Treatment-resistant schizophrenia
Schizophrenia with inadequate response to at least two antipsychotic trials from different classes at adequate doses for at least 6 weeks each. Occurs in about 30% of patients with schizophrenia. NICE NG181 (2014, 2024 update) and the Spanish SEP 2024 set clozapine as drug of choice, titrated from 12.5–25 mg up to 300–450 mg daily. Clozapine works in 30–60% of patients previously unresponsive to other antipsychotics but requires regular CBC monitoring for agranulocytosis risk.
ICD-10 F20.92 drugs
Also known as: TRS, Refractory schizophrenia
Drugs used for this condition
- AClozapineFirst line · Atypical antipsychotic – dibenzodiazepineDrug of choice for treatment-resistant schizophrenia per NICE NG181 (2014, 2024 update) and the Spanish SEP 2024. Used after failure of at least two sequential antipsychotic trials from different classes at adequate doses. Starting dose 12.5 mg at bedtime with slow titration to 300–450 mg daily in 2–3 doses. Works in 30–60% of resistant cases – the highest rate among antipsychotics. Also indicated for suicidal behavior in schizophrenia.
- CAmisulprideThird line · Atypical antipsychotic of the benzamide classUsed for treatment-resistant schizophrenia after failure of two adequate trials of antipsychotics from different classes. Clozapine remains the gold standard for treatment-resistant schizophrenia. Evidence for amisulpride is limited and conflicting: it mostly concerns amisulpride augmentation of clozapine, and RCT results diverge (the large AMICUS trial showed no significant effect on the primary endpoint, while some smaller studies report partial improvement). Co-administration with clozapine raises amisulpride plasma levels and increases the risk of QT prolongation. The main risks are dose-dependent QT prolongation and hyperprolactinaemia.
How to read tiers A–F
Evidence tiers belong to the drug-indication pair, not to the drug as a whole: the same substance can be A for one condition and F for another.
- AStrong evidence – first or second line in international guidelines
- BRecommended in selected cases
- CStudy data contradict each other
- DBenefit not proven, no documented harm
- FDocumented harm, or risk without benefit