Compounded Amlexanox for MCAS: What the Evidence Actually Supports

WWW.DIRECTINTEGRATIVECARE.COM   ·   MAST CELL ACTIVATION SYNDROME

Compounded Amlexanox for MCAS: What the Evidence Actually Supports

An old Japanese anti-allergy drug is being repositioned as a mast cell therapy. The mechanism is interesting. The human evidence is not what the marketing suggests — and the dose being prescribed does not match any study ever published.

Yoon Hang Kim, MD, MPH

Board-Certified in Preventive Medicine | Integrative & Functional Medicine Physician

Over the past two years, compounded oral amlexanox has moved from obscurity into the functional medicine conversation as a possible treatment for mast cell activation syndrome (MCAS). Compounding pharmacy newsletters, practitioner podcasts, and patient forums have all picked it up. Clients now bring it to me by name.

The interest is understandable. MCAS is a difficult condition with a short list of tolerable options, and clients who have cycled through antihistamines, cromolyn, ketotifen, and leukotriene-directed therapy without adequate relief are looking for something else. When a drug with genuine mast cell pharmacology appears, it draws attention.

But there is a wide gap between what amlexanox has been shown to do in humans and what is being claimed for it in MCAS. That gap deserves a clear description — not because the drug is unreasonable to consider, but because the reasons to consider it are not the ones circulating.

What amlexanox actually is

Amlexanox is a small-molecule anti-allergic and anti-inflammatory compound first brought to market in Japan in 1987 for allergic bronchial asthma, and later extended there to allergic rhinitis and conjunctivitis.¹ In the United States it had a much narrower life: it was approved as a 5% topical oral paste for aphthous ulcers, and that product was discontinued around 2017 as other options became available.¹ The oral systemic formulation used in Japan was never approved in the United States.

So when a compounding pharmacy dispenses an oral amlexanox capsule today, it is producing a dosage form that has no United States approval history at all. That is not automatically disqualifying — compounded therapies routinely fall outside approved formulations — but it is the correct starting point for the conversation, and it is frequently left out.

Two mechanisms, one marketing story

Most promotional material for compounded amlexanox describes it as a mast cell stabilizer that reduces histamine and leukotriene release. That description is not wrong. It is the original 1980s Japanese pharmacology, and it is the basis on which the drug was used for asthma and rhinitis for decades.

Where the story gets muddied is with the second mechanism. Beginning with work published in 2013, amlexanox was rediscovered as an inhibitor of two protein kinases, TBK1 and IKKε, and shown in mouse models to improve obesity-related metabolic dysfunction.³ That finding generated real scientific momentum — further work established TBK1 as a node connecting inflammation and energy homeostasis in adipose tissue⁴ — and it is the reason amlexanox appears in recent review literature as a compound of interest for metabolic and inflammatory disease.¹

These are two different arguments. The kinase work is a metabolic and adipose-tissue story. It does not extend the mast cell claim, and it does not supply evidence for MCAS. When a newsletter cites the excitement generated by the TBK1 research and then pivots to mast cell stabilization in the next paragraph, it is borrowing credibility from a literature that was not studying mast cells.

What the human evidence shows

The most rigorous human data on systemic amlexanox comes from the University of Michigan diabetes program. In a randomized, double-blind, placebo-controlled study of 42 participants with obesity, type 2 diabetes, and non-alcoholic fatty liver disease, roughly one-third of treated participants achieved a hemoglobin A1c reduction of 0.5% or greater.^1,2 The finding that has held clinicians’ attention since is the responder profile: participants who responded had higher baseline inflammatory markers in serum and in gene expression profiles than those who did not.¹

That is a genuinely interesting result. It suggests amlexanox may act meaningfully only in people whose disease has a substantial inflammatory component — which is precisely why the idea of extending it to inflammatory conditions is attractive.

It is also, as evidence for MCAS, indirect. The endpoint was glycemic. The population was metabolic. No tryptase, no mediator panel, no symptom instrument, no mast cell disease in the inclusion criteria. Reasoning from this trial to MCAS is a mechanistic hypothesis, not a demonstration.

Beyond this, the published record on amlexanox includes a 2020 narrative review in the Yale Journal of Biology and Medicine that surveys its atopic, metabolic, and inflammatory applications.¹ It is a useful orientation piece and it is the citation most often attached to amlexanox marketing. It is a four-page review article, not primary MCAS data, and it should not be presented as though it were.

The dosing problem

This is the point I would most want a prescribing colleague to notice.

SYSTEMIC AMLEXANOX DOSING IN THE PUBLISHED HUMAN RECORD

The regimen circulating in compounding channels does not correspond to anything in the human literature. It is lower in total daily exposure than the studied doses and it abandons the three-times-daily schedule that every human use of this drug — commercial and investigational — has employed. Amlexanox has a short reported elimination half-life, on the order of a few hours, which is presumably why divided dosing was used in the first place.

Safety, stated precisely

In the 42-participant randomized trial, no serious significant adverse events were reported with three-times-daily oral dosing over 12 weeks.¹ That is reassuring as far as it goes, and it is worth stating plainly rather than hedging. It is also the entire controlled safety database for systemic amlexanox in modern Western practice: one trial, forty-two people, three months, a metabolic population.

The review literature also documents an allergic reaction to the drug itself — a young woman treated with oral amlexanox for allergic rhinitis who developed an eruptive skin rash, with patch testing confirming amlexanox as the culprit.¹ A hypersensitivity reaction to a drug is a routine footnote in most populations. In a population defined by aberrant mast cell reactivity, it is worth more attention than a footnote.

What we do not have: any long-term safety data, any data in pregnancy or lactation, any pediatric data, and any data at all in people with mast cell disease, who as a group tolerate excipients, fillers, and novel compounds less predictably than the general population.

The compounding and regulatory questions

Because amlexanox is not currently available as a marketed product in the United States,¹ a compounded oral capsule raises questions that are worth resolving in writing before prescribing rather than after. Under section 503A, a compounder may generally use a bulk drug substance only if it complies with an applicable United States Pharmacopeia or National Formulary monograph, or is a component of an FDA-approved drug product, or appears on FDA’s 503A bulks list.⁶

I could not confirm from FDA’s published materials where amlexanox currently sits within that framework. That is not an accusation — it is a gap I would want closed. Before writing a prescription, I would ask the pharmacy directly, in writing, for:

• the specific basis on which they consider amlexanox permissible as a bulk drug substance under 503A;
• the certificate of analysis for the lot being used;
• confirmation that the bulk supplier is registered with FDA;
• the full excipient list, which matters more in MCAS than in almost any other population.

Where this might reasonably fit

None of the above makes amlexanox unreasonable to consider. It makes it a late-line, explicitly experimental option.

In my view it does not displace anything with a stronger evidence base — H1 and H2 blockade, cromolyn, ketotifen, and leukotriene-directed therapy all remain ahead of it. The client for whom it becomes a defensible conversation is one who has worked systematically through those options, has documented inadequate control, and has a clinical picture with prominent mediator-type features. If that client also carries the metabolic and inflammatory phenotype that predicted response in the diabetes trial, that is a reasonable — though entirely speculative — selection heuristic.

If a trial is undertaken

These steps are sequential for a reason: each one exists to make the eventual answer interpretable.

1. Resolve the sourcing questions first. Certificate of analysis, excipient list, and regulatory basis before the prescription, not after.
2. Document the informed consent conversation explicitly. The client should understand that no MCAS trial exists, that the formulation has no United States approval, and that this is a trial rather than a treatment.
3. Define two or three target symptoms and score them at baseline. Use a written instrument. Recall is unreliable in chronic multi-system illness, in both directions.
4. Use a dose with some grounding. The studied escalation was 25 mg three times daily, then 50 mg three times daily. Starting lower is defensible in a sensitive population; ending at once-daily dosing is not.
5. Set a hard stop point before starting. Eight to twelve weeks, with a pre-specified threshold for what counts as benefit. Write it down at the start, when it is still an honest question.
6. Stop if the threshold is not met. The failure mode in refractory illness is accumulation — adding therapies and never subtracting them, until no one can tell what is working.

The bottom line

Amlexanox is a real drug with real pharmacology and a legitimate scientific second life. What it does not have is evidence in mast cell activation syndrome. The material promoting it for MCAS leans on a metabolic literature that studied something else, and recommends a dose that appears in no published human study.

My position is not that it should not be tried. It is that it should be tried honestly: last in line, at a dose someone has actually studied, with a defined endpoint and a defined stop. Luck favors the prepared bold — but the preparation has to come first.

References

1. Dosanjh A, Won CY. Amlexanox: a novel therapeutic for atopic, metabolic, and inflammatory disease. Yale J Biol Med. 2020;93(5):759–763. PMID: 33380937; PMCID: PMC7757066.
2. Oral EA, Reilly SM, Gomez AV, et al. Inhibition of IKKε and TBK1 improves glucose control in a subset of patients with type 2 diabetes. Cell Metab. 2017;26(1):157–170.e7. doi:10.1016/j.cmet.2017.06.006. PMID: 28683283; PMCID: PMC5663294.
3. Reilly SM, Chiang SH, Decker SJ, et al. An inhibitor of the protein kinases TBK1 and IKK-ε improves obesity-related metabolic dysfunctions in mice. Nat Med. 2013;19(3):313–321. doi:10.1038/nm.3082. PMID: 23396211; PMCID: PMC3594079.
4. Zhao P, Wong KI, Sun X, et al. TBK1 at the crossroads of inflammation and energy homeostasis in adipose tissue. Cell. 2018;172(4):731–743.e12. doi:10.1016/j.cell.2018.01.007. PMID: 29425491; PMCID: PMC5808582.
5. ClinicalTrials.gov. Amlexanox in obesity and diabetes mellitus type 2 and non-alcoholic fatty liver disease. NCT01842282.
6. U.S. Food and Drug Administration. Bulk drug substances used in compounding under section 503A of the FD&C Act.