Methodology · MCAS Reference

What this is, and what it is not

This is an index of treatments and triggers discussed in mast cell activation disease. For each entry it records what the proposed mechanism is, how it relates to mast cells, what kinds of studies are cited, and which conditions those studies and approvals concern. Patient-facing sources often compress those different questions into one confident conclusion.

It is not a treatment guide. Specifically, it contains:

No dosing information, anywhere, for anything.
No efficacy claims. No entry says a thing works, helps, or is effective. Entries describe mechanisms and cite evidence; the reader and their clinician draw the conclusions.
No untraceable personal experience. Anecdote, forum consensus, second-hand claims about a clinician, and practitioner opinion without a durable first-party source do not create entries here. Documented specialist use is shown separately and never counted as trial evidence.

A build-time check greps every entry for dosage patterns and efficacy verbs and fails on a match, so these are not aspirations either.

The inclusion bar

An entry needs a citable published mechanism of action. In practice that means at least one citation of type peer-reviewed, review, drug-label, reference — a peer-reviewed study, a review article, a drug label, or a reference work.

Organization pages and preprints can appear as supporting citations but do not clear the bar on their own. Neither does anything else: if the only support for a substance is that people report it helping, it is not listed, and no amount of popularity changes that.

A specialist-use source can support the practice tag without clearing this bar by itself. The entry still needs a qualifying source for its mechanism, so an interview or clinic-authored page cannot turn an otherwise unsourced treatment into a mechanism claim.

How this site separates evidence

No single badge here scores a treatment for MCAS. That would collapse different questions into one verdict in a field where nothing is approved for MCAS, and almost nothing has been tested in it.

Instead, each entry keeps these facts separate so a reader can see what the evidence actually establishes:

How it relates to mast cells — where it acts or where the cited work was done.
Types of studies cited — the designs actually represented in the sources.
Approved for / trials in — the conditions it is actually licensed for, or has published randomised trials in.

None says “effective”. An approval is a decision a regulator made and a trial is a study that reported; drawing the conclusion is the reader's job and their clinician's, not this site's.

A third, optional fact—documented specialist use—appears only when a durable source records a named MCAS clinician-researcher's own practice. It does not alter the relationship or study design.

How it relates to mast cells

This field describes either where the cited work was done or, for downstream treatments, where the intervention acts. It does not describe clinical importance.

MCAS patients: Studied in people diagnosed with mast cell activation syndrome. Current evidence consists of case reports, uncontrolled series, or retrospective review rather than randomised comparisons — so it establishes that people have taken it, not that it worked.
Mast cell disease: Studied in patients with a different mast cell disease, usually systemic mastocytosis. Those disorders have diagnostic criteria and disease biology that differ from MCAS, so a result there does not establish an MCAS outcome.
Mast-cell-mediated condition: Given to people with a condition in which mast cells are central to the disease process, with a human mast-cell effect also measured. This is human evidence, but it is not evidence in MCAS or in a clonal mast cell disorder, and it cannot establish an MCAS outcome.
Mast cells in the laboratory: Studied on mast cells directly, but in cell culture or animals rather than in people. The study-design field distinguishes human cells from non-human models. A mechanism demonstrated in a dish is a reason to investigate, not a result — and the concentrations used are frequently ones that human exposure does not reach.
Related inflammatory condition: Studied in people with a condition in which mast cells can contribute, but the study did not measure an effect on mast cells. This is a clinical bridge for inclusion, not a direct observation of mast cells and not an MCAS outcome.
Downstream of the mast cell: Acts after mast-cell mediators are released rather than stabilising the mast cell itself. This includes treatments that block mediator receptors, degrade a mediator, or oppose the physiology of an acute reaction. It describes where the intervention acts, not its clinical importance.

Why treatments are not scored

A previous version derived a visible score from the strongest study design that directly involved mast cells. The calculation was internally consistent, but its presentation inverted the clinical context a new reader needed. Epinephrine and routine H1 and H2 antihistamines act after mediators are released, so the badge assigned them its bottom category. Experimental refractory treatments with direct case reports appeared to rank above them.

The score described direct mast-cell observation, not usefulness, clinical importance or the overall evidence for the treatment's real role. A short badge could not hold that distinction, so it is gone. Every input fact remains visible as relationship, study design, studied condition, approval and evidence limits.

Approval appears only beside the named conditions it applies to. A second generic approval marker would repeat that information without telling the reader anything more, and approval for one condition is not evidence of approval for another.

Study design

Recorded per entry and displayed beside the relationship above. This used to be inferred from the population instead, which meant every laboratory entry claimed both cell lines and animal models regardless of what was actually done — wrong in opposite directions for a cell-line-only entry and an animal-only one.

Randomised controlled trial: Participants allocated at random to treatment or comparison, with published results.
Cohort study: A defined group followed over time, without randomisation.
Cross-sectional survey: A group assessed at one point in time. Treatment histories and ratings are retrospective and can be affected by recall and selection.
Case series: A set of individual patients described together, with no comparison group.
Case report: One patient, or a very small number, described individually.
Human mast cells in vitro: A human mast-cell line or mast cells isolated from human tissue and studied outside the body.
Non-human mast cells in vitro: Mast cells taken from a non-human animal and studied outside the body.
Animal model: Live animals rather than people or isolated human cells.

The order treatments appear in

Medications are grouped by the sequence a 2025 open-access practical-management review describes, because that is the order a reader is most likely to meet them in. This site is reporting that sequence, not recommending it — the same footing on which the drug classes appear on the appointment page. Parts of it are specific to Canadian practice, and entries say so where it matters.

Anything outside that sequence is grouped by its real catalog role rather than being ranked as though all entries answered the same question. Emergency treatment appears before maintenance; trigger-specific, emerging or refractory, and local-route treatments have separate groups. Unclassified entries, including many clonal-disease agents and every supplement, remain outside the sequence.

Order within a group follows the same source. Where the review describes an order inside a step — step 1 prioritises second-generation antihistamines over first-generation ones — entries carry a stepOrder recording it, and the build rejects a stepOrder on any entry that is not in the sequence at all. Where the review describes no order, an internal rank uses the strongest cited study design that directly involved mast cells, followed by alphabetical order. Sorting purely by that rank once put a sedating first-generation antihistamine above a second-generation one, which reversed the literature the group was built from; sourced stepOrder now prevents that. The fallback rank affects browsing position only. It is not displayed as a score and does not describe usefulness, effectiveness or clinical importance. Every card shows the relationship and cited study types, so the basis of the order remains visible.

Emergency intervention: Used for a time-critical acute event, particularly one meeting anaphylaxis criteria. It appears before the maintenance sequence because emergency treatment and prevention answer different questions; this placement is not an instruction to diagnose an event or use a device.

Trigger-specific treatment: Directed at a defined sensitization or trigger, such as venom allergy. It may change risk from that trigger but is not a general treatment for every MCAS episode.

Emerging / refractory evidence: Included because a published report or survey documents use in people with MCAS, usually after extensive prior treatment. The evidence is uncontrolled, and this placement is not a recommendation or a routine treatment step.

Local-route symptom treatment: The studied product is delivered to a particular site, such as the eye, nose, or airway. Its route-specific evidence does not establish systemic MCAS treatment or equivalence to an oral or injected formulation.

The build rejects an entry that carries both a numbered treatment step and one of these outside-sequence contexts.

1. H1 antihistamines — described as the initial treatment, with second-generation agents prioritised over first
2. H2 antihistamines — described as add-on therapy where episodes involve gastrointestinal symptoms
3. Leukotriene receptor antagonists — the next class described; in Canada montelukast is the only available option
4. Cromolyn — described for persistent gastrointestinal symptoms despite H1 and H2 blockade
5. Ketotifen — described for patients not controlled on high-dose second-generation H1 antihistamines
6. Aspirin — described as limited evidence, for persistent episodes despite H1 blockade
7. Oral corticosteroids — described for frequent episodes, with use limited by adverse effects
8. Omalizumab — described for severe recurrent reactions despite the preceding treatments

Approved for, and trials in

Approved in the US for: Carries this indication on its current FDA label.
Approved outside the US for: Approved by at least one non-US regulator for this indication.
Randomised trials in: Has published, evaluable randomised trial results in this condition. It says trials exist and reported, not that the drug worked.

Both are checked at build time. An entry listing an approved indication must cite the current drug label, and one listing randomised trials must cite the published report — because an approval list is exactly the kind of thing that gets written from memory, and a wrong one is a false claim about what a regulator decided.

One page per evidence base

Where several drugs share one evidence base, this site carries one entry and names the rest on it. Cetirizine's page stands for the second-generation H1 antihistamines — levocetirizine, fexofenadine, loratadine, desloratadine and others — because six pages would repeat the same studies six times and imply six separate bodies of evidence.

The drugs it stands for are listed on the page under Also covers, so searching for a brand name finds the entry that actually applies. They are not recorded as alternative names: fexofenadine is a different drug from cetirizine, not another word for it.

A class member gets its own page when the evidence separates it. Rupatadine is separate from the other second-generation antihistamines because it also antagonises the platelet-activating-factor receptor and has a randomised mastocytosis trial; ketotifen is separate from the first-generation antihistamines because of its additional stabiliser rationale; zileuton is separate from montelukast because it blocks leukotriene synthesis rather than the receptor.

Documented specialist use

Documented specialist use means a durable first-party source documents that a treatment was used or discussed in a clinician's own MCAS practice. It exists because “specialists use this despite thin published evidence” and “nobody has tried this” look identical when both have weak evidence, and those are different things to know. It is not proof of efficacy, a regulatory approval, a guideline recommendation, or professional consensus.

The tag records where the reported use came from and at what scale — “47 patients drawn from six specialist MCAS practices” — rather than naming the clinicians. It named them until August 2026. The attributions were accurate, but naming them contradicted a position this page takes a few sections above: that this site does not adjudicate between the consensus-1 and consensus-2 criteria. The clinicians named were the authors and advocates of one of them, and citing that camp's leaders as the practice authority quietly picked the side the site says it does not pick. The authors remain one click away in the citation.

The build requires each tag to describe its provenance, identify the kind of first-party source, and point to an exact URL already present in that entry's citations. Drug labels and trial registries cannot support the tag.

Treating-author report: A named clinician is an author and the source identifies that clinician as having treated or managed the reported patient or cohort.
Authored clinical guidance: A named clinician describes the treatment in authored clinical guidance or a practice-focused publication rather than an outcome comparison.
Recorded first-party discussion: A durable recording or transcript captures the named clinician discussing their own practice; a second-hand recap does not qualify.

What the evidence cannot establish

Mast cell activation disease relapses and remits. That single fact makes uncontrolled evidence unusually treacherous here: a patient who starts something while symptomatic and improves over the following weeks would, in many cases, have improved anyway. Open-label series record the improvement and attribute it to the intervention. Publication then favours the series that improved.

So every entry must state, in plain prose, what its evidence cannot establish. This is a schema rule: an entry that omits it fails the build. It is not possible to quietly ship one.

This replaced an earlier badge that read “Confound risk: high”. That label was a mistake worth describing, because it failed in a direction that matters: readers took it to mean the drug was dangerous, when it meant the study design could not support a causal claim. A patient-facing site that accidentally implies a safety warning has done real harm to the reader's decision-making, so the ordinal score is gone and the same information is now prose under a heading that says what it means.

Regulatory status

Kept separate from relationship and study design, because they answer different questions. A drug can be approved and barely studied in this population, or unapproved and backed by a randomised trial. “Approved outside the US” is its own bucket rather than being folded into “off-label”, and an entry using it must state which regulator approved it and why that is relevant.

FDA-approved for MCAS: Carries a US label for mast cell activation disease.
FDA-approved, other indication: Carries a US label for a different condition. Any use in MCAS is off-label.
Approved outside the US: Approved by at least one non-US regulator but not by the FDA. Entries in this bucket must state the off-label rationale explicitly.
Investigational: In clinical trials; not approved by any regulator.
Over-the-counter drug: A non-prescription drug, regulated under an FDA over-the-counter monograph or an approved application. Reserved for drugs; a dietary supplement is a different thing and has its own status below.
US dietary supplement — not FDA-approved as a drug: Sold as a dietary supplement, which in the US means it does not undergo premarket FDA review for safety or effectiveness. Availability and approval are separate axes, and being easy to buy says nothing about either.
Compounded: Not commercially manufactured in this form; prepared by a compounding pharmacy.
No regulatory approval: Not approved by any regulator for any indication.

Trial statuses, and why they come from one place

Trial statuses on this site are read from ClinicalTrials.gov or from the sponsor directly. Never from an aggregator or a news tracker. The data model does not even have a field value for one — the list of permitted status sources contains exactly two entries, so citing an aggregator is a build error.

This rule exists because of a specific failure. An aggregator reported an active status for a masitinib trial whose registry record had gone quiet; the registry itself was carrying it as status unknown, its own flag for a record the sponsor has not updated past the expected completion date. Aggregators tend to cache the last confident-sounding status and show it indefinitely, which is worse than showing nothing.

Every trial status also carries the date it was read. A status older than 180 days fails the build, which forces a re-check rather than letting a stale status sit there looking current.

A registered trial is also not, by itself, evidence. One masitinib study in MCAS was registered and then stopped updating; a palmitoylethanolamide study terminated after two participants, which cannot estimate an effect of any size. Both are facts about trials, not results from them. The schema enforces the distinction: listing a condition as having randomised trials requires citing the peer-reviewed primary report, so a protocol or an abandoned record cannot be presented as a result.

Freshness

Every entry records the date a human last checked it. After 120 days it is marked due for re-check; after 180 days, overdue. Both states are visible on the entry itself, so a reader can see the age of what they are reading without taking anyone's word for it.

A scheduled job re-checks every cited link weekly and opens a tracking issue for anything broken or overdue. Dead citations are the normal failure mode of a reference site, and the only defence is automation.

A note on diagnostic criteria

MCAS does not have one agreed definition, and this affects how any source about it should be read. Two criteria sets are in active use. The AAAAI work group report requires objective evidence of mediator release, including a documented rise in serum tryptase from the patient's own baseline during an episode. The “consensus-2” criteria are broader, and admit a substantially larger patient population.

This site does not adjudicate between them, because it indexes mechanisms and evidence rather than making diagnoses. What it does instead is say so plainly: when two sources about MCAS appear to contradict each other, the disagreement is often about which patients are being described, not about the biology. The paper Using the Right Criteria for MCAS is the best single entry point to that debate.

Potential food-trigger reports

Potential trigger is kept separate from every measured-content axis. A food can contain little measured histamine and still appear on an elimination list, but that does not establish that it releases histamine from mast cells. The table therefore says Reported, never “yes” or “histamine liberator.”

The evidence boundary is unusually important here. A comparative review found that foods such as strawberry have been described as possible endogenous- histamine triggers, while also finding no supporting human clinical studies and only a few inconclusive laboratory or animal studies. A review focused on mastocytosis likewise found no double-blind placebo-controlled food challenges of proposed histamine-releasing foods. The signal records that the claim exists; it does not promote the claim into established biology.

The schema requires every signal to use a registered, restatable source whose full text was opened, plus a food-specific explanatory note. It rejects unknown sources, link-only sources and duplicate signals. A blank cell means no qualifying signal is recorded here—not that the food is guaranteed to be tolerated by every person.

Sources: comparative low-histamine-diet review and mastocytosis evidence review .

Food rating sources and their terms

Food ratings are shown per source rather than merged, and each source is tagged with whether its values may be reproduced here. Sources marked link-only are cited and linked but never restated: converting a proprietary database's measured values into low / moderate / high reduces their precision without changing where they came from. The schema enforces this — a rating attached to a link-only source fails validation.

There is a second rule, and it cost this directory some entries. A food rating is only attributed to a source whose full text was actually read — not its abstract, and not a secondary description of it. Several of the standard food-composition papers, including the 1985 salicylate survey most published salicylate lists descend from, sit behind publisher paywalls. Citing them for a specific value without having opened them would be the same failure this site exists to avoid, one level down, so those axes are absent rather than approximated. A thinner table that can be checked is worth more than a complete one that cannot.

Biogenic Amines in Plant-Origin Foods: Are They Frequently Underestimated in Low-Histamine Diets? (Foods, 2018) May be restated

Peer-reviewed food-composition review, open access

CC BY 4.0 and read in full text. Table 4 reports measured ranges of histamine, tyramine, putrescine, cadaverine, spermidine and spermine in non-fermented plant foods, in mg/kg fresh weight. Other-amine ratings here follow the paper's own high-versus-low groupings rather than any clinical threshold, and tyramine stays on its separate axis. Two reading rules matter: a dash means the value was not reported and is never recorded here as low, and 'not detected' means below the cited study's detection limit rather than absent. The paper proposes that putrescine and the polyamines may compete with histamine for the DAO enzyme; that is a hypothesis, and notes here describe content first.
Dietary Oxalate Intake and Kidney Outcomes (Nutrients, 2020) May be restated

Peer-reviewed review, open access

Open access and read in full text. Reports oxalate content per 100 g of food, which is why its rankings can differ from sources reporting per serving. Its table gives numbers rather than categories, so the low/moderate/high shown here is a mapping: at or above 50 mg per 100 g is high, using the threshold the companion kidney-stone review publishes for its own high band; 20 to 50 mg is moderate and below 20 mg is low. Those two lower boundaries are this site's, not the paper's, and every entry states the measured figure so you can apply your own.
Histamine in Brazilian Foods: Occurrence and Risk Assessment for Intoxication and Intolerance (Food Science & Nutrition, 2025) May be restated

Peer-reviewed systematic food-composition review, open access

CC BY 4.0 and read in full text. A PRISMA systematic review compiling histamine measurements in Brazilian-market foods from 1980 to 2024, in Tables 3 and 4. The paper uses four bands; this site merges its two lowest, mapping under 10 mg/kg or mg/L to low, 10 to 50 to moderate, and above 50 to high, with variable where the reported studies or range cross a band. This is a regional dataset rather than a universal property of the food, and sample sizes and product forms vary. Composition labels only — the paper's safety and intolerance language is deliberately not carried across.
Histamine Intolerance: The Current State of the Art (Biomolecules, 2020) May be restated

Peer-reviewed narrative review, open access

CC BY 4.0 and read in full text. Used here only when the review explicitly names a food as having been reported to trigger endogenous histamine release. The article states that the responsible mechanism has not been elucidated, so this source supports a reported potential-trigger signal rather than a proven food property or measured-content rating.
Low-Histamine Diets: Is the Exclusion of Foods Justified by Their Histamine Content? (Nutrients, 2021) May be restated

Peer-reviewed comparative review, open access

Open access and read in full text for every rating attributed to it here. Compares ten published low-histamine diets against measured histamine and biogenic amine content.
Mast Cell 360 low-histamine food list Link only

Practitioner-authored food list

Widely read by patients, and included because readers will arrive having already seen it. Its authors describe it as drawn from 'the best research available about low histamine foods and our experience in the clinic with hundreds of very sensitive clients', and individual foods are not attributed to particular sources. That makes it a clinical opinion rather than a measurement, which is why it is link-only: its value lies in which foods it selects and how it groups them, and that selection is the authors' own work. Note that it places several foods — bananas and citrus among them — in its higher-histamine section, where laboratory analysis has found no detectable histamine.
Monash University FODMAP Diet App Link only

Proprietary laboratory database

Monash laboratory-tested the foods in this database and licenses access through their app. Their measured values are facts, and facts are not themselves copyrightable — but the database is accessed under terms of use, the compilation and its thresholds reflect their editorial judgement, and the certification mark is theirs. This site links out to their published high- and low-FODMAP food page rather than restating, which also keeps every rating here traceable to something a reader can open.
Oxalate in Foods: Extraction Conditions, Analytical Methods, Occurrence, and Health Implications (Foods, 2023) May be restated

Peer-reviewed food-composition review, open access

CC BY 4.0 and read in full text. Table 5 compiles total, soluble and insoluble oxalate per 100 g fresh weight across many analytical studies. Ratings use the same per-100-g mapping this site applies to its other numeric oxalate source — 50 mg or more is high, 20 to under 50 is moderate, under 20 is low — and those bands are this site's, not the paper's. Total oxalate is used where reported and soluble oxalate is never substituted for it. Preparation is kept in every note, because the review reports raw, boiled, dried and canned values that are not interchangeable, and it warns that cultivar, soil, harvest and analytical method all move the result.
Risk of Kidney Stones: Influence of Dietary Factors, Dietary Patterns, and Vegetarian-Vegan Diets (Nutrients, 2020) May be restated

Peer-reviewed review, open access

Open access and read in full text. Bands foods by oxalate per serving rather than per 100 g, so its ranking of dense foods such as bran relative to leafy greens differs from per-weight sources. Both of its bands — above 200 mg and above 50 mg per serving — are shown here as high, with the band named in each entry's note.
SIGHI Food Compatibility List for Histamine Intolerance Link only

Patient organization compatibility list

Widely used by patients and published by the Swiss Interest Group Histamine Intolerance under their own terms. Listed because readers will encounter it, and linked rather than reproduced. Note that comparative review has found published low-histamine lists disagree substantially with each other and with measured histamine content.

Corrections

A reference index with no error-reporting route is asking to be trusted more than it deserves. If something here is wrong, out of date, or overstated — particularly a mechanism or evidence description — open an issue. The full contents of this site, including every entry and the validation rules above, are in the repository.

Disclosures. This site is built and maintained by one person, is not funded by anyone, sells nothing, and has no affiliate relationships with any manufacturer or retailer of anything listed on it.