Clozapine is the only evidence-based antipsychotic indicated for treatment-resistant psychosis.¹ Clozapine has proven efficacy in the reduction of recurrent suicidal ideation in patients with schizophrenia and schizoaffective disorder.² A study of the death registers in Finland, spanning from 1996 to 2006, indicated that their clozapine-prescribed population had overall lower all-cause mortality rates when compared with the population on alternative antipsychotics. This benefit was attributed to clozapine’s reduction in suicidality.³ Clozapine is indicated for the management of tardive dyskinesia,² and has proven efficacy in treating anxiety, irritability, aggression and negative symptoms of schizophrenia,⁴ and benefits in cognitive functioning.⁵

Clozapine has an extensive receptor binding and pharmacological profile, responsible for both its therapeutic efficiency and wide range of potential side effects.⁶ With some exceptions, most clozapine-related side effects occur within 3 to 6 months of beginning treatment. Rarely clozapine discontinuation may be required for severe or life-threatening side effects. Indications for discontinuation, without re-challenge, include agranulocytosis (absolute neutrophil count [ANC] < 0.5 × 10⁹/L)⁷ myocarditis, cardiomyopathy and prolonged cardiac ventricular repolarisation (QTc) interval (> 500 ms). Ileus, neuroleptic malignant syndrome, venous thromboembolism and hyperglycaemic emergencies are also side effects requiring clozapine discontinuation. However, in such cases, once the patient is medically stabilized, clozapine may be re-challenged. Discontinuation of clozapine however does increase the risk of relapse, aggression and suicidality, and as such careful consideration is taken when planning intervention in the management of side effects.⁸ Most clozapine-related side effects do not warrant discontinuation and commonly attenuate with time or dose reduction.⁹ In a recent systematic quantitative meta-review, clozapine proved to have superior outcomes with regard to hospitalisations, discontinuation rates and mortality when compared with other antipsychotics.¹⁰

Regarding clozapine’s more commonly occurring side effects, prevalence rates of metabolic syndrome range between 11% and 64% in clozapine users, compared with a prevalence of 20% in first-episode schizophrenia patients, pre-treatment. Weight gain is experienced in up to 35% of patients, at an average of 4.4 kg over 10 weeks. Gastrointestinal side effects include hypersalivation, constipation, gastrointestinal reflux disease, nausea and vomiting, and occur more frequently with clozapine compared with other antipsychotics. Tachycardia is the most common cardiac side effect, with a prevalence of 25%–35%. The historically synonymous clozapine-related side effect, neutropenia, however, has an overall risk of about 3%.⁹

Many of clozapine’s side effects are dose-dependent, yet because of significant inter-individual variation in clozapine metabolism, as well as drug–drug interactions,¹¹ some patients experience side effects at much lower doses than others. Clozapine is metabolised by the hepatic cytochrome P-450 isoenzyme system, specifically CYP1A2 and CYP3A4, and has active metabolites, norclozapine and clozapine-N-oxide. It has an extensive list of pharmacokinetic drug interactions, resulting in inhibition and induction of this system and altering the plasma concentration of the active metabolites. Consequently, co-administered medications may affect clozapine plasma concentration. Furthermore, certain genetic polymorphisms of cytochrome enzymes are associated with changes in clozapine clearance. Other factors such as gender, age and cigarette smoking can also affect the pharmacokinetics, and thus plasma concentration of clozapine.^12,13 Therapeutic drug monitoring is a tool used in some settings to optimise pharmacotherapy,¹⁴ but is not currently routinely performed in South Africa (SA).

Although the side effect burden of clozapine is high, side effects can generally be managed by dose reduction, dose-splitting or with pharmacological intervention.¹⁵ A recent audit performed in the United Kingdom (UK) indicated that two-thirds of patients on clozapine were on further medications to manage side effects.¹⁶ Regarding polypharmacy in schizophrenia, aside from medications added to manage side effects or comorbid diseases, patients may also be treated with multiple psychotropic medications. Antipsychotic polypharmacy is concerning because of the lack of standardised research describing efficacy and safety, and unclear practice patterns.¹⁷ The benefit of polypharmacy must always outweigh the risks,¹⁷ which include increased pill burden, increased risk of side effects and drug–drug interactions, increased risk for accidental medication errors, or non-adherence by virtue of complicated treatment schedules, as well as higher economic expense.¹⁸ Although polypharmacy should generally be avoided, in the case of clozapine, augmentation strategies and combination therapies may improve patient outcomes,¹⁸ and minimise the need for clozapine discontinuation. Despite clozapine monotherapy being strongly advised,¹⁸ evidence to support the so-called ‘clozapine combination therapy’ is growing.¹⁵

With regard to patterns of clozapine use worldwide, despite clozapine being well-recognised as the gold standard antipsychotic, a retrospective study performed in the United States of America (USA) from 2002 to 2005 indicated clozapine prescription rates were low for the population of patients who met criteria for clozapine trial.¹⁹ In the 2018–2019 National Clinical Audit of Psychosis in the UK, it was found that only 54% of patients meeting the criteria for treatment resistance were offered clozapine. Based on this finding, the Healthcare Quality Improvement Partnership recommends that reasons for not prescribing clozapine be documented in records and advised that mental health pharmacists should assist with identifying patients who are eligible for clozapine.¹ A cohort study performed in Denmark assessing the global assessment of the functioning of patients initiated on clozapine between 2004 and 2011 found that there was moderate to substantial functional improvement in more than 24% of clozapine users, but that chance of substantial improvement in females decreased by 15% for each year that clozapine initiation was delayed.²⁰

Clozapine is broadly underutilised and, based on evidence to support its efficacy, this is a public health dilemma. While a study on international clozapine trends across 17 countries found that clozapine prescription is on the increase, the countries that were included were high-income, and there remains little data on clozapine use in the third world.²¹ Compounding the aforementioned, there is limited research describing the landscape of psychiatric illness and the use of clozapine in our setting. In a study performed in the Western Cape, clozapine use in Xhosa-speaking patients diagnosed with schizophrenia or schizoaffective disorder was low, at 10%, with a rate of antipsychotic polypharmacy almost three times higher, at 28%.²² Small studies on clozapine users have described the prevalence of clozapine-related side effects,²³ and monitoring patterns of patients on clozapine.²⁴ However, research describing clozapine prescribing patterns in SA is sparse.

This study aimed to explore the magnitude of clozapine use, and the demographic, clinical and prescription characteristics of inpatients treated with clozapine, at a specialised psychiatric hospital in Johannesburg, SA.

The study objectives were:

A third (33.2%) of all the patients admitted to the biological wards at Tara Hospital over the study period received a trial of clozapine. The percentage of in-patients diagnosed with schizophrenia or other psychotic disorders that were treated with clozapine was not calculated for the purpose of this study, but rather the percentage of all the biological admissions, representing a patient population inclusive of all severe mental illnesses, not only psychotic disorders. Nonetheless, this result demonstrates that clozapine is not uncommonly used in the biological wards, which was not an unexpected finding as Tara Hospital is a medium to long-term treatment facility, likely to manage a greater proportion of patients who have responded poorly to standard first-line treatment and are treatment-resistant.

Treatment-resistant schizophrenia (or psychosis), defined as ‘the persistence of positive symptoms despite ≥ 2 trials of adequate dose and duration of antipsychotic medication with documented adherence’,²⁵ was found to be the indication for clozapine use in almost all of the patients in this study. Furthermore, evidence of delayed initiation of clozapine was noticed, with 48% of the patients with treatment resistance having had between three and five documented antipsychotic trials prior to clozapine. Delays in the appropriate commencement of clozapine lead to poorer long-term outcomes.²⁶ Barriers to clozapine use have been described in three categories: patient and medication-related, clinician-related and health system-related. Patient and medication-related barriers include non-compliance with blood testing schedules, intolerance of side effects and polypharmacy. Clinician-related barriers refer to a lack of knowledge and inexperience in identifying suitable clozapine candidates, prescribing clozapine and managing side effects, as well as the need for close monitoring and perceived poor adherence. Health system barriers refer to resource limitations and care continuity concerns regarding a more complex monitoring protocol. These barriers impact both clozapine initiation and continuation.²⁷ However, in a study performed on psychiatric outpatients in KwaZulu-Natal (KZN), overall high rates of encouraging attitude and knowledge towards clozapine were reported by patients.²⁸ Another study, conducted in Australia, compared the subjective experience of clozapine side effects and outcomes of clozapine-treated patients, with the perceived experience as assessed by their clinicians. In this study, stable clozapine users self-reported greater satisfaction with clozapine compared with what their clinicians believed them to be experiencing. Furthermore, only 19% of users reported disdain at blood tests, compared with the clinician estimate of 52%.²⁹

Regarding the characteristics of clozapine users, the KZN study found a male and black race predominance of 67% and 71%.²⁸ similar to the demographics of the population in this study (67.6% and 89.2%, respectively). Rates of pre-existing diagnoses of hypertension (7.8%) and diabetes mellitus type 1 and 2 (3%) in this study were lower compared with that of the general SA adult population, where hypertension rates of 44% (males) and 46% (females), and diabetes mellitus rates of 8% (males) and 13% (females) have been found.³⁰ Likewise, with regard to pre-existing dyslipidaemia (5.9%) in this study. Little data exist on the prevalence of dyslipidaemia in the SA general population, but a study performed in Mpumalanga found a rate of 67%, however, with less than 1% being diagnosed and on treatment.³¹ These study findings were noteworthy, as not only higher rates of medical comorbidity (and reduced life expectancy) have been found in patients with schizophrenia but also increased risk of undetected (and inadequately treated) medical illness relative to the general population.³² Cigarette smoking was found to be highly prevalent in this study (48%) compared with the SA general population, where rates have been reported at 8% (females) and 37% (males),³⁰ which was not a surprising finding, as smoking prevalence is known to be up to fivefold greater among patients with schizophrenia.³³

High rates of polypharmacy were found across both groups in this study. Psychiatric polypharmacy, the prescription of two or more psychotropic medications, occurs commonly and can be described as same-class, multi-class, adjunctive, augmentation and total polypharmacy.³⁴ Regarding antipsychotic polypharmacy specifically, this was common in both the discharged-on-clozapine and clozapine-discontinued groups (65.9% and 50.0%). In the discharged-on-clozapine group, the most common second antipsychotic was flupenthixol decanoate (30.7%), followed by amisulpride (23.9%). Despite foundational evidence behind clozapine combinations growing, it is not currently sufficient to delineate a treatment schedule. Although combination strategies may not lead to substantial clinical improvement, they can allow for clozapine dose reduction, and may thus benefit tolerability.¹⁵ A significantly greater use of flupenthixol decanoate was found in the clozapine-discontinued group (64.3%; p = 0.0322), compared with the discharged-on-clozapine group, which may also explain the significantly greater use of anticholinergics in the clozapine-discontinued group (p = 0.0474). This trend is echoed in a systematic review and meta-regression of global and regional trends of antipsychotic polypharmacy, which found antipsychotic polypharmacy significantly associated with the following: inpatient status, diagnosis of schizophrenia, greater long-acting injectable antipsychotic prescription and anticholinergics.³⁵ There were no significant differences with regard to the use of other psychiatric and non-psychiatric medications between groups. The most frequently prescribed non-psychiatric medications were stool softeners or laxatives (33.0%), beta-blockers (29.5%) and lipid-lowering agents (21.8%) in the discharged-on-clozapine group and anticholinergics (34.7%), lipid-lowering agents (28.6%) and beta-blockers (21.4%) in the clozapine-discontinued group.

The most common side effects found in this study were weight gain, tachycardia and constipation. A recent UK-based data review of electronic health records of clozapine users from 2007 to 2016 similarly showed high rates of weight gain (56%), tachycardia (25%) and constipation (25%), with other notable side effects including hypersalivation (48%), sedation (46%) and insomnia (33%).³⁶ As a result of the high rates of polypharmacy in this study, it was not possible to attribute side effects exclusively to clozapine. Aside from medications, other considerations are of multi-causative processes in the development of side effects such as dyslipidaemia and include sedentary lifestyle, poor diet, age and genetics.³⁷ The discharged-on-clozapine patients had significantly greater weight gain, as well as increased rates of lipid-lowering agent initiations during the course of their admission. However, there was no significant difference between the overall rate of lipid-lowering prescriptions (i.e. for both pre-existing and newly diagnosed dyslipidaemias) in both groups. There were no cases of agranulocytosis, myocarditis, cardiomyopathy, prolonged QTc or clozapine-associated deaths in this population, all side effects, which are known to be rare.⁸ Regarding neutropenia, this occurred in two patients (2.3%) from the discharged-on-clozapine group, and three patients from the clozapine-discontinued group, accounting for 21.4% of clozapine discontinuations.

Consensus on the threshold for treatment interruption because of neutropenia is not conclusive. According to USA guidelines, the threshold is < 1.0 × 10⁹/L, or < 0.5 × 10⁹/L for patients with benign ethnic neutropenia (BEN), compared with the UK, which uses < 1.5 × 10⁹/L, or < 1 × 10⁹/L in cases of BEN.³⁸ Benign ethnic neutropenia, which occurs in up to a third of individuals of African and Middle Eastern descent, is well-described and poses no increased risk for infection.^9,39 Larger epidemiological studies have described BEN as an important contributor to under-prescription and early discontinuation of clozapine. Consequently, lower thresholds for clozapine initiation and discontinuation have been recommended in such patients.³⁹ ‘Morning pseudoneutropenia’, caused by diurnal variation in circulating white blood cells, may also lead to unnecessary clozapine discontinuations.¹¹

The rate of clozapine discontinuation in this study was relatively low (13.7%); however, only in-patient discontinuations were reviewed. A retrospective medication utilisation review, performed in the Eastern Cape, of a small cohort of patients discharged on clozapine found discontinuation rates of 29% and 35% at 3 and 6 months, respectively.²⁴ Worsening of obsessive-compulsive symptoms, found in one clozapine-discontinued patient in this study, is a rare side effect that has been found to be associated with clozapine, as well as with other atypical antipsychotics.⁴⁰ Only one patient had a potentially life-threatening side effect, namely ileus, which was resolved upon medical management and discontinuation of clozapine. There were no fatalities. Clozapine doses in this study were all less than the maximum licenced dose for clozapine.¹⁵