Management of clozapine-resistant schizophrenia

The problem of delay

One of the most modifiable contributors to poor outcomes is delayed clozapine initiation (Shah et al., 2018). Despite strong evidence supporting clozapine’s superior efficacy in TRS, real-world data consistently demonstrate often a 10-year delay between meeting criteria for TRS and commencing clozapine (Thien and O’Donoghue, 2019). This delay is clinically and functionally consequential. Earlier clozapine introduction is associated with higher response rates, and retrospective data suggest a potential ‘critical treatment window’ – approximately within 3 years of TRS onset – during which the likelihood of meaningful response is substantially higher (Yoshimura et al., 2017).

Response rates of 75% have been reported when clozapine is introduced early as a third-line treatment in FES cohorts (Agid et al., 2011), notably higher than rates typically seen in multi-episode, delayed-treatment samples (Siskind et al., 2017). Conversely, longer delays are associated with poorer outcomes (Griffiths et al., 2021). Although duration of active psychosis (DAT) and delay to effective treatment are distinct constructs from CRS, the convergent implication is that earlier effective intervention is associated with improved long-term trajectory (Crowley et al., 2025). It is plausible that longer DAT will impact clozapine response and contribute to CRS. Delayed clozapine may not merely reflect severity; it may contribute to development of clozapine resistance.

Defining true clozapine resistance

Before diagnosing CRS, clinicians must exclude pseudoresistance. International consensus criteria emphasise three requirements: confirmed adherence, adequate duration (at least 12 weeks, and longer – up to 4 months – for negative symptoms) and adequate clozapine exposure (plasma concentration ⩾350 ng/mL) (Howes et al., 2017 Wagner et al., 2020).

These requirements emphasise the need to confirm adequate clozapine exposure before attributing non-response to true pharmacological resistance rather than pseudoresistance and highlight the importance, where available, of implementing clozapine TDM early in clozapine treatment (Flanagan et al., 2023a). TDM is therefore foundational. Yet TDM remains underutilised, contributing to subtherapeutic dosing, suboptimal clozapine concentrations and pseudoresistance rather than true clozapine resistance.

Most patients respond within trough plasma clozapine concentrations of 350–600 ng/mL (Flanagan et al., 2023a; Siskind et al., 2021), with a concentration of 372 ng/mL as a point of optimal sensitivity and specificity for response prediction (Northwood et al., 2023), but variability is considerable.

Where TDM is unavailable, clinicians may need to rely on structured clinical assessment to guide dose optimisation. Careful evaluation of adherence, symptom response and dose-related adverse effects may help estimate adequate exposure. Clinical markers such as new or worsening dose-related sedation, hypersalivation, orthostatic hypotension, constipation or myoclonus may indicate elevated clozapine concentrations or approaching tolerability thresholds, although these remain imperfect surrogates for plasma concentrations and should not replace TDM where available.

Identification and management of pharmacokinetic factors that reduce clozapine plasma concentrations are essential to optimising treatment response. Smoking status is a particularly important modifiable factor. Cigarette smoking induces CYP1A2, reducing clozapine concentrations; dose adjustments are often required in smokers, and careful reduction is needed after cessation to avoid toxicity (Flanagan et al., 2023b). Other medications affecting CYP1A2 or CYP3A4 similarly alter plasma levels and require proactive management. Notably, CYP1A2 inhibitors fluvoxamine and ciprofloxacin can significantly raise clozapine levels, while fluoxetine can inhibit CYP2D leading to milder increases in clozapine levels.

Other important kinetic factors affecting clozapine metabolism include systemic inflammation or infection. Acute inflammatory states and infections, often accompanied by elevated C-reactive protein (CRP), can suppress CYP1A2 activity and produce rapid rises in clozapine concentrations, increasing the risk of toxicity (Flanagan et al., 2023a). Emerging sedation, delirium, seizures, cardiovascular adverse effects or marked hypersalivation during infection should prompt urgent clinical review and consideration of dose reduction and repeat TDM where available (de Leon et al., 2020).

Adverse effects also limit adequate dosing. Sedation, constipation, weight gain and cardiometabolic disturbance frequently restrict titration (Flanagan et al., 2020). Early identification and targeted management of adverse effects can facilitate safe clozapine dose escalation, improve adherence and enable attainment of therapeutic plasma clozapine concentrations (Tanzer et al., 2024). Proactive management of side effects improves adherence and enables attainment of therapeutic concentrations. Without optimising exposure and tolerability, a diagnosis of CRS may be premature.

Clinical response may also be influenced by comorbid conditions, including depressive disorders and substance use, which can adversely affect clozapine adherence, symptom burden and outcomes. Broader psychosocial factors, such as environmental deprivation and limited social support, may further impede recovery and should be addressed within a comprehensive multidisciplinary treatment framework.

Dose optimisation and supratherapeutic strategies

If persistent symptoms occur despite sustained concentrations above 350 ng/mL for at least 12 weeks (Howes et al., 2017; Sanahan et al., 2025), cautious dose escalation towards the upper reference range (500–600 ng/mL) is a reasonable next step. Assessment at higher therapeutic concentrations should allow a further period of up to 8–12 weeks (Fabrazzo et al., 2002), as clozapine response is associated with both plasma concentration and duration of treatment.

Evidence suggests a plateau effect beyond 600–800 ng/mL, with diminishing returns and increased adverse effects (Northwood et al., 2023; Remington et al., 2013), particularly seizures and cardiotoxicity, at concentrations of 800–1000 ng/mL (Northwood et al., 2023). Observational data indicate that some individuals may benefit from concentrations in the 600–800 ng/mL range (Northwood et al., 2023; Remington et al., 2013; Yada et al., 2021), but improvements above 600 ng/mL are inconsistent, and may be associated with increased adverse effects especially dose-related side effects such as seizure and sedation without increased clinical benefit. If pursued, supratherapeutic dosing should occur with close monitoring – electrocardiogram (ECG) monitoring, cardiometabolic management and vigilance for myoclonus, seizures and severe constipation. There is an absence of data on time to response at supratherapeutic concentrations, but in practice, a further period of assessment with tolerability of 8–12 weeks is suggested.

Older adults may achieve therapeutic or supratherapeutic plasma concentrations at substantially lower clozapine doses due to reduced clearance and age-related pharmacokinetic changes, necessitating more cautious dose escalation and closer monitoring for adverse effects.

Any dose adjustments aimed at achieving supratherapeutic clozapine plasma concentrations, or the introduction of augmentation strategies, should be accompanied by structured pre- and post-intervention assessment of clinical symptoms. Standardised rating instruments such as the Positive and Negative Syndrome Scale (PANSS) or the Brief Psychiatric Rating Scale (BPRS) are recommended to systematically monitor clinical response and inform ongoing treatment decisions. If there is inadequate clinical response at supratherapeutic plasma concentrations, dose reductions to obtain plasma concentrations of 350–600 ng/mL are indicated. The key principle is structured escalation with explicit safety thresholds, rather than indefinite dose increases without clear therapeutic aims.

Augmentation strategies

While CRS definitions enhance research comparability and clinical clarity, they provide limited guidance regarding optimal augmentation strategies. When optimised clozapine fails to produce an adequate clinical response, augmentation strategies are commonly used. Evidence for augmentation strategies in CRS is mixed, with no universally accepted approach. In the following sections, we expand on the recommendations outlined in the 2026 ANZJP guideline (Suetani et al., 2026). Figure 1 provides recommendations for the management of CRS.

OPEN IN VIEWER

Figure 1.

Clinical recommendations on the management of clozapine resistance in schizophrenia.

Antipsychotic augmentation

Adding a second antipsychotic is common in practice, occurring in up to half of clozapine-treated cases (Preobrazenski et al., 2025). Meta-analyses demonstrate at best small effect sizes, without consistently replicated robust symptom efficacy for any agent (Etchecopar-Etchart et al., 2024; Galling et al., 2017; Lally and Gaughran, 2019; Siskind et al., 2018; Wagner et al., 2019). Nonetheless, expert consensus supports cautious use in selected patients with persistent positive symptoms (Wagner et al., 2020). Antipsychotic medication with complementary receptor activity might be considered to augment clozapine response in treating positive symptoms. Amisulpride and aripiprazole are most often recommended, where tolerability permits, particularly at low-to-medium doses (Wagner et al., 2020). Large Nordic cohort studies suggest that low to medium-dose aripiprazole (9 mg to <16.5 mg/day) augmentation may reduce relapse-related hospitalisation by 20–30% compared with clozapine monotherapy, whereas higher doses of augmenting antipsychotics (including aripiprazole) are associated with increased relapse risk and rehospitalisation (Tiihonen et al., 2025). These findings underscore an important theme: more medication is not necessarily better. Targeted, modest augmentation may be preferable to high-dose polypharmacy.

For negative symptoms, longer clozapine trials (⩾4 months) are advised before declaring resistance (Wagner et al., 2020). Low-dose aripiprazole, amisulpride or cariprazine may be considered.

Antidepressants

Antidepressant augmentation, particularly with selective serotonin reuptake inhibitors (SSRIs) such as fluoxetine, shows small but statistically significant improvements in overall symptom severity in meta-analyses (Siskind et al., 2018). However, the quality of evidence is limited by small sample sizes and short trial durations. A bi-national cohort study from Sweden and Finland found reduction in psychiatric rehospitalisation with lower doses of sertraline, duloxetine and escitalopram (Taipale et al., 2025). Reassuringly, antidepressant augmentation at these lower doses was not associated with higher rates of physical health hospitalisation. However, it remains unclear whether these reductions of psychiatric hospitalisation reflect treatment of comorbid depression rather than CRS. These agents may be most rational when depressive symptoms are prominent.

Mood stabilisers

Evidence for anticonvulsant augmentation in CRS remains mixed. Sodium valproate has shown potential benefits for total psychotic symptoms and aggression in systematic reviews, but controlled CRS-specific evidence is weak (Siskind et al., 2018; Sommer et al., 2012). Meta-analyses report inconsistent findings for lamotrigine, with both positive (Grover et al., 2023) and negative (Zheng et al., 2017) results compared with placebo. Topiramate has demonstrated superiority over placebo (Grover et al., 2023; Zheng et al., 2017), but its use is limited by poor tolerability and significantly increased all-cause discontinuation (Zheng et al., 2017). Evidence supporting lithium for symptom augmentation is limited. In practice, it is more commonly co-prescribed to address low neutrophil counts in an attempt to avoid clozapine discontinuation due to neutropenia (Verdoux et al., 2024); however, this represents an unlicensed and potentially unwarranted use.

Valproate co-prescription requires caution, particularly early in clozapine treatment, given associations with myocarditis and neutropenia (Malik et al., 2017; Vickers et al., 2022). Valproate is associated with weight gain, further complicating the metabolic picture of people using clozapine (Verrotti et al., 2011), and has also been associated with cognitive impairment (Alsukhni et al., 2023). The teratogenic risk associated with valproate further restricts its clinical use (Alsdorf and Wyszynski, 2005). Given these risks, its use among people on clozapine is questionable. If clinicians chose to augment clozapine with valproate, close laboratory monitoring and monitoring for myocarditis is mandatory.

Across augmentation strategies, a central principle applies, trials should be time-limited and systematically evaluated. If no meaningful improvement occurs, the intervention should be discontinued rather than continued indefinitely.

Electroconvulsive therapy and neurostimulation

Clozapine augmentation with electroconvulsive therapy (ECT) is endorsed by expert consensus and highlighted in the ANZJP guideline (Suetani et al., 2026). A prospective randomised study demonstrated significant symptom reduction with clozapine plus ECT compared with clozapine alone (Petrides et al., 2015). Meta-analyses report response rates in CRS exceeding 50%, particularly for persistent positive symptoms (Grover et al., 2023; Lally et al., 2016b). However, most studies are small, heterogeneous and include few sham-controlled trials. There remains a substantial paucity of high-quality evidence regarding ECT use in schizophrenia, including TRS and CRS. Available network meta-analytic data have not demonstrated a statistically significant difference in efficacy between ECT augmentation of antipsychotics (including clozapine) and sham ECT combined with antipsychotics in TRS (Wei et al., 2025). However, these findings require cautious interpretation given substantial clinical and methodological heterogeneity. Included studies comprised TRS and CRS patients, employed variable ECT protocols and utilised differing outcome measures (Wei et al., 2025).

The optimal number of ECT treatments in schizophrenia remains uncertain. A meta-analysis in CRS identified a mean of 11.3 treatments (Lally et al., 2016b), with evidence of clinical response plateauing at 12–16 treatments (Petrides et al., 2015). Although schizophrenia is often considered to require a higher number of ECT treatments to achieve a response than depressive disorders, this is comparable to the mean of 10.6 ECT treatments reported across all indications in a large observational cohort, in which 92% of cases involved depressive disorders (Buley et al., 2017).

ECT response is more robust for positive symptoms than for negative domains. Observational data suggest better outcomes with fewer prior antipsychotic trials and shorter episode duration (Grover et al., 2023).

Repetitive transcranial magnetic stimulation (rTMS) has been studied with mixed results in CRS. Evidence remains limited and inconsistent, precluding firm clinical recommendations (Wei et al., 2025).

Psychological and social interventions

As the ANZJP GRADE Guidelines highlight, psychological and social interventions are important aspects of management of people with schizophrenia, even, or especially, those with TRS and CRS (Suetani et al., 2026).

Cognitive-behavioural therapy for psychosis (CBTp) has modest evidence in schizophrenia overall. While the largest randomised trial in clozapine-resistant patients did not demonstrate significant symptom improvement over treatment as usual (Morrison et al., 2018), a 2023 meta-analysis showed some small improvements in positive symptoms, with low risks associated with the intervention (Todorovic et al., 2023).

Family interventions, social skills training, metacognitive therapy, and rehabilitation approaches lack CRS-specific evidence but remain important components of comprehensive care, particularly for adherence and functional recovery.

A structured approach to CRS

Management of CRS requires a systematic and multidisciplinary approach. First, confirm clozapine adherence and therapeutic exposure through TDM. Optimise dose within safe limits, address smoking and drug interactions, and proactively manage adverse effects. Second, consider cautious dose escalation guided by TDM towards the upper therapeutic range and if inadequate response, consider a trial of clozapine at higher plasma concentration 600–800 ng/mL based on response and tolerability. Third, evaluate targeted augmentation – antipsychotic, antidepressant or ECT – based on symptom profile, tolerability, comorbidity and risk–benefit assessment.

Augmentation should never substitute for optimisation, nor continue without demonstrable benefit. Equally, proactive management of clozapine-associated adverse effects, psychosocial interventions and management of comorbid conditions should proceed in parallel.

Given the complexity and burden of CRS, shared decision making and psychoeducation are essential throughout treatment planning. Discussions regarding higher clozapine dosing and concentrations, augmentation strategies or ECT should include clear communication regarding the expected magnitude of benefit, uncertainty of evidence, potential adverse effects, monitoring requirements and alternative approaches. Involving patients, families and carers where appropriate may improve engagement, adherence and alignment of treatment decisions with individual goals and preferences.

Ultimately, the most effective strategy to reduce CRS may lie earlier in the treatment pathway: prompt recognition of TRS and timely initiation of clozapine once resistance criteria are met. In this sense, clozapine resistance should be understood not solely as a pharmacological phenomenon but also a systems-level challenge reflecting delays in recognition, access and implementation of evidence-based care.