

In mg/day

Efficacy

measures

Adverse drugs reaction

(ADR)

Results summary

Sporn et al., 2007 [43] n = 35a

Age at follow-up:

19 (4.1)

OL Outpatient 2–6 years Mean dose at

follow-up: CLZ

360.3 (96.9)

CGAS Discontinuation due to

ADRs: 5.5%

Sustained clinical

improvement on CLZ;

further improvement was

seen in 56% of patients who

had poor response at 6

weeks

Fleischhaker et al.,

2008 [14]

n = 33a

Age at baseline

CLZ: 17.2 (1.6)

OLZ: 15.7 (1.3)

RIS: 14.3 (2.6)

Prospective

observational

Inpatient/

outpatient

45 weeks Mean dose at study

entry

CLZ: 311.7

(137.5)

OLZ: 10.2 (3.5)

RIS: 2.6 (1.7)

Not assessed Weight gain (kg)

CLZ: 9.5

OLZ: 16.2

RIS: 7.2

BMI increase (kg/m2)

CLZ: 2.9 (3.5)

OLZ: 5.2 (3.7)

RIS: 1.9 (1.3)

OLZ associated with greater

weight gain compared to

CLZ and RIS

Kim et al., 2008 [24] n = 26

Age at baseline:

14.4 (2.1)

Retrospective

observational

Outpatient 3.6 years Mean maintenance

dose of CLZ: 278.8

(122)

Number of

hospitalizations

per year

Hospital

days per year

Neutropenia

66.5% of males

7% of females

Discontinuation due

to ADRs after 3 years of

CLZ treatment: 6.5%

Patients had fewer hospital

days per year after CLZ

treatment

Age and dosage are shown as mean (standard deviation); weight gain and BMI are shown as mean; BMI (Body Mass Index); CLZ (clozapine); BPRS (Brief Psychiatric Rating Scale); CGAS (Child Global Assessment Scale); GCI-S (Global

Clinical Impression-Severity); DB (double-blind); DBRCT (double-blind randomised controlled trial); FGAs (first generation antipsychotics); GAS (Global Assessment Scale); HAL (haloperidol); OL (open label); OLRT (open label

randomised trial); OLZ (olanzapine); PANNS (Positive and Negative Syndrome Scale); PER (perphenazine); RIS (risperidone); SGAs (second generation antipsychotics); SANS (Schedule for the Assessment of Negative Symptoms);

SAPS (Schedule for the Assessment of Positive Symptoms); SUL (sulpiride); ZIP (ziprasidone).a Overlapping samples.b FGAs = haloperidol (n = 9), levomepromazine (n = 5), fluphenazine (n = 4), flupenthixol (n = 3), chlorprothixene (n = 2), promethazine (n = 1), perazine (n = 1) and thioridazine (n = 1) 80% of patients received more than one drug.

C.

Schn

eider

et a

l. /

Eu

rop

ean

Psy

chia

try 2

9 (2

01

4)

1–

10

5

Table 2Common adverse drug reactions to clozapine.

Adverse drug reaction Incidence (%) Period of highest risk Possible risk factors

Hypersalivation 80–90 Present throughout treatment

Sedation 56–90 Worse at initiation

;ay persist throughout treatment

High doses

Constipation 13–50 Present throughout treatment Low-fibre diet

Inadequate fluid intake

Lack of exercise

Concomitant use of anticholinergics

Enuresis 15–61 Highest risk at treatment initiation Childhood enuresis

Concomitant use of a second antipsychotic

Akathisia 15–31 Worse at initiation

May persist throughout treatment

Female sex

EEG abnormalities 10–60 Highest risk at initiation and dose escalation Pre-treatment seizures and/or EEG abnormalities

Brain lesions

Seizures 3 Highest risk after first year of treatment

Neutropenia 6–15 Highest risk during the 18 first weeks

Risk lower thereafter but persist

Female

Low baseline WBC counts

Ethnicity

Young age

Concomitant use of immunosuppressant drug

Weight gain 20–64 Risk increases with length of exposure Lack of physical activity

Race

Genetic

Metabolic abnormalities 8–22 Risk increases with length of exposure Weight gain

Genetic factors

Lack of physical activity

Diabetes 6 Risk increased with length of exposure Genetic factors

Weight gain or metabolic syndrome on previous antipsychotics

Tachycardia 35 Highest risk early in treatment

May persist

Rapid increase in doses

Hypotension 12.50 Highest risk early in treatment

Tolerance

Rapid increase in doses

Concomitant use of b blockers (e.g. for akathisia)

Hypertension 6 Highest risk early in treatment

Very common (� 1/10), common (� 1/100, < 1/10), uncommon (� 1/1000, < 1/100), rare (� 1/10,000, < 1/1000), very rare (< 1/10,000).

C. Schneider et al. / European Psychiatry 29 (2014) 1–106

3.2.1.4. Obsessive-compulsive symptoms. CLZ is associated withincreased prevalence of obsessive-compulsive symptoms [32] inadults. This association has not been systematically examined inEOS and requires clinical vigilance.

3.2.2. Cardiovascular effects

Clinically significant postural hypotension in CLZ-treated youthis frequent in the first fortnight. Tachycardia and hypertension arealso common early in the treatment and may persist [41,40,24]. Inadults, CLZ treatment is associated with an incidence ofmyocarditis of up to 3% [20]. The incidence of myocarditis inEOS is unknown. Although no cases were reported in the studiesreviewed here, there have been case reports [31], which suggestthat clinicians should remain vigilant when prescribing CLZ toyoung patients.

3.2.3. Metabolic side effects

3.2.3.1. Weight gain. Weight gain is one of the most commonadverse effects reported for SGAs in adult patients [35], and this isalso the case for the pediatric population [15,9]. With regards toCLZ, Fleischhaker et al. in 2008 [14] compared weight gain in 45youth aged 9 to 21 years, treated for 45 weeks with olanzapine,risperidone or CLZ. The average weight gain associated with CLZwas 2.5 � 2.9 kg at 6 weeks and 9.5 � 10.4 kg at 45 weeks. This wascomparable to the weight gain associated with risperidone at 45

weeks (7.2 � 5.3 kg) while the greatest weight gain over the sameperiod was seen with olanzapine (16.8 � 8.8 kg).

3.2.3.2. Laboratory changes in lipid, triglycerides and glucose. Inchildren and adolescents, the diagnosis of the metabolic syndromerequires at least three of the following: obesity (waist circumfer-ence > 90th percentile or BMI > 95th percentile), hypertriglycer-idemia (fasting serum triglyceride levels > 1.24 mmol/L [110 mg/dL]), low high-density lipoprotein (HDL) cholesterol levels (fastingHDL cholesterol < 1.0 mmol/L [40 mg/dL]), hypertension (bloodpressure > 90th percentile for age and sex) and hyperglycaemia(fasting glucose > 110 mg/dL) [15,9].

The prevalence of the metabolic syndrome in adults treatedwith CLZ around 50% [28]. Although, the rate of metabolicsyndrome in EOS is not known, abnormalities in lipid andglucose regulation have been reported. Hypertriglyceridemia isthe most frequent abnormality occurring in about 8-22% of CLZ-treated EOS patients [16,27]. The second most frequentabnormality is emergent diabetes, which occurs in about 6%of CLZ-treated youth [27]. Koller et al. identified all cases withhyperglycemia in children and adolescents treated with CLZthat were spontaneously reported to the Food and DrugAdministration between January 1993 and March 2001 [25].There were 11 reports of hyperglycemia in adolescents aged 13to 18 years (seven males and four females) who had been

C. Schneider et al. / European Psychiatry 29 (2014) 1–10 7

prescribed CLZ in daily doses from 100 to 1000 mg. Eight werenewly diagnosed cases, half of whom presented within the first6 weeks of treatment with further cases presenting over a 6-month period. CLZ was discontinued or the dose was decreasedin six patients.

3.2.4. Hematological effects

CLZ has greater propensity to cause serious hematologicaladverse events (HAEs) than any other antipsychotic [12]. The mostconcerning are neutropenia (absolute neutrophil count < 1,500/mm3) and agranulocytosis (absolute neutrophil count < 500/mm3).

3.2.4.1. Neutropenia. The annual incidence of CLZ-induced neu-tropenia in adults may vary from 2.3% [33] to 22% [23], if a moreliberal definition of neutropenia (absolute neutrophil count< 2000/mm3) is used. Sporn et al. in 2007 reported a 6% rate ofneutropenia based on data from 54 COS patients from the NIMHcohort, of whom 33 were evaluated over an average period of 4years (range 2–6 years) [43]. The prevalence of neutropenia overthe period of a year was estimated at 13% based on aretrospective chart review of 172 children and adolescentstreated with CLZ at the Bronx Children’s Psychiatric Center [18].In this study, half of the patients who experienced neutropeniawere successfully rechallenged with CLZ. In a similar retro-spective study of long-term CLZ treatment in 26 Korean childrenwith EOS, neutropenia developed in 34.6% of patients over a 2-year period; in all cases neutropenia was transient and CLZ wascontinued or re-instated successfully [24]. Overall, the datasuggest that neutropenia is relatively more common in EOS thanin adult patients but it is transient [16,40] and does not precludecontinued CLZ treatment [24].

3.2.4.2. Agranulocytosis. The risk of agranulocytosis in adultsranges between 0.5 and 1% and is not dose-dependent [33]. Therisk of agranulocytosis in EOS is comparable ranging from 0%[40,24] to 0.99% [18].

3.2.4.3. Fatalities. None of the studies reviewed reported any CLZ-related deaths in EOS.

3.2.5. Endocrinological effects

Unlikely, all other SGAs (olanzapine, risperidone, ziprasidoneand quetiapine) [15,9] CLZ is not associated with increased inprolactin levels [47,1].

3.2.6. Urinary side effects

Enuresis: nocturnal enuresis has been reported in up to 15% ofthe patients participating in the NIMH clinical trials [41] and up to61.5% of patients in observational trials [24].

3.2.7. Gastrointestinal effects

3.2.7.1. Hypersalivation. Hypersalivation is the most commonside-effect of CLZ, reported in 80–90% of EOS patients[16,26,43,44,13]. Hypersalivation may be dose-related, it usuallypersists throughout treatment and is more pronounced duringsleep [38].

3.2.7.2. Constipation. In adults, antipsychotic treatment is com-monly associated with constipation [10] which is particularlysevere with CLZ [10,36,37]. In rare cases, constipation mayprogress to paralytic ileus, faecal impaction, necrotizing colitisand intestinal perforation, all of which can be fatal [37]. The effecthas been attributed to the peripheral anticholinergic andantiserotinergic properties of CLZ. In EOS, constipation is commonand reported in about 30 to 50% of patients [13,24]. Although we

could not identify any cases of severe CLZ-related gastrointestinalADRs in EOS, clinicians should remain vigilant and consider thispossibility in patients presenting with constipation, abdominaldistension and pain [10,36,37].

3.3. Therapeutic Dose Monitoring

In adults, therapeutic drug monitoring is routinely performed toassess adherence and toxicity and to assist in judging therapeuticresponse based on plasma concentrations of CLZ and its majormetabolite, nor-clozapine. Conventionally, therapeutic plasma CLZconcentrations range between 350 and 600 ng/mL. A specific rangethat constitutes toxicity has not been established although the riskincreases with serum levels of 650 ng/mL. Couchman et al. in 2012published the most comprehensive review of therapeutic drugmonitoring for CLZ in young patients [8]. They examined 1408samples from the UK and Eire obtained between 1994 and 2010from 454 patients (males = 267; females = 187), aged 8–17 years.They found that plasma CLZ levels reflected the prescribed dose;median plasma CLZ increased with age up to 13 years but wasrelatively stable thereafter. Generally, given the same dose, plasmaCLZ levels were approximately 30% higher in girls than boys andhigher in those with lower body weight. Interestingly, thepercentage of patients smoking cigarettes increased from 10% inthose aged 13 years to 52% in those 17 or older. This is important asnon-smokers had on average 40% higher CLZ plasma levels thansmokers. The authors concluded that the determinants of CLZplasma concentration in EOS are similar to those reported in CLZ-treated adults.

3.4. Clinical guidelines for the use of CLZ in youths with schizophrenia

We identified two guidelines that were explicit in theirrecommendations regarding the use of CLZ in EOS. The AmericanAcademy of Child and Adolescent Psychiatry discusses CLZ in the‘‘Practice parameters for the assessment and treatment of childrenand adolescents with schizophrenia’’ [2] and the National Institutefor Health and Clinical Excellence in the ‘‘Recognition andManagement of Psychosis and schizophrenia in Children andAdolescents’’ [34]. Both guidelines recognise the superior efficacyof CLZ for treatment-refractory schizophrenia in youth. Because ofsignificant individual variability, they recommend that response toany antipsychotic should be judged after patients have beentreated at an adequate dose for at least 6 weeks. They alsorecommend CLZ only for patient who either failed to respond to atleast two therapeutic trials of other antipsychotics and/ordeveloped significant extrapyramidal side effects. The guidelinesoutline the level of monitoring for ADRs that is consideredappropriate for patients in this age group and make recommenda-tions regarding possible interventions. We have summarised thisinformation in Tables 3 and 4 and Fig. 1.

4. Discussion

The evidence presented here is consistent in highlighting thesuperior efficacy of CLZ in young patients with EOS. Symptomaticimprovement in CLZ-treated EOS is expected in most patients whohave failed to respond to other antipsychotics (Table 1). Moreover,after the initial response to CLZ further clinical improvement maybe seen in the subsequent 6–8 months [43]. The benefits of CLZtreatment are sustained during long-term maintenance based onstudies that have followed-up patients for periods of 2–9 years(Table 1).

Tolerability does not seem to present a particular challenge intreating young patients with CLZ and this is most convincinglydemonstrated by the low discontinuation rates (3–6%) [40,24]. CLZ

Fig. 1. Algorithm for clozapine treatment in youth with schizophrenia.

Recommendation based on references [2,34].

C. Schneider et al. / European Psychiatry 29 (2014) 1–108

Table 3Interventions to ameliorate clozapine related adverse drug reactions.

Adverse drug reaction Suggested interventions

Hypersalivation Chewing gum (sugar free) during the day

Anticolinergic drugs (hyiscine hydrobromide,

biperiden, trihexyphenidyl)

Ophthalmologic drops (atropine) as mouthwash

before bedtime

Raising the pillow during the night

Sedation Slow rate of titration during treatment initiation

Use minimally effective dose

Use single night time dose

Modafinil co-administration

Constipation Maintain fluid intake

Healthy, fibre-rich diet

Add bulk-forming laxatives

Enuresis Fluid restriction after 6.00 pm

Double-voiding before bedtime

Avoid drinks with caffeine or bladder irritants

(citrus or cranberry drinks)

Desmopressin nasal spray before bedtime

Akathisia Slow rate of titration during treatment initiation

Use minimally effective dose

Add long-acting benzodiazepines

EEG abnormalities/

Seizures

Use minimally effective dose

Add an anticonvulsant

Neutropenia Monitor and discontinue treatment if

persistent (see Table 4)

Add Lithium

Weight gain /

metabolic abnormalities

Dietary education and advice

Regular physical activity

Add Metformin

Add low dose Aripiprazole

Hypotension Slow rate of titration during treatment initiation

Use minimally effective dose

Hypertension Dietary education and advice

Table 4Neutropenia and agranulocytosis.

White Blood Cell

Count/mm3

Absolute Neutrophil

Count/mm3

Intervention

� 3500 � 2000 None

3000–3500 1500–2000 Continue clozapine treatment

Twice weekly blood sampling until

counts stabilise or increase

> 3000 > 1500 Stop clozapine treatment

Daily blood sampling until counts

normalise

Monitor for infection

C. Schneider et al. / European Psychiatry 29 (2014) 1–10 9

carries a higher risk of haematological ADRs compared to all otherantipsychotics which necessitates frequent monitoring. Thelogistics of arranging regular venopuncture and the discomfortassociated with this procedure are possibly greater barriers thanthe ADRs to prescribing CLZ. On the other hand, the necessity ofrigorous monitoring increases contact between patients, parentsand clinicians, which could enhance therapeutic relationships andpromote engagement.

Neutropenia was observed in up to 36% of patients prescribedCLZ [24] but it was commonly transient and did not requireintervention. In contrast, agranulocytosis was a rare event (< 0.6%)[18]. There is currently no evidence to suggest that EOS patients areat higher risk than adults with regards to agranulocytosis.Similarly, non-specific and clinically silent EEG changes werecommon while seizures were rare (< 3%) [41].

CLZ treatment was also associated with metabolic abnormal-ities but at a level comparable to olanzapine and otherantipsychotics SGAs [14,27,22,13,15,9].

Sedation and hypersalivation were observed in nearly everyEOS patient treated with CLZ [16,26,44,13]. Although not life-threatening these ADRs have a negative impact on patientexperience with CLZ.

Hyperprolactinaemie was not observed during CLZ treatment[47,1]. This is a distinct advantage of CLZ over all antipsychotics.Hyperprolactinaemia is a particular concern in young patientsbecause of its potential adverse effect on development includingheight, bone density, menstruation and sexual maturation.

Therapeutic drug monitoring has proved helpful in personalis-ing CLZ treatment in adults with schizophrenia. Available evidence[8] suggests that therapeutic drug monitoring could prove helpfulin EOS in establishing the optimal dose of CLZ in terms of risk–benefit ratio, and assessing adherence.

Evidence-based recommendations are helpful in supportingclinical decision-making but this should not diminish the value oflocal support. Clinical decision making with regards to CLZinitiation and monitoring can be enhanced through a variety ofmechanisms tailored to each clinical setting. These can take theform of second opinion assessments by colleagues, clinical casepresentations, and consultations with senior pharmacists andclinicians from other specialties. The latter is particularly useful inassessing and managing neurological, metabolic, endocrine andcardiological risk.

5. Conclusions

Systematic review of the evidence regarding the efficacy andtolerability of CLZ in EOS confirmed the superior efficacy of CLZ inpatients that had failed to respond to two previous trials ofantipsychotic medication. Most patients experienced multipleADRs but life-threatening events were infrequent and thediscontinuation rate was low. Fig. 1 synthesizes the availableinformation on screening and monitoring patients during CLZtreatment.

Disclosure of interest

CS, RC, DH and MK have no competing interests to report. SF hasreceived honoraria for her contribution to advisory meetings forEnzymotec and Janssen.

Acknowledgments

This review has been supported by funding from the EuropeanCommunity’s Seventh Framework Programme (FP7/2007–2013)under grant agreement no. 279227. The funding agency has had noinput in any aspect of data review, interpretation and manuscriptwriting.

References

[1] Alfaro CL, Wudarsky M, Nicolson R, Gochman P, Sporn A, Lenane M, et al.Correlation of antipsychotic and prolactin concentrations in children andadolescents acutely treated with haloperidol, clozapine, or olanzapine. J ChildAdolesc Psychopharmacol 2002;12:83–91.

[2] American Academy of Child Adolescent Psychiatry. Practice parameter for theassessment and treatment of children and adolescents with schizophrenia. JAm Acad Child Adolesc Psychiatry 2001;40(7 Suppl.):4–23.

[3] Armenteros JL, Davies M. Antipsychotics in early onset schizophrenia:systematic review and meta-analysis. Eur Child Adolesc Psychiatry2006;15:141–8.

[4] Cannon M, Jones P, Huttunen MO, Tanskanen A, Huttunen T, Rabe-Hesketh S,et al. School performance in Finnish children and later development ofschizophrenia: a population-based longitudinal study. Arch Gen Psychiatry1999;5:457–63.

C. Schneider et al. / European Psychiatry 29 (2014) 1–1010

[5] Cannon TD, Rosso IM, Bearden CE, Sanchez LE, Hadley T. A prospective cohortstudy of neurodevelopmental processes in the genesis and epigenesis ofschizophrenia. Dev Psychopathol 1999;11:467–85.

[6] Cirulli G. Clozapine prescribing in adolescent psychiatry: survey of prescribingpractice in in-patient units. Psychiatr Bull 2005;29:377–80.

[7] Citrome L. A systematic review of meta-analyses of the efficacy of oral atypicalantipsychotics for the treatment of adult patients with schizophrenia. ExpertOpin Pharmacother 2012;13:1545–73.

[8] Couchman L, Bowskill SV, Handley S, Patel MX, Flanagan RJ. Plasma clozapineand norclozapine in relation to prescribed dose and other factors in patientsaged < 18 years: data from a therapeutic drug monitoring service, 1994–2010.Early Interv Psychiatry 2013;7:122–30.

[9] De Hert M, Dobbelaere M, Sheridan EM, Cohen D, Correll CU. Metabolic andendocrine adverse effects of second-generation antipsychotics in children andadolescents: a systematic review of randomized, placebo controlled trials andguidelines for clinical practice. Eur Psychiatry 2011;26:144–58.

[10] De Hert M, Hudyana H, Dockx L, Bernagie C, Sweers K, Tack J, et al. Second-generation antipsychotics and constipation: a review of the literature. EurPsychiatry 2011;26:34–44.

[11] Eggers C, Bunk D. The long-term course of childhood-onset schizophrenia: a42-year follow-up. Schizophr Bull 1997;23:105–17.

[12] Fitton A, Heel RC. Clozapine: a review of its pharmacological properties, andtherapeutic use in schizophrenia. Drugs 1990;40:722–47.

[13] Fleischhaker C, Heiser P, Hennighausen K, Herpertz-Dahlmann B, Holtkamp K,Mehler-Wex C, et al. Clinical drug monitoring in child and adolescent psychi-atry: side effects of atypical neuroleptics. J Child Adolesc Psychopharmacol2006;16:308–16.

[14] Fleischhaker C, Heiser P, Hennighausen K, Herpertz-Dahlmann B, Holtkamp K,Mehler-Wex C, et al. Weight gain in children and adolescents during 45 weekstreatment with clozapine, olanzapine and risperidone. J Neural Transm2008;115:1599–608.

[15] Fraguas D, Merchan-Naranjo J, Laita P, Parellada M, Moreno D, Ruiz-Sancho A,et al. Metabolic and hormonal side effects in children and adolescents treatedwith second-generation antipsychotics. J Clin Psychiatry 2008;69:1166–75.

[16] Frazier JA, Gordon CT, McKenna K, Lenane MC, Jih D, Rapoport JL. An open trialof clozapine in 11 adolescents with childhood-onset schizophrenia. J Am AcadChild Adolesc Psychiatry 1994;33:658–63.

[17] Freedman JE, Wirshing WC, Russel AT, Palmer Bray M, Jergen U. Absence statusseizures after successful long-term clozapine treatment of an adolescent withschizophrenia. J Child Adolesc Psychopharmacol 1994;4:53–62.

[18] Gerbino-Rosen G, Roofeh D, Tompkins DA, Feryo D, Nusser L, Kranzler H, et al.Hematological adverse events in clozapine-treated children and adolescents. JAm Acad Child Adolesc Psychiatry 2005;44:1024–31.

[19] Gochman P, Miller R, Rapoport JL. Childhood-onset schizophrenia: the chal-lenge of diagnosis. Curr Psychiatry Rep 2011;13:321–2.

[20] Haas SJ, Hill R, Krum H, Liew D, Tonkin A, Demos L, et al. Clozapine-associatedmyocarditis: a review of 116 cases of suspected myocarditis associated withthe use of clozapine in Australia during 1993–2003. Drug Saf 2007;30:47–57.

[21] Hafner H, Nowotny B. Epidemiology of early-onset schizophrenia. Eur ArchPsychiatry Clin Nerurosci 1995;245:80–92.

[22] Hrdlicka M, Zedkova I, Blatny M, Urbanek T. Weight gain associated with atypicaland typical antipsychotics during treatment of adolescent schizophrenic psy-choses: a retrospective study. Neuro Endocrinol Lett 2009;30:256–61.

[23] Hummer M, Kurz M, Barnas C, Saria A, Fleischhacker WW. Clozapine-inducedtransient white blood count disorders. J Clin Psychiatry 1994;55:429–32.

[24] Kim Y, Kim BN, Cho SC, Kim JW, Shin MS. Long-term sustained benefits ofclozapine treatment in refractory early onset schizophrenia: a retrospectivestudy in Korean children and adolescents. Hum Psychopharmacol 2008;23:715–22.

[25] Koller E, Malozowski S, Doraiswamy PM. Atypical antipsychotic drugs andhyperglycemia in adolescents. JAMA 2001;286:2547–8.

[26] Kumra S, Frazier JA, Jacobsen LK, McKenna K, Gordon CT, Lenane MC, et al.Childhood-onset schizophrenia. A double-blind clozapine-haloperidol com-parison. Arch Gen Psychiatry 1996;53:1090–7.

[27] Kumra S, Kranzler H, Gerbino-Rosen G, Kester HM, DeThomas C, Cullen K, et al.Clozapine versus ‘‘high-dose’’ olanzapine in refractory early-onset schizophre-nia: an open-label extension study. J Child Adolesc Psychopharmacol 2008;18:307–16.

[28] Lamberti JS, Olson D, Crilly JF, Olivares T, Williams GC, Tu X, et al. Prevalence ofthe metabolic syndrome among patients receiving clozapine. Am J Psychiatry2006;163:1273–6.

[29] Lay B, Blanz B, Hartmann M, Schmidt MH. The psychosocial outcome ofadolescent onset schizophrenia: a 12-year follow-up. Schizophr Bull2000;26:801–16.

[30] Maziade M, Bouchard S, Gingras N, Charron L, Cardinal A, Roy M-A, et al. Long-term stability of diagnosis and symptom dimensions in a systematic sample ofpatients with onset of schizophrenia in childhood and early adolescence. II:positive/negative distinction and childhood predictors of adult outcome. Br JPsychiatry 1996;169:371–8.

[31] Merrill DB, Dec GW, Goff DC. Adverse cardiac effects associated with clozapine.J Clin Psychopharmacol 2005;25:32–41.

[32] Mukhopadhaya K, Krishnaiah R, Taye T, Nigam A, Bailey AJ, Sivakumaran T,et al. Obsessive-compulsive disorder in UK clozapine- treated schizophreniaand schizoaffective disorder: a cause for clinical concern. J Psychopharmacol2009;23:6–13.

[33] Munro J, O’Sullivan D, Andrews C, Arana A, Mortimer A, Kerwin R. Activemonitoring of 12,760 clozapine recipients in the UK and Ireland. Beyondpharmacovigilance. Br J Psychiatry 1999;175:576–80.

[34] National Institute for Health and Clinical Excellence. Psychosis and schizo-phrenia in children and young people, recognition and management, 2013.(http://www.nice.org.uk/nicemedia/live/12951/60475/60475.pdf).

[35] Newcome JW. Second-generation (atypical) antipsychotics and metaboliceffects: a comprehensive literature review. CNS Drugs 2005;19(Suppl. 1):1–93.

[36] Palmer SE, McLean RM, Ellis PM, Harrison-Woolrych M. Life-threateningclozapine-induced gastrointestinal hypomotility: an analysis of 102 cases.Clin Psychiatry 2008;69:759–68.

[37] Peyriere H, Roux C, Ferard C, Deleau N, Kreft-Jais C, Hillaire-Buys D, et al.Antipsychotics-induced ischaemic colitis and gastrointestinal necrosis: areview of the French pharmacovigilance database. Pharmacoepidemiol DrugSaf 2009;18:948–55.

[38] Praharaj SK, Arora M, Gandotra S. Clozapine-induced sialorrhea: pathophysi-ology and management strategies. Psychopharmacology (Berl) 2006;185:265–73.

[39] Ropcke B, Eggers C. Early-onset schizophrenia: a 15-year follow-up. Eur ChildAdolesc Psychiatry 2005;14:341–50.

[40] Schulz E, Fleischhaker C, Remschmidt HE. Correlated changes in symptomsand neurotransmitter indices during maintenance treatment withclozapine or conventional neuroleptics in adolescents and youngadults with schizophrenia. J Child Adolesc Psychopharmacol 1996;6:119–31.

[41] Shaw P, Sporn A, Gogtay N, Overman G, Greenstein D, Gochman P, et al.Childhood-onset schizophrenia. A double-blind, randomized clozapine-olan-zapine comparison. Arch Gen Psychiatry 2006;63:721–30.

[42] Sikich L, Frazier JA, McClellan J, Findling RL, Vitiello B, Ritz L, et al. Double-blindcomparison of first and second-generation antipsychotics in early-onsetschizophrenia and schizo-affective disorder: findings from the treatment ofearly-onset schizophrenia spectrum disorders (TEOSS) study. Am J Psychiatry2008;165:1420–31.

[43] Sporn AL, Vermani A, Greenstein DK, Bobb AJ, Spencer EP, Clasen LS, et al.Clozapine treatment of childhood-onset schizophrenia: evaluation of effec-tiveness, adverse effects, and long-term outcome. J Am Acad Child AdolescPsychiatry 2007;46:1349–56.

[44] Turetz M, Mozes T, Toren P, Chernauzan N, Yoran-Hegesh R, Mester R, et al. Anopen trial of clozapine in neuroleptic-resistant childhood-onset schizophre-nia. Br J Psychiatry 1997;170:507–10.

[45] Vyas NS, Hadjulis M, Vourdas A, Byrne P, Frangou S. The Maudsley early onsetschizophrenia study. Predictors of psychosocial outcome at 4-year follow-up.Eur Child Adolesc Psychiatry 2007;16:465–70.

[46] Wehmeier PM, Schuler-Springorum M, Heiser P, Remschmidt H. Chart reviewfor potential features of myocarditis, pericarditis, and cardiomyopathy inchildren and adolescents treated with clozapine. J Child Adolesc Psychophar-macol 2004;14:267–71.

[47] Wudarsky M, Nicolson R, Hamburger SD, Spechler L, Gochman P, Bedwell J,et al. Elevated prolactin in pediatric patients on typical and atypicalantipsychotics. J Child Adolesc Psychopharmacol 1999;9:239–45.