Treatment-resistant schizophrenia

INTRODUCTION — Antipsychotic medication is first-line treatment for schizophrenia. Most patients show substantial improvement in psychotic symptoms in response to antipsychotics; however, for many, improvement is insufficient to meet stringent criteria for remission, and a substantial proportion experience residual treatment-resistant symptoms.

This topic addresses the evaluation and management of individuals with treatment-resistant schizophrenia. The epidemiology, pathogenesis, clinical manifestations, course, assessment, diagnosis, and treatment of schizophrenia are reviewed separately, as are the guidelines for prescribing clozapine. Treatment of schizophrenia with co-occurring disorders is also reviewed separately.

●(See "Schizophrenia in adults: Epidemiology and pathogenesis" and "Schizophrenia in adults: Maintenance therapy and side effect management".)

●(See "Schizophrenia in adults: Clinical features, assessment, and diagnosis".)

●(See "Schizophrenia in adults: Epidemiology and pathogenesis" and "Schizophrenia in adults: Maintenance therapy and side effect management".)

●(See "Schizophrenia in adults: Pharmacotherapy with long-acting injectable antipsychotic medication".)

●(See "Schizophrenia in adults: Guidelines for prescribing clozapine".)

●(See "Co-occurring schizophrenia and substance use disorder: Epidemiology, clinical features, assessment, and diagnosis".)

EPIDEMIOLOGY — A sizeable proportion of patients with schizophrenia do not respond sufficiently to antipsychotics [1]. Additionally, epidemiologic data suggest that the diagnosis of treatment-resistant schizophrenia may often be overlooked [2,3] and is often made years after the criteria were met [2]. (See 'Early diagnosis' below.)

A meta-analysis (50 studies, n = 23,390), examining the prevalence of treatment-resistant schizophrenia reported the following [4]:

●Overall prevalence of treatment-resistant schizophrenia among patients with schizophrenia spectrum disorder (ie, schizophrenia, schizoaffective disorder, schizotypal personality disorder) was 36.7 percent (95% CI 33.1-40.5). The prevalence in studies rated as low risk of bias was 28.4 percent (95% CI 23.9-33.0).

●Among individuals experiencing a first episode psychosis, 22 percent (95% CI 18.4-25.8) had treatment-resistant schizophrenia. The incidence was higher, 39.5 percent (95% CI 32.2-47) among individuals with multiple episodes of psychosis.

●Primary treatment resistance (ie, no antipsychotic response) appears to be more frequent than later onset treatment resistance (24 versus 9 percent). However, short duration of follow-up may have led to underestimate of treatment resistance in longitudinal studies.

DIAGNOSING TREATMENT-RESISTANT SCHIZOPHRENIA — To make a diagnosis of treatment-resistant schizophrenia, we review the chart thoroughly and confirm treatment history with the patient and available collateral sources. Prior to establishing a diagnosis of treatment-resistant schizophrenia, we exclude factors that may impede clinical response (eg, pseudoresistance), and confirm adequacy of medication trials [5-10]. (See 'Pseudoresistance' below and "Schizophrenia in adults: Guidelines for prescribing clozapine".)

Early diagnosis — We are proactive in diagnosing treatment-resistant schizophrenia. There is increasing evidence suggesting that delay of a diagnosis of treatment-resistant schizophrenia and clozapine initiation may have a negative impact on the likelihood that clozapine will help.

As examples:

●In a multicenter retrospective cohort study, individuals starting clozapine later in the course of illness were more likely to be rehospitalized after initiating clozapine that those starting earlier in illness (hazard ratio 2.94, 95% CI 1.10-8.55) [11]. Similarly, in another observational trial, including 159 subjects with schizophrenia or schizoaffective disorder, individuals who achieved remission with clozapine treatment, as compared with those who did not achieve remission, had shorter time periods and fewer number of antipsychotic treatments prior to clozapine initiation [12].

●In a Danish national database study, among 633 patients with schizophrenia who were starting clozapine, a shorter time to admission was predicted by the each of the following, as compared with those without these factors [13]:

•Greater number of different antipsychotics used (hazard ratio 1.1, 95% CI 1.0-1.2)

•Greater number of prior admissions (hazard ratio 1.04, 95% CI 1.03-1.05)

•Lower clozapine dose (hazard ratio 0.07/100 mg, 95% CI 0.03-0.13/100 mg)

Additionally, the study suggested that a lower number of psychiatric hospitalization and antipsychotic trials prior to clozapine initiation appears to be associated with greater clozapine response.

Prioritization of clozapine for individuals deemed unlikely to respond to a non-clozapine antipsychotic has advanced but is not yet ready for clinical practice. Neuroimaging biomarkers have demonstrated promise, but still lack the robust validation necessary for their clinical use [14]. Individual prediction of response with only clinical data has shown promise. For instance, in a robust validation process, using age, sex, ethnicity, triglycerides, alkaline phosphatase levels, and lymphocyte counts, treatment response predictions showed accuracy above chance, but remained below the 80 percent that is generally recommended for individual decision making [14,15].

Additionally, ethnic disparities in access to clozapine, may perpetuate health disparities [16,17].

Pseudoresistance — Pseudoresistance is treatment nonresponse due to reasons other than medication nonresponse. Each of the following may be considered a cause of pseudoresistance and are assessed for and addressed as part of the determination of treatment-resistant schizophrenia.

Misdiagnosis — We reevaluate the primary diagnosis in patient who are unresponsive to antipsychotic treatment. The differential diagnosis and diagnostic evaluation of patients presenting with schizophrenia-like symptoms is discussed separately. (See "Schizophrenia in adults: Clinical features, assessment, and diagnosis", section on 'Diagnosis'.)

Presence of co-occurring disorders — Co-occurring mental disorders, substance use disorders, and medical conditions contribute to the illness burden of patients with schizophrenia; if untreated, these conditions can impede its effective treatment.

●Mental disorders – Mental disorders such as depression, obsessive-compulsive disorder, and anxiety may occur concurrently with schizophrenia. Identification and management of mental disorders commonly co-occurring with schizophrenia are discussed separately. (See "Anxiety in schizophrenia" and "Depression in schizophrenia" and "Management of obsessive-compulsive disorder in adults".)

●Substance use disorders – The identification and management of substance use disorders co-occurring with schizophrenia are discussed separately. (See "Co-occurring schizophrenia and substance use disorder: Epidemiology, clinical features, assessment, and diagnosis".)

●Other medical conditions – We complete a detailed past medical history and physical examination in evaluating for other medical conditions that may be limiting response to treatment. We refer for a complete medical evaluation by a consulting clinician in unclear cases or in those with prior history or abnormalities on investigation. As an example, obesity, often leading to sleep apnea, is common in chronically ill patients with schizophrenia. Sleep apnea can affect response to treatment and can worsen symptoms of fatigue and irritability. We ask about snoring and interrupted breathing and may refer patients with obesity for medical evaluation for sleep apnea. (See "Clinical presentation and diagnosis of obstructive sleep apnea in adults".)

Medication-related effects — Medication-related side effects and interactions with other medications may be a cause of pseudoresistance. Additionally, we review for nonadherence to medication as a cause of pseudoresistance. These are addressed below and elsewhere. (See "Schizophrenia in adults: Maintenance therapy and side effect management", section on 'Side effect management'.)

●Antipsychotic adverse effects – Side effects can mimic ongoing agitation or negative symptoms and can lead to functional disability and a greater tendency for ongoing psychopathology and relapse [18]. Management of side effects in individuals on antipsychotic medications for schizophrenia is discussed elsewhere. (See "Schizophrenia in adults: Maintenance therapy and side effect management", section on 'Side effect management'.)

●Medication interactions – We perform a thorough review of potential drug-drug interaction. Efficacy of a specific medication may be affected by other prescribed medications or over the counter agents [19].

Examples include:

•Interactions between carbamazepine and antipsychotic medications may lead to reduced levels of antipsychotic medications.

•Smoking stimulates cytochrome p450 1A2 (CYP1A2) and may lead to reduced levels of olanzapine and clozapine.

•Estrogens, contraceptive pills, obesity, caffeine and acute inflammation may increase levels of antipsychotic medications metabolized by CYP1A2, but especially of clozapine that has a narrow therapeutic index [20].

The drug interactions tool can be used to assess potential drug interactions.

●Medication nonadherence – Until proven otherwise, we consider nonadherence as a potential reason for nonresponse to medication. Nonadherence to antipsychotic medication is often overlooked by patients and clinicians, yet it is a common reason for pseudoresistance. In one cross-sectional study, it was reported that 35 percent of individuals provisionally diagnosed with treatment-resistant schizophrenia had a subtherapeutic antipsychotic plasma level [21-24].

The use of LAI medications, supervised medication intake (typically with an orally dissolving or liquid formulation) and digital medicine approaches (eg, medication with a sensor that signals to a torso-worn patch) are other considerations [25-28]. These are discussed elsewhere. (See "Schizophrenia in adults: Maintenance therapy and side effect management", section on 'Side effect management' and "Schizophrenia in adults: Pharmacotherapy with long-acting injectable antipsychotic medication".)

●Adequacy of trial – We consider an adequate trial to include at least six weeks at the maximally tolerated dose within the medication's recommended therapeutic dose range. As clinical trials do not show convincing efficacy of dose recommendation above those recommended by the manufacturer, we do not consider supratherapeutic dosing to be a necessary component in determining treatment-resistant schizophrenia [29-32]. (See "Schizophrenia in adults: Maintenance therapy and side effect management", section on 'Treatment-resistant schizophrenia'.)

Discussion of medication adjustments for poor response to antipsychotic treatment is discussed elsewhere. Antipsychotic doses in the treatment of schizophrenia are on the tables (table 1 and table 2). (See "Schizophrenia in adults: Maintenance therapy and side effect management", section on 'Medication adjustments'.)

Optimize nonpharmacologic treatment — Prior to diagnosing treatment resistance, we optimize non pharmacologic treatment.

●Triggers and stressors – We identify and address triggers for worsening symptoms of schizophrenia (eg, stressors that may be interfering with effective treatment). Additionally, we review the potential impact of the support system and engage them in treatment when possible. (See "Schizophrenia in adults: Clinical features, assessment, and diagnosis", section on 'Factors influencing course'.)

●Psychosocial interventions – We suggest psychosocial intervention as an adjunct to antipsychotic medications. Although psychosocial interventions given alone are not sufficiently effective in schizophrenia, providing them as an adjunct to antipsychotic medication has been found to improve patient outcomes.

Psychosocial interventions studied in the treatment of schizophrenia include cognitive-behavioral therapy, social skills training, family psychoeducational interventions, assertive community treatment, and crisis intervention [33-41]. These are discussed elsewhere. (See "Schizophrenia in adults: Psychosocial management" and "Assertive community treatment for patients with severe mental illness".)

Diagnostic criteria — Varying definitions of treatment-resistant schizophrenia have posed challenges interpreting studies evaluating its treatment. We prefer the use of the optimal criteria for treatment-resistant schizophrenia; however, when this is not possible (eg, absence of standardized scales defining treatment responsiveness longitudinally, plasma levels not drawn), the minimal Treatment Response and Resistance in Psychosis (TRRIP) criteria are acceptable.

Minimal and optimal diagnostic criteria for treatment-resistant schizophrenia are found below [42,43]:

●Minimal TRRIP criteria for treatment-resistant schizophrenia – Based on cross-sectional clinical assessment supplemented by collateral sources of information (eg, medical record documentation, caregiver's report), patients should meet each of the following:

•Diagnosis – The American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision (DSM-5-TR) diagnosis of schizophrenia. (See "Schizophrenia in adults: Clinical features, assessment, and diagnosis", section on 'Diagnosis'.)

•Symptom severity – At least moderate symptom severity (>3 in psychotic symptom items) as rated using a standardized scale (eg, Positive and Negative Syndrome Scale [PANSS] or Brief Psychiatric Rating Scale [BPRS]).

•Functional impairment – At least moderate impairment measured using a validated scale (eg, Social and Occupational Functioning Assessment Scale).

•Prior treatment – At least two trials of ≥6 weeks at a therapeutic dose (equivalent to ≥600 mg chlorpromazine) with adherence ≥80 percent of prescribed doses.

●Optimal TRRIP criteria for treatment-resistant schizophrenia – The optimal TRRIP criteria include the minimal criteria (above) with the addition of:

•Prospective evaluation of symptom severity using a standardized scale (eg, PANSS or BPRS) confirming <20 percent symptom reduction over six weeks of treatment and persistence of at least moderate positive symptom severity and moderate functional impairment.

•One of the two antipsychotic trials should be a long-acting injectable antipsychotic (table 3). (See "Schizophrenia in adults: Pharmacotherapy with long-acting injectable antipsychotic medication".)

•Antipsychotic adherence should be confirmed by ≥2 unannounced antipsychotic plasma levels.

Although some have advocated reducing the number of antipsychotic trials required by the TRRIP Working Group definition from two to one before clozapine is used [44,45], we believe this recommendation is premature, given the limitations of the supporting data and the risks of agranulocytosis and other serious side effects with clozapine [46].

CLOZAPINE AS FIRST-LINE TREATMENT — Our preference for patients with treatment-resistant schizophrenia is treatment with clozapine rather than other medications or interventions. (See 'Diagnosing treatment-resistant schizophrenia' above.)

Eligibility — To be eligible for a clozapine trial, a patient with treatment-resistant schizophrenia must meet the following criteria (see 'Diagnostic criteria' above):

●Absolute neutrophil count ≥1500 cells/microliter (or >1000 in individuals diagnosed with benign ethnic neutropenia [47]

●Determination by the clinician as well as the patient and family if possible that the benefits of clozapine outweigh the risks

●Ability to adhere to treatment monitoring, including regular blood draws

●Patient/family agreement

Prescribing clozapine — We treat with clozapine for at least 24 weeks and confirm adequate exposure by assuring that plasma levels are above 350 ng/dL. Studies confirm that clinical improvement can continue for at least this duration [5,48]. However, while studies confirm that clozapine blood level is optimal at >350 ng/dL, some patients may need levels >550 ng/dL [49].

Guidelines for prescribing clozapine, including medical contraindications, pharmacology, dosing, monitoring, and adverse effects are described separately. (See "Schizophrenia in adults: Guidelines for prescribing clozapine".)

Efficacy — Randomized trials have shown that clozapine has greater efficacy, as compared with other antipsychotics, in the treatment of patients with schizophrenia who have responded poorly to prior antipsychotic trials. Among individuals with treatment-resistant schizophrenia, response rates to clozapine have been estimated to be between 40 and 50 percent [9].

Support for the use of clozapine comes from head-to-head studies and network meta-analyses of comparative effectiveness [5,6,8-10], as well as from nationwide naturalistic studies [50] and a meta-analysis of naturalistic studies [51] with more generalizable patient populations than in randomized controlled trials.

●In a meta-analysis including 25 trials and 2364 subjects with treatment-resistant schizophrenia, the efficacy of clozapine was compared with other first- or second-generation antipsychotic medications [5]. When all time frames were combined, subjects treated with clozapine experienced greater clinical improvement in total symptoms as compared with those treated with other antipsychotics (standardized mean difference -0.29, 95% CI -0.49 to -0.09). When only short-term trials were included in the analysis (eight studies, n = 1218), the improvement with clozapine, as compared with other antipsychotics, was more robust, (standardized mean difference -0.39, 95% CI -0.61 to -0.17) and individuals treated with clozapine were more likely to meet variably described threshold for response to treatment (relative risk 1.17, 95% CI 1.07-2.73). The absolute risk reduction was 12.48 percent (95% CI 7.52-17.43). Furthermore, clozapine had a greater effect on positive symptoms at all times points, as compared with negative symptoms where benefits were seen in the short-term trials only. (See "Schizophrenia in adults: Clinical features, assessment, and diagnosis", section on 'Positive symptoms' and "Schizophrenia in adults: Clinical features, assessment, and diagnosis", section on 'Negative symptoms'.)

●In a meta-analysis including 63 studies (n = 109,341, study duration 19 months) clozapine treatment, as compared with other antipsychotic treatment, was associated with lower hospitalization risk (19 studies, n = 49453; relative risk 0.82, 95% CI 0.72-0.92; number needed to treat [NNT] 18, 95% CI 12-40) and all-cause discontinuation (16 studies, n = 56,368; relative risk 0.732, 95% CI 0.64-0.84; NNT 8, 95% CI 6-12) [51]. This superiority was observed despite greater illness severity in patients prescribed clozapine. Moreover, clozapine was also associated with better outcomes regarding overall symptom reduction (standardized mean difference -0.3, 95% CI -0.6 to -0.03) and Clinical Global Impressions scale severity (standardized mean difference -1.2, 95% CI -2.2 to -0.12). Individuals continuously treated with clozapine also have been observed to have lower mortality rates compared with those treated with other antipsychotics (mortality rate ratio 0.56, 95% CI 0.36-0.85) [52].

●In a network meta-analysis of 60 studies involving 6838 participants with treatment-resistant schizophrenia, clozapine and olanzapine were more effective in improving overall symptoms of schizophrenia than risperidone, haloperidol, fluphenazine, sertindole, chlorpromazine, and quetiapine (range of mean standardized mean differences -0.11 to -0.48) [10]. Treatment with either clozapine or olanzapine led to similar effects (standardized mean difference -0.05, 95% CI -0.21 to 0.11) and were associated with increased weight gain (both clozapine and olanzapine) or greater sedation (clozapine) as compared with other antipsychotics. The lack of superiority of clozapine versus olanzapine in this meta-analysis, however, has been put in question, due to heterogeneity in studies (eg, differences in defining treatment resistance between studies) [7,53]. Observational data have demonstrated the superiority of clozapine over olanzapine with regards to all-cause discontinuation (clozapine versus olanzapine, n = 27,212; relative risk 0.792, 95% CI 0.669-0.939) [51].

Furthermore, data from an adjusted within-individual analysis of nationwide Finnish registry data (n = 2250) appear to support clozapine as most effective and safest treatment option in patients with schizophrenia who had been using clozapine for ≥1 year before clozapine treatment cessation for undefined reasons [54]. In that study, the risk of psychosis-related hospitalization (adjusted hazard ratio 0.49, 95% CI 0.40-0.61) and the risk of mortality (adjusted hazard ratio 0.18, 95% CI 0.09-0.36) as compared with placebo were found to be lowest for reinitiation of clozapine. The risk of treatment failure (ie, switching, addition of another antipsychotic, and/or stopping the current antipsychotic) was lowest for aripiprazole long-acting injectable (adjusted hazard ratio 0.42, 95% CI 0.27-0.65; p<0.0001) followed by clozapine (adjusted hazard ratio 0.49, 95% CI 0.43-0.57).

Although evidence exists for the efficacy of electroconvulsive therapy (ECT) for treatment-resistant schizophrenia [55], we prefer clozapine over ECT (see 'Electroconvulsive therapy' below). Our preference is based on well controlled trials, the need for general anesthesia for ECT, potential adverse effects of ECT, and limited availability of ECT and maintenance ECT in some locations. Additionally, clozapine is the only approved treatment for treatment-resistant schizophrenia.

Discontinuation — Just as we do with beginning a clozapine trial, we use shared decision making in the decision to discontinue clozapine. We discuss the risks and benefits with the patient and, with permission, chosen family members or other informed individuals. Benefits of clozapine may include patient satisfaction, increased well-being, and functionality that are not always as obvious to clinicians as they might be to patients and family members. We aim to confirm that the benefits of clozapine are sufficiently clear to justify continuation of the drug.

PATIENTS INELIGIBLE FOR OR REFUSING CLOZAPINE — Our preference for treatment of individuals with treatment-resistant schizophrenia who are ineligible for or refuse clozapine treatment is discussed here. Several alternatives to clozapine have been tested, however most of the findings are mixed or from trials with methodologic limitations [56-64]. Nonetheless, these treatments are used by patients who have responded inadequately to prior treatments and have a clinical status and health-related quality of life that is sufficiently poor that the patient, family, and treatment team have decided that the treatments are justified despite the very limited evidence of efficacy.

Prior to considering alternative treatments we discuss the viability of clozapine and other treatment options with the patient and authorized family members or supports [47]. (See 'Diagnosing treatment-resistant schizophrenia' above.)

Electroconvulsive therapy — When clozapine is not an option, our preference is treatment with an eight-week trial of electroconvulsive therapy (ECT) augmentation of a non-clozapine antipsychotic. The following data appears to support the use of ECT as an alternative to clozapine treatment in treatment-resistant schizophrenia when clozapine is not an option:

●In a meta-analysis of 11 randomized clinical trials including 818 participants with treatment-resistant schizophrenia, treatment with ECT augmentation of a non-clozapine antipsychotic was compared with treatment with a non-clozapine antipsychotic alone [56]. ECT augmentation of a non-clozapine antipsychotic was superior to antipsychotic alone, for total symptom improvement (standardized mean difference -0.67, 95% CI -0.95 to -0.39), and study defined response (relative risk 1.48, 95% CI 1.24-1.77), with a number needed to treat (NNT) between four and nine. The most frequent side effects of ECT were headache and memory impairment.

●In an open label trial, 60 individuals with treatment-resistant schizophrenia were randomly assigned to ECT (acute course followed by 24 weeks of maintenance) or to clozapine. While both groups showed improvements in PANSS scores, individuals randomized to ECT treatment had a greater improvement in PANSS score than those treated with clozapine, at week 6 (nine points, p<0.001) week 12 (eight points, p<0.001) and 24 weeks (6.5 points, p<0.001) [57]. However, trial study design (open label) and size limit generalizability of findings.

For patients who do not respond to an eight-week trial of ECT augmentation other medication strategies are discussed below. Limited data support their efficacy. (See 'Medication strategies' below and "Overview of electroconvulsive therapy (ECT) for adults".)

Medication strategies

Augmentation of non-clozapine antipsychotic — For individuals who do not respond to an eight-week trial of ECT augmentation, we suggest an 8-week trial of lamotrigine augmentation. If this is unsuccessful, we try sequential trials of topiramate and minocycline augmentation.

However, augmentation of a non-clozapine antipsychotic with another medication has not shown convincing evidence of efficacy. Our preference is based on indirect evidence.

In a meta-analysis of 381 trials, while initial analysis identified 42 antipsychotic augmentation strategies that were superior to placebo for measures of total psychopathology (14 trials), positive symptoms (six trials), negative symptoms (four trials), methodologic concerns including small sample, lack of replication, heterogeneity, and publication bias limit interpretation. Furthermore, an inverse relationship was found between the quality of the meta-analyzed clinical trials and the effect size of any advantage [58]. This study quantified the effect size of the augmentation strategies, using standardized mean difference, as well as the risk of bias for that publication, using a systematic assessment. Considering both the effect size and the quality of the data from which the effect size emerged, these were the top three interventions with best balance.

●Lamotrigine (typical dose 100 to 200 mg/day) – Among three trials including 98 patients, lamotrigine augmentation superior to placebo in reducing symptoms of antipsychotic-treated schizophrenia (standardized mean difference -0.73, 95% CI -1.26 to -0.20 [58].

●Minocycline (typical dose 100 to 200 mg/day) – Meta analysis of eight trials including 548 patients with schizophrenia (on antipsychotic medication), adjunctive minocycline, as compared with placebo led to a greater reduction in total psychopathology (standardized mean difference -0.64, 95% CI -1.02 to -0.27) [59].

●Topiramate (typical dose 100 to 400 mg/day) – Among 13 trials including 651 patients, topiramate augmentation led to greater symptom reduction than comparator (placebo augmentation or antipsychotic monotherapy; standardized mean difference -0.58, 95% CI -0.82 to -0.35) [47,56-59,65].

Neuromodulation — Transcranial magnetic stimulation (TMS) is a treatment option for treatment-refractory schizophrenia when other medications strategies and ECT are not an option or have failed.

Short-term clinical trials have found mixed results regarding the efficacy of repetitive TMS (rTMS) [60-62,66,67]. Most studies were of short-term rather than long-term treatment. Heterogeneity in patient groups and treatment modalities among various studies limit firm conclusions.

In a meta-analysis of 22 randomized trials including 827 patients with schizophrenia, rTMS applied to the frontal cortex reduced negative symptoms in patients with schizophrenia already being treated with antipsychotic drugs compared with antipsychotic treatment alone with an effect size = 0.64 (95% CI 0.32-0.96) [60]. Trials including younger patients found larger effect sizes compared with trials including older patients.

However, in a separately published meta-analysis including 30 trials and 768 patients with schizophrenia, rTMS (as compared with sham TMS) reduced hallucinations and negative symptoms but was associated with a modest trend toward worsening of some positive symptoms [61]. Other trials have not shown a benefit for rTMS [62,66].

Side effects of rTMS include pain at the site of stimulation, muscle twitching during treatment sessions, posttreatment headache and toothache, and, rarely, seizures. We do not treat patients with a seizure disorder with rTMS.

A study examining the effects of transcranial direct current stimulation, a neuromodulatory intervention that uses direct current, as adjunctive treatments in schizophrenia found a nonsignificant trend towards reducing positive and overall symptoms [61].

TREATMENT OF CLOZAPINE-RESISTANT SCHIZOPHRENIA — We define clozapine resistance as the failure to show adequate improvement after a 24-week trial of clozapine. Among individuals with treatment-resistant schizophrenia, up to half of the patients do not respond to treatment with clozapine [5,6,9,48]. (See "Schizophrenia in adults: Guidelines for prescribing clozapine".)

Augmentation with electroconvulsive therapy — For clozapine-resistant patients with schizophrenia, we favor electroconvulsive therapy (ECT) augmentation of clozapine rather than medication augmentation [55,58,63,64,68].

In a meta-analysis (18 trials, n = 1769) of treatments for clozapine resistant schizophrenia, ECT augmentation of clozapine was superior to clozapine alone regarding symptom improvement at and endpoint assessment (mean 5.3 weeks; standardized mean difference -1.44, 95% CI -2.05 to -0.84) and rate of remission at endpoint assessment (23.6 percent versus 13.3 percent, relative risk 1.80, 95% CI 1.39 to 2.35; p<0.0001, NNT 14, 95% CI 6-50) [55]. Patient-reported memory impairment and headache each occurred more frequently in individuals treated with adjunctive ECT than those that did not receive adjunctive ECT (24 versus 0 percent and (14 versus 2 percent, respectively). In one eight week trial included in the meta-analysis, 39 individuals with treatment-resistant schizophrenia were assigned to continue clozapine or to receive bilateral ECT (two to three session/week) along with continuing clozapine [63]. Patients treated with both ECT and clozapine were more likely to meet criteria for clinical response (≥40 percent reduction in psychosis) compared with patients receiving clozapine alone (50 versus 0 percent). The control group demonstrated a similar response to ECT upon crossing over to receive ECT after the randomized phase was completed. Two patients required the postponement of an ECT session because of mild confusion.

However, two small but sham-controlled trials of ECT in patients with clozapine-resistant schizophrenia [69,70] and a post hoc power calculation of a prior trial [63] did not replicate prior strong findings in favor of ECT.

Augmentation with medication — Clinical trials of the medications below have very limited evidence of efficacy and multiple methodologic limitations, and therefore should be used only if ECT augmentation is not viable.

●Non-clozapine antipsychotic – We add a non-clozapine antipsychotic to clozapine for individuals with residual symptoms despite clozapine treatment and after augmentation with ECT has been ineffective. Our preference is augmentation of clozapine with aripiprazole when ECT is not an option.

The addition of a non-clozapine antipsychotic to clozapine for individuals with residual symptoms despite clozapine treatment is controversial. A meta-analysis of randomized trials with a total of 612 patients found a benefit to clozapine augmentation with a first- or second- generation antipsychotic (14 trials, n = 612, standardized mean difference -0.52, 95% CI -0.90 to -0.14), although when only high-quality studies were included the difference was no longer significant [71]. This result was supported by an overview of a meta-analysis of combination treatments that showed none of the agents when combined with clozapine showed superior results to control in measurements of total symptom reduction [58].

Augmentation with aripiprazole has been studied more extensively than other antipsychotics because of the unique pharmacodynamic characteristics of the second-generation antipsychotics (see "Second-generation antipsychotic medications: Pharmacology, administration, and side effects"). Analysis of data from a national registry of health care data in Finland found that individuals who were treated with clozapine were 14 percent less likely to be hospitalized for psychosis during periods when they were also prescribed aripiprazole (hazard ratio 0.86, 95% CI 0.79-0.94) [72]. The beneficial effect of the combination was more pronounced in first-episode patients (hazard ratio 0.78, 95% CI 0.63-0.96).

●Other medications – The combination of other drugs with clozapine has been consistently ineffective in meta-analyses of randomized clinical trials, and for this reason their use is generally not recommended [58]. Weak evidence provides some support for augmentation with certain antidepressants (fluoxetine, duloxetine, citalopram) for persistent negative symptoms. Still weaker evidence supports clozapine augmentation with mood-stabilizers, anticonvulsants, pro-glutamatergic agents [73], rTMS, or transcranial direct current stimulation [68].

A network meta-analysis including 35 trials, 1472 patients with clozapine-resistant schizophrenia [average clozapine dose of 440.8±91.3 mg for 1168.2±710.3 days] examined 23 active augmentation treatments as compared with placebo. Three augmentation treatments appeared to be most efficacious for overall symptoms: mirtazapine (standardized mean difference -4.41, 95% CI -5.61 to -3.21), electroconvulsive therapy (ECT) (standardized mean difference -4.32, 95% CI -5.43 to -3.21), and memantine (standardized mean difference -2.02, 95% CI -3.14 to -0.91). None of the antipsychotics, NMDA receptor agonists, or antiseizure medications tested showed greater efficacy than placebo augmentation [64].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Psychotic disorders".)

SUMMARY AND RECOMMENDATIONS

●Treatment-resistant schizophrenia – While many patients show clinical improvement in psychotic symptoms in response to antipsychotic medication, a substantial proportion experience residual treatment-resistant symptoms. (See 'Introduction' above.)

●Establishing a diagnosis – We take a proactive approach to the evaluation and management of treatment-resistant schizophrenia. We undertake the following steps in all patients with suspected treatment resistance. (See 'Diagnosing treatment-resistant schizophrenia' above.)

•Reevaluate the primary diagnosis

•Assess for and address co-occurring mental or substance-use disorders, or medical conditions

•Rule out drug-drug interactions

•Rule out nonadherence

•Assess for medication related effects including adverse effects, medication interactions, nonadherence, and adequacy of trials

•Optimize non pharmacologic treatments.

●Diagnostic criteria - Varying definitions of treatment-resistant schizophrenia have posed challenges interpreting studies evaluating its treatment. We prefer the use of the optimal criteria for treatment-resistant schizophrenia, however, when this is not possible (eg, absence of standardized scales defining treatment responsiveness longitudinally, plasma levels not drawn) the minimal Treatment Response and Resistance in Psychosis (TRRIP) criteria are acceptable. (See 'Diagnostic criteria' above.)

●Management – For patients with treatment-resistant schizophrenia, we suggest treatment with clozapine rather than other antipsychotic medications (Grade 2C). (See 'Clozapine as first-line treatment' above.)

●Prescribing clozapine – We confirm the following prior to beginning clozapine treatment (see 'Eligibility' above):

•Absolute neutrophil count ≥1500 cells/microliter (or >1000 for individuals with diagnosed benign ethnic neutropenia)

•Determination by clinician as well as the patient and family if possible that the benefits of clozapine outweigh the risks

•Ability to adhere to treatment monitoring

•Patient/family agreement

We treat with clozapine for at least 24 weeks and confirm adequate exposure by assuring that plasma levels are above 350 ng/dL. (See 'Prescribing clozapine' above.)

●Patient ineligible or refusing clozapine – For patients with treatment-resistant schizophrenia who are not eligible for (or refuse) clozapine, we favor augmentation of a non-clozapine antipsychotic with electroconvulsive therapy (ECT) rather than other medications. (See 'Patients ineligible for or refusing clozapine' above.)

If ECT is not effective, we would use lamotrigine however, minocycline and topiramate are reasonable alternatives. (See 'Electroconvulsive therapy' above.)

Treatment of treatment-resistant schizophrenia with rTMS has shown inconsistent support. (See 'Neuromodulation' above.)

●Clozapine-resistant schizophrenia – For patients who have not responded to a 24-week trial of clozapine we favor ECT augmentation of clozapine rather than other medication options (Grade 2C). (See 'Augmentation with electroconvulsive therapy' above.)

If this is ineffective, we consider trials of augmentation of clozapine with a non-clozapine antipsychotic, or other agents such as mood stabilizers, anticonvulsants, or antidepressants. Minimal evidence supports these approaches. (See 'Treatment of clozapine-resistant schizophrenia' above.)