Fluoxetine
| Evidence Level: L4 | Predicted Indications: 10 |
Table of Contents
Fluoxetine: From Depressive and Anxiety Disorders to Schizoid Personality Disorder
One-Sentence Summary
Fluoxetine is a selective serotonin reuptake inhibitor (SSRI), originally used for depressive disorders, anxiety disorders, obsessive-compulsive disorder and bulimia nervosa. The TxGNN model's top-ranked prediction is schizoid personality disorder, but there are no clinical trials and only 2 publications (one narrative review and one cohort study, neither specific to fluoxetine in this disorder) behind it. Among the other predictions, phobic disorder has the strongest support (a completed Phase 3 trial, NCT00043537, in childhood social phobia), and this is discussed in the conclusion.
Quick Overview
| Item | Content |
|---|---|
| Original Indication | Not stated in the New Zealand licence records; the knowledge graph lists depressive disorders (major depressive disorder, dysthymia, unipolar and endogenous depression), anxiety disorder, obsessive-compulsive disorder and bulimia nervosa |
| Predicted New Indication | Schizoid personality disorder (top-ranked prediction) |
| TxGNN Prediction Score | 99.92% |
| Evidence Level | L4 |
| New Zealand Market Status | Cancelled only (no currently valid licences) |
| Number of Authorizations | 2 (both cancelled) |
| Recommended Decision | Hold |
Why is This Prediction Reasonable?
Currently, detailed mechanism of action data is not available in the data compiled for this report. Based on known information, fluoxetine is an SSRI that blocks the serotonin transporter (SERT). Its efficacy in depression and anxiety disorders is established. Pharmacology data also show binding to several serotonin receptors (5-HT2A, 5-HT2B, 5-HT2C, 5-HT6).
For schizoid personality disorder, the rationale is weak. Serotonergic modulation may influence affective flattening and social anxiety features, but there is no disorder-specific evidence. The only relevant source is a narrative review of drug treatment in cluster A personality disorders, which gives indirect support at best. The very high TxGNN score is likely driven by network similarity to other psychiatric conditions rather than by direct clinical findings.
Several other predictions have more credible links, since they sit close to fluoxetine's existing uses:
- Phobic disorder (rank 8): SSRIs dampen fear and anxiety responses, and a completed Phase 3 trial includes a fluoxetine arm.
- Agoraphobia (rank 6): fluoxetine has panic disorder studies, and SSRI class evidence supports use in panic disorder with agoraphobia.
- Schizotypal personality disorder (rank 3): a small 1991 open-label study showed reductions in self-injury and symptom scores with fluoxetine.
- Melancholia (rank 10): a severe subtype of major depression, which is already an approved use, so this is a refinement rather than true repurposing.
Clinical Trial Evidence
Currently no related clinical trials are registered for schizoid personality disorder.
For reference, trials linked to the better-supported predictions:
| Trial Number | Phase | Status | Enrollment | Key Findings |
|---|---|---|---|---|
| NCT00043537 | Phase 3 | Completed | 139 | Long-term comparison of behavioural treatment, fluoxetine and placebo for childhood social phobia (phobic disorder) |
| NCT02476136 | N/A | Unknown | 8800 | Individual patient data meta-analysis of antidepressant efficacy in anxiety disorders by initial severity (phobic disorder); not fluoxetine-specific |
| NCT01916824 | Phase 4 | Completed | 53 | Treatment effects on decision-making in major depression (melancholia); not melancholia-specific |
| NCT00961961 | Phase 4 | Completed | 177 | Antidepressant plus mood stabiliser versus mood stabiliser alone for preventing relapse of bipolar depression (manic bipolar disorder); only indirectly relevant |
Literature Evidence
Literature for the top prediction (schizoid personality disorder):
| PMID | Year | Type | Journal | Key Findings |
|---|---|---|---|---|
| 29955451 | 2016 | Review | The Mental Health Clinician | Narrative review of drug treatment in cluster A personality disorders (paranoid, schizoid, schizotypal); indirect support only |
| 16390895 | 2006 | Cohort | American Journal of Psychiatry | Course and predictors of outcome in depressed patients over 6 months of treatment; not specific to schizoid personality disorder |
Selected literature for the better-supported predictions:
| PMID | Year | Type | Journal | Key Findings |
|---|---|---|---|---|
| 1853957 | 1991 | Open-label | American Journal of Psychiatry | 12-week trial of fluoxetine in 22 patients with borderline or schizotypal personality disorder; reduced self-injury and symptom scores |
| 10362436 | 1999 | RCT | Journal of Clinical Psychiatry | Once-weekly fluoxetine dosing to maintain remission in panic disorder |
| 38014714 | 2023 | Network meta-analysis | Cochrane Database of Systematic Reviews | Drug treatments for panic disorder in adults; fluoxetine is not the best-supported agent |
| 11110016 | 2000 | Review | International Clinical Psychopharmacology | SSRIs including fluoxetine were superior to placebo in panic disorder and agoraphobia |
| 11001241 | 2000 | RCT | Journal of Clinical Psychopharmacology | Pindolol augmentation of fluoxetine in treatment-resistant panic disorder |
| 9812120 | 1998 | RCT | American Journal of Psychiatry | Fluoxetine versus placebo in panic disorder, assessing outcome measures |
| 15466674 | 2004 | Clinical trial | Archives of General Psychiatry | Fluoxetine, cognitive behavioural therapy and placebo in generalised social phobia |
| 16714326 | 2007 | Meta-analysis | Journal of Psychopharmacology | SSRIs in adult social anxiety disorder: meta-analysis of double-blind placebo-controlled trials |
| 9664779 | 1998 | Case report | Psychosomatics | Transient psychosis with psychogenic polydipsia in a schizotypal patient taking fluoxetine (safety signal) |
| 23218251 | 2013 | Meta-analysis | Journal of Affective Disorders | Olanzapine plus fluoxetine for the depressive phase of bipolar disorder |
New Zealand Market Information
| Authorization Number | Product Name | Dosage Form | Approved Indication |
|---|---|---|---|
| 2618575 | Fluox (cancelled) | Dispersible tablet 20 mg, scored | Not recorded |
| 2450860 | Arrow-Fluoxetine (cancelled) | Capsule 20 mg | Not recorded |
Both licences are cancelled, so there is currently no valid fluoxetine licence in the records compiled for this report.
Safety Considerations
- Contraindications flagged by the knowledge graph (model-derived associations, not package insert text): bipolar disorder and major affective disorder, epilepsy, liver disease and liver failure, cirrhosis, diabetes and diabetic ketoacidosis, hyperglycaemia, neutropenia and agranulocytosis, blood coagulation disorders, benign prostatic hyperplasia, low blood pressure disorders (orthostatic hypotension, postural orthostatic tachycardia syndrome, multiple system atrophy), stroke and cerebrovascular disorders, myocardial infarction, coronary thrombosis, hyperlipidaemia, hypertriglyceridaemia, obesity, gallbladder disease and female breast carcinoma.
- Drug interactions: the interaction search returned no drug-drug interactions. The nine records returned are pharmacological target bindings: SERT, 5-HT2A, 5-HT2B, 5-HT2C, 5-HT6, Kir3.2, Kv3.1, the calcium-activated chloride channel (CaCC/ANO1) and the plasma membrane monoamine transporter.
- Prediction-specific concerns:
- Antidepressant use in manic bipolar disorder carries a risk of manic switching.
- Children and adolescents need monitoring for suicidality (boxed warning), relevant to the social phobia trial population.
- A case report links fluoxetine to transient psychosis with polydipsia in a schizotypal patient.
- Package insert warnings and contraindications were not available. Please refer to the package insert for the official safety information.
Conclusion and Next Steps
Decision: Hold
Rationale: The top-ranked prediction, schizoid personality disorder, has no clinical trials and only indirect literature, so it is supported by the model score alone (L4). Fluoxetine has no valid New Zealand licence at present, and package insert safety data were not available. Within the broader prediction list, phobic disorder (L1, with a completed Phase 3 trial) would be the most reasonable candidate to advance under guardrails, and agoraphobia and melancholia sit at L2 as research questions.
To proceed, the following is needed:
- Package insert warnings and contraindications from Medsafe (the current safety screening gap)
- Mechanism of action data
- For phobic disorder: fluoxetine arm details and outcome results of NCT00043537, plus a suicidality monitoring plan for children and adolescents
- For schizoid personality disorder: controlled studies of fluoxetine in cluster A personality disorders
- Clarification of the regulatory pathway given that both New Zealand licences are cancelled
Disclaimer
This content is for research purposes only and does not constitute medical advice. Clinical validation is required before any clinical application.