Loxapine

Limits on what this shows

• Null/negative findings: In the same experiment, levomepromazine (also a phenothiazine), ziprasidone, and olanzapine showed no significant complex I inhibition - a real null result demonstrating the effect is not universal even among structurally related or typical agents.
• At-risk subgroups: None established for loxapine. Patients with pre-existing mitochondrial disease are a theoretical class-wide concern, unconfirmed for this drug.
• Counter-evidence: Levomepromazine's null result argues against assuming uniform phenothiazine-class behavior. The active concentration (50 µmol/L) is supratherapeutic for an in vitro system and has not been validated against clinical plasma levels for any of the drugs tested.

Regulatory status

Not addressed in either loxapine product label.

Population evidence

In vitro study, Naunyn-Schmiedeberg's Arch Pharmacol (2019), isolated pig brain mitochondria. Ten antipsychotics tested: typical - chlorpromazine, levomepromazine, haloperidol; atypical - risperidone, ziprasidone, zotepine, aripiprazole, clozapine, olanzapine, quetiapine. At 50 µmol/L, chlorpromazine produced complete complex I blockade; haloperidol produced significant inhibition (IC50 64.9 µmol/L, a full inhibitor). Loxapine was not tested.

Mechanism

Direct inhibition of NADH:ubiquinone oxidoreductase (complex I) at the inner mitochondrial membrane, demonstrated for two structurally different typical antipsychotics (a phenothiazine and a butyrophenone) - suggesting the effect is not restricted to phenothiazine chemistry, though this remains unconfirmed for loxapine's dibenzoxazepine structure.

Counter-evidence held on record

Class extrapolation only (Hard Rule 6); loxapine was absent from the tested cohort. Haloperidol's inclusion (a non-phenothiazine butyrophenone) supports treating this as a defensible typical/first-generation-antipsychotic class signal rather than a phenothiazine-specific one, but it remains unconfirmed for loxapine. Shares its underlying source with the ATP production entry below - see Overlap map.

Limits on what this shows

• Null/negative findings: Both studies found oxidative-stress markers abnormal even in some SGA-treated patients and, in one analysis, "regardless of treatment type" - evidence that illness itself, not only the drug, contributes.
• At-risk subgroups: None loxapine-specific established.
• Counter-evidence: Both studies are open-label/non-blinded with small first-generation arms (n=5-9 for haloperidol alone); confounding by indication (illness severity, chronicity) is not excluded in either design (Hard Rule 6).

Regulatory status

Not addressed in either loxapine product label.

Population evidence

(1) Open comparative study, J Neuropsychiatry Clin Neurosci, 92 patients over 21 days: first-generation group = haloperidol (n=5, mean 6.2 mg/day) and flupentixol (n=17, mean 3.5 mg/day); second-generation group n=70 (clozapine, amisulpride, quetiapine, risperidone, olanzapine). Day-21 plasma malondialdehyde (MDA) was highest in the first-generation group (mean 0.81 µmol/L); among SGAs, clozapine was lowest (0.21) and olanzapine matched the FGA group (0.80). (2) Separate comparative cohort (haloperidol n=9 vs olanzapine/quetiapine/risperidone n=26): significantly decreased glutathione peroxidase activity and increased MDA in schizophrenia patients overall "regardless of type of treatment," with superoxide dismutase activity raised particularly in the haloperidol and quetiapine subgroups. Loxapine was absent from both cohorts.

Mechanism

Named only generically (oxidative/antioxidant-enzyme imbalance) in both sources; no drug-structure-specific biochemical mechanism is established.

Counter-evidence held on record

Loxapine was absent from both cohorts. Haloperidol (butyrophenone) and flupentixol (thioxanthene) are both non-phenothiazine typical agents showing the signal, which supports treating this as a defensible typical/first-generation-antipsychotic class claim rather than a phenothiazine-specific one - but it remains unconfirmed for loxapine, and confounding by indication has not been excluded in the source studies.

Limits on what this shows

• Null/negative findings: No subgroup in this study showed low copper; the direction reported is uniformly toward excess.
• At-risk subgroups: None established for loxapine specifically.
• Counter-evidence: Confounding by indication is not excluded: elevated ceruloplasmin/copper is separately documented in unmedicated schizophrenia, so the drug-attributable fraction of this signal, if any, is unclear (Hard Rule 6).

Regulatory status

Not addressed in either loxapine product label.

Population evidence

Psychiatry Research cohort, 62 schizophrenia outpatients (17 on depot neuroleptics, remainder on atypical antipsychotics) versus healthy controls. Serum copper was significantly higher in the schizophrenia group overall (mean 117.4 µg/dL) than in controls (105.6 µg/dL). Specific depot drug names were not given in the accessible text; loxapine has no depot/long-acting-injectable formulation and is not expected to be represented in this cohort.

Mechanism

Not established. The authors speculate that "altered levels of trace elements... may be a consequence of antipsychotic treatment," but ceruloplasmin (the dominant copper-carrying plasma protein) is independently documented as an acute-phase reactant elevated in psychosis itself, apart from any medication.

Counter-evidence held on record

This is an excess/retention-direction finding, not a depletion (Hard Rules 4 and 5) - it must never be read as a reason to add copper. No class member confirmed to include loxapine was studied, and no copper-deficiency signal of any kind was found anywhere in the loxapine or typical-antipsychotic literature reviewed. Supplementation would be the hazard here, not the remedy, if the signal is drug-related at all.

Limits on what this shows

• Null/negative findings: Serum iron normalized after recovery from NMS in the case series - evidence against a chronic deficiency and for a reversible, state-dependent marker.
• At-risk subgroups: Patients who develop NMS on loxapine - a labeled risk shared across neuroleptics generally, not shown to be higher or lower specifically for loxapine versus other typical agents in the evidence reviewed.
• Counter-evidence: No loxapine-specific serum iron measurement exists anywhere located in this review; the link between "loxapine causes NMS" and "NMS lowers serum iron" rests on two separate evidence lines that have not been directly joined in a single published loxapine case.

Regulatory status

NMS itself is a labeled, class-wide adverse reaction for loxapine; serum iron is not mentioned in either loxapine label.

Population evidence

Prospective case series, 26 NMS episodes in 23 patients: serum iron was ≤10 µmol/L in 25 of 26 episodes (96%), and ≤4 µmol/L in 11 episodes (36%); levels normalized after NMS resolved. The specific neuroleptic(s) used were not individually named in the accessible abstract. Separately, two loxapine-specific case reports confirm loxapine itself can precipitate NMS: an 81-year-old patient on loxapine (10 mg each morning, 50 mg each evening, increased to 65 mg at bedtime) who developed hyperthermia, rigidity, and an elevated creatine kinase; and a second case reported in the British Journal of Psychiatry (full text could not be opened in this review - see Sources/access notes). Neither loxapine case report measured serum iron.

Mechanism

Proposed as part of an acute-phase response accompanying the hypermetabolic, hyperthermic NMS state - iron is thought to redistribute into storage/inflammatory compartments, not to be lost from the body, and the case-series data show it renormalizes once NMS resolves.

Counter-evidence held on record

This is an acute, transient, drug-toxicity-associated laboratory change, not a dietary iron deficiency (Hard Rules 2 and 3) - it is not evidence that routine loxapine use depletes iron, carries no supplementation implication, and the marker is only clinically relevant when NMS is already suspected, where it resolves with NMS treatment, not with iron repletion.

Limits on what this shows

• Why no grade: Dual condition met. (1) No Tier B/C evidence: a 2024 systematic review of antipsychotics and gastrointestinal microbiota (Prog Neuropsychopharmacol Biol Psychiatry) covered only risperidone (5 studies), quetiapine (3), amisulpride (1), olanzapine (1), and unspecified atypicals (2) - zero first-generation/typical antipsychotics and zero loxapine - and its own authors describe even the atypical findings as "inconclusive." The one typical-antipsychotic-adjacent fact located, that chlorpromazine (a phenothiazine) has documented direct in vitro antibacterial/antifungal activity, is a real pharmacological property but has never been tested as an in vivo gut-dysbiosis mechanism and is not transferable to loxapine's distinct dibenzoxazepine ring structure without further justification (Hard Rule 6). (2) No circulating consumer claim naming loxapine or typical/first-generation antipsychotics for gut microbiome disruption was located during this review.

Channels tried

PubMed search-listing pages: returned a bot-verification/reCAPTCHA interstitial on every attempt this session, not a clean zero-result. Europe PMC REST API: HTTP 429 (rate-limited) on every attempt. Crossref REST API and OpenAlex search API: HTTP 403 policy-denial when reached directly through the sandboxed shell, HTTP 429 when reached through the fetch tool. DailyMed/FDA labels for both loxapine formulations: read cleanly, silent on microbiome/GI flora. Semantic Scholar: not reached (session web-search budget was exhausted before this channel could be tried). Direct publisher-page fetches succeeded for the one directly relevant systematic review (ScienceDirect), which is the substantive negative result reported above.

Disposition

Mixed: part genuine literature gap (the dedicated systematic review's own exclusion of loxapine/FGAs), part tooling limitation at the time of writing (PubMed/Europe PMC/Crossref/OpenAlex/Semantic Scholar blocked or rate-limited). A later revision should re-run those searches once access is restored.

Limits on what this shows

• Null/negative findings: Olanzapine was the one tested drug in this study with no significant respiratory inhibition.
• At-risk subgroups: None established for loxapine.
• Counter-evidence: Same caveats as Mitochondrial burden: supratherapeutic in vitro concentrations, no clinical ATP or lactate correlate measured in patients, loxapine absent from the tested cohort.

Regulatory status

Not addressed in either loxapine product label.

Population evidence

Same source as Mitochondrial burden (Naunyn-Schmiedeberg's Arch Pharmacol, 2019, isolated pig brain mitochondria, 10 antipsychotics tested). ADP-stimulated mitochondrial respiratory rate was significantly inhibited by all tested antipsychotics except olanzapine; aripiprazole (IC50 13.1 µmol/L) and zotepine (IC50 39.5 µmol/L) were full inhibitors of complex-I-linked respiration, haloperidol IC50 64.9 µmol/L. Loxapine was not tested.

Mechanism

Reduced complex I electron flow lowers the proton-motive force available to ATP synthase, lowering maximal ATP-generating (State 3) respiration - shown directly for haloperidol and chlorpromazine among the typical agents tested.

Counter-evidence held on record

This entry shares its entire underlying source with Mitochondrial burden above - see Overlap map. Do not read the two entries as two independent confirmations; they are one in vitro finding reported as two physiological endpoints. Class extrapolation only; unconfirmed for loxapine.

Limits on what this shows

• Why no grade: No Tier B/C evidence connects loxapine, or a defensible class member, to a nutrient- or mineral-mediated change in hemoglobin, hematocrit, red-cell mass, or tissue oxygen delivery. Two real hematologic/respiratory signals exist in the label but neither qualifies as a nutrient entry: (1) rare agranulocytosis, leukopenia, and thrombocytopenia are white-cell/platelet effects, not red-cell/oxygen-carrying effects, and are the drug's own labeled idiosyncratic toxicity, not a nutrient deficiency (Hard Rule 2). (2) Inhaled loxapine (Adasuve) carries a boxed warning for bronchospasm (FEV1 dropped up to 537 mL after a second dose in asthmatic study volunteers versus 0% bronchospasm with placebo), and oral/parenteral loxapine can cause orthostatic hypotension via alpha-adrenergic blockade; both are the drug's own direct local-irritant or pharmacologic effects on respiratory and vascular tone, excluded from grading under Hard Rule 2 exactly as a beta-blocker's intended heart-rate reduction would be. No circulating consumer claim linking loxapine to a nutrient-based oxygen-delivery or anemia mechanism was located.

Channels tried

Same blocked/rate-limited pattern as Gut microbiome risk for PubMed, Europe PMC, Crossref, OpenAlex, and Semantic Scholar. DailyMed and the FDA Adasuve label were both read cleanly and are the source of the agranulocytosis/bronchospasm/orthostatic findings above - a genuine clean read, not a tooling failure, for the primary-source channel.

Disposition

Primarily a genuine absence of any oxygen-delivery/hematologic-nutrient signal in the readable primary sources, though the blocked literature-search channels mean a drug-specific hemoglobin/hematocrit study cannot be ruled out with full confidence; re-check on revision.