Azelastine

Limits on what this shows

• Null findings: ROS, ATP and oxygen consumption were not reported in the colorectal paper [9], and mitochondria were not measured in the glioblastoma paper, which cites the colorectal fission result without testing it [10]. No human tissue or blood measurement was found.

Regulatory status

No mitochondrial statement in the labels retrieved [1][2][3].

Population evidence

None in humans. Tier B, azelastine-specific, in vitro or animal. HeLa cells, 48 hours, 15 to 90 uM [8]: cells with depolarized mitochondrial membranes 9.76% (15 uM), 14.05% (25 uM), 15.89% (45 uM), above 50% (60 and 90 uM); viability by MTT 86%, 68.33%, 51.33%, about 8% and about 4% across the same doses. Colorectal cancer lines HT29 and DLD1 [9]: weaker mitochondrial fission on electron microscopy and immunofluorescence, lower phospho-Drp1, and a dose-dependent fall in the JC-1 red/green ratio at 48 hours (concentrations given only as "indicated"); proteomics ranked mitochondrial dysfunction first (P = 8.77 x 10^-7). HT29 xenograft mice given 10 or 20 mg/kg orally every second day (n=7) showed tumor inhibition of 40.7% and 72.1%, no body weight change, no obvious ALT or AST change and no liver, kidney or lung toxicity on histology [9]. Glioblastoma lines [10]: mitochondria were not measured.

Mechanism

Named in the colorectal work: azelastine binds Thr-48 of ARF1 (Kd 1.74 x 10^-9 M for wild-type versus 5.94 x 10^-7 M for the T48S mutant), reducing Drp1 phosphorylation and mitochondrial fission [9]. Normal cells were not tested [9]. The label mechanism (H1 antagonism, mast cell stabilization [7]) is unrelated to mitochondria.

Exposure gap

Cmax of 200 pg/mL (0.1%) is about 0.52 nM using a molar mass of 381.90 g/mol [2][7] (arithmetic by this review); 409 pg/mL (0.15%) is about 1.07 nM; about 88% is protein bound [2]. The lowest HeLa dose that depolarized mitochondria, 15 uM, is roughly 29,000 times the 0.1% peak. Concentration in the nasal lining after a spray was not measured in any retrieved source.

Counter-evidence held on record

Counter-evidence The cytotoxic and antiproliferative context (cancer lines, 48-hour exposure) is not representative of topical allergy dosing. The mouse study found no organ toxicity at 10 to 20 mg/kg orally [9], although plasma exposure was not reported. Framing caution This is not a mitochondrial toxicity claim for nasal or eye-drop users and not a reason to take any mitochondrial supplement.

Limits on what this shows

• Null findings: No effect on cell-free ROS [11]. The glioblastoma paper used an oxidative stress inhibitor but did not measure ROS [10]. No human glutathione or malondialdehyde measurement in azelastine users was found; the only human in vivo redox statement [20] is an abstract-level neutrophil respiratory burst result in a confounded chemoradiotherapy population, and neutrophil SOD was negligibly changed [19][20].

Regulatory status

No oxidative stress statement in the labels retrieved [1][2][3].

Population evidence

No systemic redox marker in humans. Tier B, azelastine-specific. Human neutrophils [11]: azelastine significantly reduced superoxide, hydrogen peroxide and hydroxyl radical production and phagocytosis, with no marked effect on chemotaxis and none on a cell-free xanthine/xanthine oxidase system; concentrations were not stated on the page read (abstract only). Human neutrophils [12]: superoxide generation fell in cells activated by PMA or fMLP, and leukotriene B4 formation was inhibited only at 20 to 200 uM across the five drugs tested; per-drug values not given in the abstract. A trial in 43 patients with recurrent aphthous ulcers [13] reports that azelastine suppressed neutrophil superoxide and chemiluminescence dose-dependently in laboratory tests (aggregator summary only; authors and year not captured). HeLa cells [8]: ROS-positive cells 3.29% in controls versus 12.4% (15 uM), 24.8% (25 uM), above 45% (45 uM), 48.93% (60 uM) and 49.99% (90 uM). Ueta 1995 [19] (human neutrophils and mononuclear cells, rabbit alveolar macrophages, fibroblasts, plus rabbits in vivo): azelastine ("Azeptin") dose-dependently suppressed reactive oxygen generation in vitro, and 10^-5 M azelastine made the peplomycin-induced increase negligible; in rabbits, azelastine at 0.04 mg/kg/day given with peplomycin at 0.1 mg/kg/day produced no rise in reactive oxygen generation; azelastine did not change superoxide dismutase activity (peplomycin suppressed it). Human in vivo, confounded: Osaki 1994 [20], 63 patients receiving chemoradiotherapy with peplomycin and 5-fluorouracil, of whom 37 received azelastine 2 mg daily plus vitamin C 500 mg, vitamin E 200 and glutathione 200 mg and 26 received the same vitamins and glutathione without azelastine; the abstract states azelastine suppressed the neutrophil respiratory burst in vivo and in vitro, with neutrophil SOD activity negligibly suppressed; the mucositis outcome (grade 3 or 4 at completion 16 of 37 versus 21 of 26 as read from the abstract) is a clinical outcome, not a redox marker, and the abstract numbers were reconstructed from an index and flagged as needing verification. Rat cisplatin model: Abo El-Ela 2025 [21], azelastine 4 or 6 mg/kg orally for 10 days; at 6 mg/kg malondialdehyde fell (values reported jointly with tissue iron, 33.6% to 70.9%, p < 0.0001), and glutathione (2.7-fold renal, 2.1-fold testicular), Nrf2 (6.9 and 15.6-fold) and GPX4 (3.9 and 4.8-fold) rose against the cisplatin-injured state.

Mechanism

Present in two opposing forms. Reduced ROS output from activated neutrophils is a cellular effect on production, not direct scavenging, since the cell-free system was unaffected [11]. Increased ROS in HeLa cells accompanies mitochondrial depolarization at 15 uM and above [8]. Neither form was linked to depletion of an antioxidant nutrient. The rat work names an Nrf2/SLC7A11-GPX4 pathway [21].

Counter-evidence held on record

Counter-evidence The neutrophil results use abstracts only, without stated concentrations, and the HeLa concentrations exceed label plasma exposure by about four orders of magnitude (see the mitochondrial entry). Ueta 1995 [19] and Osaki 1994 [20] come from the same Kochi research group and are one line of work, and [20] gave antioxidant vitamins and glutathione to both arms, so it cannot isolate an azelastine effect on systemic antioxidant status. The rat work [21] used an injury model with doses far above human nasal exposure. The direction is protective or neutral, which keeps this entry yellow. Framing caution Not an antioxidant benefit claim, not a pro-oxidant claim, and not antioxidant-nutrient depletion. Do not suggest antioxidant supplements.

Limits on what this shows

• Null findings: No ferritin, serum iron, transferrin saturation, hemoglobin or iron-absorption measurement in any azelastine user was found. PubMed listing "azelastine AND (iron OR anemia OR hemolysis)" returned 4 records; the other three concerned red-cell membrane leakage [22], a nasal nanosuspension and a rat paw edema photoproduct, none an iron status finding.

Regulatory status

No iron interaction in the label sections retrieved [1][2][3][4][5].

Population evidence

None in humans. Tier B, azelastine-specific, animal [21] (abstract only, read as a PubMed record through a summarizing tool; Crossref and OpenAlex records carry no abstract): male Sprague-Dawley rats received azelastine 4 or 6 mg/kg orally for 10 days and a single intraperitoneal cisplatin dose of 10 mg/kg on day 5; group sizes are not stated in the abstract. At 6 mg/kg, malondialdehyde and tissue iron decreased by 70.9%, 51%, 44.5% and 33.6% across renal and testicular tissue (p < 0.0001; the abstract does not label which value belongs to which marker and tissue). Transferrin receptor decreased by 27.6% (p < 0.05) in one tissue and 39% (p < 0.0001) in the other. Cisplatin-elevated serum creatinine, BUN and urinary protein were reduced by 38.5%, 32.3% and 60%. NCOA4 and FTH1 mRNA decreased, NCOA4 protein increased by 44.5% and FTH1 protein decreased; the abstract does not reconcile the discordant NCOA4 mRNA and protein directions. The comparator is presumably the cisplatin-only group, which the abstract does not state explicitly; an azelastine-alone group was not reported.

Mechanism

Named by the authors: Nrf2/SLC7A11-GPX4 signaling and NCOA4-mediated ferritinophagy, that is, inhibition of ferroptosis in cisplatin-injured tissue [21]. This is tissue iron redistribution in an injury model. No classical iron mechanism (chelation, absorption interference, blood loss, catabolism) is present for azelastine.

Exposure gap

6 mg/kg orally per day in rats compared with a nasal defined daily dose of 0.56 mg [6]: for a 70 kg adult that is about 0.008 mg/kg, so the rat dose is roughly 750 times higher per kg (arithmetic by this review, ignoring species scaling and the 40% nasal bioavailability).

Counter-evidence held on record

Counter-evidence and confounding The measured iron fall is relative to a cisplatin-injured state and reflects the drug's cytoprotective pharmacology, not depletion (Hard Rule 2). Only the abstract was read, the tissue and marker assignments are unlabeled, and the NCOA4 mRNA and protein results disagree. Framing caution Direction is protective or neutral. This is not iron deficiency and gives no basis for iron supplementation or iron restriction.

Limits on what this shows

• Why no grade: Dual condition met: no Tier B or C evidence, and no circulating copper claim for azelastine (the generic depletion articles opened do not mention it [16][17]). Not a finding of safety. Graded separately from Iron per Hard Rule 7.

What was searched

Web search "azelastine zinc OR iron OR copper chelation OR serum levels study": results were iron-chelator therapy papers unrelated to azelastine. PubMed listing "azelastine AND (zinc OR copper OR ferritin OR hemoglobin OR methemoglobin)": 4 results (a lipid and aortic calcification paper, a membrane-interaction paper on antihistamines, a ZINC-database virtual screen, an in silico NTCP study); none concerned copper status. Titles only, not opened. Label sections retrieved list no mineral interaction [1][2][3][4][5]. Update 2026-10-08: a re-run PubMed listing "azelastine AND (copper OR zinc OR selenium OR ceruloplasmin OR trace element OR chelation)" returned 4 records whose abstracts were read (ocular chitosan-alginate microspheres, a chemical stability study, a ZINC-database virtual screen, an in silico NTCP study); none measured a mineral or described metal binding. The ferroptosis rat paper [21] was published in a trace element journal but reports iron only.

Channel outcomes

PubMed opened in both reviews; other databases 429 or disallowed. A clean zero on opened channels, now on two different query phrasings with abstracts read in the second. The ungraded outcome is a genuine gap, with the residual caveat that Europe PMC, Semantic Scholar and full texts were not searched.

Limits on what this shows

• Null findings: PubMed listing "azelastine AND (microbiome OR microbiota)": 0 results. PubMed listing "azelastine AND (antibacterial OR antimicrobial OR bacteria OR bactericidal OR antibiofilm)": 74 results, first 10 read; two reported antibacterial activity [25][26], none involved gut commensals. The 2018 gut-commensal screen of more than 1,000 marketed drugs against 40 gut strains [28] did not mention azelastine on the page read, but its supplementary table could not be opened, so whether azelastine was screened is unknown and is not recorded as a null.

Regulatory status

No antibacterial or microbiome statement in the label sections retrieved [1][2][3][4][5].

Population evidence

None in humans. Tier B, azelastine-specific, in vitro, abstracts only. El-Nakeeb 2011 [24]: 10 antihistamines against 29 multidrug-resistant Gram-positive and Gram-negative clinical isolates; azelastine was among the most active (with cyproheptadine, mequitazine and promethazine), while doxylamine, fexofenadine and loratadine were inactive at 1 mg/mL; minimum bactericidal concentrations were almost identical to minimum inhibitory concentrations; in saline viable counts azelastine killed in a concentration-dependent way that fell with larger inoculum and rose with higher pH, and the post-antimicrobial effect of 100 ug/mL azelastine reached up to 3.36 hours. Organism-level MIC values were not retrieved (full text returned 403). Boyd 2021 [25]: four non-antibiotic drugs against five clinical S. aureus isolates (nasal and wound swabs) and one control strain; azelastine MIC 200 ug/mL for all isolates, MBC within one dilution. Kuratli 2020 [26]: Chlamydia trachomatis in HeLa cells; azelastine 10 and 20 uM reduced inclusion size, number and infectivity, 20 uM disrupted inclusion structure, infectivity recovered only incompletely after removal, no cytotoxicity up to 20 uM.

Mechanism

Mechanism class G is present as a plausibility only. The retrieved abstracts name no antibacterial mechanism. Azelastine's membrane surface activity [22] is a possible but untested link and is not claimed here.

Exposure gap

An MIC of 200 ug/mL [25] is 1,000,000 times the label plasma peak of 200 pg/mL (0.0002 ug/mL; arithmetic by this review) [2]. About 75% of an oral radiolabeled dose appears in feces [4], but the swallowed fraction of a nasal spray and the resulting gut luminal concentration were not found in any source.

Counter-evidence held on record

Counter-evidence and confounding The tested organisms are pathogens and clinical isolates, not commensals, and the concentrations are supratherapeutic. Allergic rhinitis and its treatments independently relate to the nasal microbiome in the literature seen only by title. Framing caution Not a dysbiosis claim and not a reason to take a probiotic. No effect on the gut microbiome of azelastine users has been shown or excluded.

Limits on what this shows

• Null findings: PubMed listing "azelastine AND (iron OR anemia OR hemolysis)": 4 records, none a clinical hemolysis, anemia or methemoglobinemia report; one 2023 nanosuspension paper used a hemolysis blood-compatibility assay and its abstract reports no result (title and snippet only). The human aerobic test [23] found no impairment of aerobic performance.

Regulatory status

No hematologic warning in the label sections retrieved [1][2][3][4][5]; only selected sections were read through a summarizing tool.

Population evidence

Tier B, azelastine-specific, in vitro (abstract only, via an OpenAlex record) [22]: nine antihistamines were compared (desloratadine, clemastine, azelastine, ketotifen, diphenhydramine, pyrilamine, emedastine, epinastine, olopatadine). Azelastine ranked third for surface activity at a phospholipid monolayer. Across the agents other than olopatadine and emedastine, which did not perturb red cells, there was concentration-dependent hemoglobin leakage from intact erythrocytes; the half-maximal leakage concentration correlated with monolayer dissociation constants (r = 0.98). Azelastine's own half-maximal leakage concentration was not given in the abstract. At the marketed ocular concentration of azelastine (0.05%), significant membrane perturbation and lactate dehydrogenase release from human conjunctival and corneal epithelial cells were reported. Tier A, human, functional [23] (abstract only; the PubMed record returned 429): 15 physically active men with allergic rhinitis or rhinoconjunctivitis took intranasal azelastine 0.56 mg/day for 5 days and were compared with 15 untreated healthy controls on a maximal incremental cycle test before and after. Recorded variables were power, heart rate, VO2, ventilation, oxygen pulse and blood lactate. In the azelastine group peak heart rate was lower after treatment (190 versus 186 beats/min, p < 0.05) and ventilation at the ventilatory threshold was higher (54.7 versus 60.2 L/min, p < 0.05); the controls did not change. The authors concluded maximal and submaximal aerobic performance were not impaired. The abstract record gave no numeric VO2max or lactate result, and one printed standard deviation looked like a transcription error and is not reproduced. Hemoglobin, hematocrit and methemoglobin were not measured.

Mechanism

Direct membrane perturbation through surface activity, described by the authors as partly receptor-mediated and partly direct [22]. There is no demonstrated route from this in vitro effect to red-cell loss in users, and no named oxygen-carrying mechanism. Not merged with ATP production (Hard Rule 7).

Exposure gap

The 0.05% ocular product is 0.5 mg/mL (arithmetic by this review), which is about 5 x 10^8 ng/mL; plasma azelastine after 56 days of eye drops was 0.02 to 0.25 ng/mL [3], so circulating red cells see a concentration roughly nine orders of magnitude below the product concentration. The leakage concentration for azelastine was not retrieved, so no direct comparison with plasma levels is possible.

Counter-evidence held on record

Counter-evidence and confounding Heart rate and ventilation changes in [23] are physiological or pharmacologic outcomes (Hard Rule 2) and are not read as an oxygen-delivery deficiency in either direction. The study was 5 days, in men with active allergic disease, with untreated controls. The membrane study used purified or isolated systems at concentrations not shown to reach blood. Framing caution Do not present azelastine as causing hemolysis, anemia or reduced oxygen-carrying capacity. No iron or other supplement follows from this entry.

Limits on what this shows

• Why no grade: Dual condition met: no Tier B or C source measured ATP for azelastine, and no circulating claim of azelastine-related low energy was found. Mitochondrial depolarization (graded separately) is not an ATP measurement, and inferring ATP from it would be unfalsifiable stacking. The whole-body aerobic test [23] is recorded under Oxygen delivery/utilization and is not ATP evidence. A disclosed gap, not a finding of safety. Kept independent of Mitochondrial burden and Oxygen delivery/utilization (Hard Rule 7).

What was searched

Web search "azelastine cellular ATP levels oxidative phosphorylation oxygen consumption respiration cells": only the HeLa paper, in which ATP was not measured and the authors only speculate that higher mitochondrial numbers may reflect ATP demand [8]. PubMed listing "azelastine AND (ATP OR oxygen consumption OR membrane potential OR respiration)": 38 results, page 1 of 4 read, with no ATP measurement among titles. The colorectal paper reports no ATP or oxygen consumption data [9].

Channel outcomes

Update 2026-10-08: the same PubMed listing was re-read with a larger page size and returned 20 titled records (none an ATP, mitochondrial respiration or oxygen consumption measurement with azelastine) plus 18 PMIDs shown without titles. Three of those were opened through OpenAlex: Ueta 1994 on neutrophil respiratory burst (no abstract in the record), a 1989 mast cell granule phosphatidylinositol kinase study in which ATP is only the radiolabeled assay substrate, and a 2026 docking paper; none measured cellular ATP. The remaining 15 untitled PMIDs and the 1990 neutrophil metabolic activation paper (no abstract in the record) were not opened. The Maier 2018 and Chicharro 1998 papers are not ATP measurements. Result: a clean zero on the channels that opened, plus a partial tooling limit.