Myoepithelial carcinoma

1If you were just diagnosed

Nobody opens this page for fun. If you are here, you or someone you love has had bad news, and you are reading under stress. So, three things first.

Written by Peter Saddington, a patient, from what he learned after his own diagnosis in 2021. His account is at the end of this page.

Take ownership

Your doctors treat the disease, but you have to run the project. With a disease this rare¹, many of the people treating you may be seeing it for the first time. You are the only person in every room, and nobody has more at stake than you do. You cannot hand your health to the system and wait.

None of this means distrusting your doctors. It means arriving prepared, asking until you understand, and keeping your own records.

A plan for the first 30 days

1. Get the pathology report in your hands
   Ask for the full written report. Find three things on it: the exact diagnosis line, the stains that were positive, and what it says about the margin. Peter’s read “adnexal carcinoma, consistent with myoepithelial carcinoma,” with invasion close to the margin⁹.
2. Get the slides read a second time
   Ask for a second pathology opinion at a center that sees these tumors. In one series most cases had first been given a different diagnosis¹⁰. Peter’s diagnosis came from a second-opinion review⁹.
3. Ask about fusion testing
   Ask whether the tumor has been tested for gene fusions such as EWSR1 and PLAG1, and for INI1. The result affects how the tumor is classified and what is known about how it behaves¹¹¹².
4. Find a team that has seen it
   Ask each doctor how many cases they have treated. The published series come from specialist units: a sarcoma unit for soft-tissue disease¹³, tertiary cancer centers for salivary disease¹⁴. Ask for your case to go to a tumor board.
5. Know what the scans show
   Ask which scans are needed to look for spread, and keep copies of the images and the reports. PET, CT and MRI were ordered for Peter⁹.
6. Talk about margins, before surgery and after
   Complete removal with clear margins was associated with better survival in the pooled data³. Ask whether the margin was clear, what happens if it was not, and whether radiotherapy afterward is worth discussing³.
7. Ask about trials, and about your tissue
   The authors of the adult soft-tissue series say trial enrollment should be encouraged¹³. Ask whether your tissue can be banked for research. cureMEC is building biospecimen collections⁵, and Peter donates his⁹.
8. Set a surveillance schedule and keep it
   This disease returns often. In a 101-tumor soft-tissue series, 42% of malignant tumors recurred, and follow-up ran as long as 21 years¹⁵. Agree how often you will be checked, and do not let it slide when you feel well.

Questions to take to your next appointment

• What exactly does my pathology report say, and has it had a second opinion?
• Where did it start, and how large and how deep was it?
• Were the margins clear?
• Has it been tested for EWSR1, PLAG1 and INI1?
• What do my scans show?
• How many patients with this diagnosis have you treated?
• Will a tumor board review my case?
• What are my options, and what happens if we wait?
• Is radiotherapy worth considering after surgery?
• Is there a trial I could join, and can my tissue go to research?
• How often will I be checked, and with which scans?
• Who do I call between appointments?

Staying calm enough to think

Fear is a normal response to this news, and it makes people decide badly. The goal is to be steady enough to make the next decision well, and that is possible while still afraid. This is the list Peter keeps¹⁶.

Tell people, and give them jobs. One drives, one takes notes, one makes the insurance calls. Peter, who by his own account rarely asks for help, wrote: “I got cancer. I need help.”⁹

Protect sleep, food and movement. They are the part of this you control today.

If you are the family or the friend

• Go to the appointments and write everything down. It is hard to listen and take notes when the news is about you.
• Take a job off their list: scheduling, records, insurance, meals, rides.
• Do not send cures you found online. Ask what would help this week.
• Let them talk about the numbers, including the bad ones. Then bring it back to the next step.

Where the hope is

The reasons for hope are in the data.

Four years after his diagnosis, past the date he had been told to expect, Peter wrote: “We’re making history by simply not dying.” His routine is plain: scans every three months, and “you cut what you find.” His advice is four words: “Stay the course. Focus.”¹⁸

This is one patient’s checklist, informed by the published research. It is not medical advice. Every step is something to do with your doctors, not instead of them.

2What it is

A cancer of the cells that squeeze glands, at the malignant end of a spectrum that begins with benign mixed tumors¹².

Myoepithelial cells sit between the secreting cells of a gland and the membrane around it. They contract, which pushes secretions out. Glands that have them include the salivary glands, the sweat glands of the skin and the breast¹⁹, which is why tumors of these cells turn up in such different parts of the body.

Tumors of myoepithelial cells form a spectrum. A mixed tumor (called pleomorphic adenoma in the salivary glands and chondroid syringoma in the skin) contains ducts as well as myoepithelial cells. A myoepithelioma is purely myoepithelial and benign. A myoepithelial carcinoma is the malignant form¹²²⁰.

The dividing line depends on the site. In the salivary glands, the tumor must be unequivocally malignant and exclusively myoepithelial¹⁰. In soft tissue, tumors with at least moderate cytologic atypia are classified as malignant¹². In one salivary series most cases had first been diagnosed as something else, usually malignant mixed tumor¹⁰.

3Names and codes

The same disease appears under several names, which is one reason it is hard to search for.

4Where it occurs

Most often the salivary glands, then soft tissue, skin and breast³.

The answer depends on who is counting. In the pooled literature the salivary glands lead³. In the U.S. National Cancer Database the most common coded site was the breast, followed by the vulva, and 70.5% of patients were women⁴. Within the major salivary glands, 76.6% of cases were in the parotid²⁵. In soft tissue, most tumors arise in the limbs and limb girdles¹⁵.

The chart counts papers, not patients: the 817 human studies in this catalog that name the disease in their title, tagged by the body sites their title or abstract mentions²¹.

5How rare it is

About one new case for every 7.7 million people each year¹.

The SEER study found the same incidence in men and women, and a higher rate in the mouth and throat than at other sites. For 2012 alone it reports higher rates: 0.058 per 100,000 in men and 0.025 in women¹. The National Cancer Database study saw a gradual rise in reported cases over its 17 years⁴. In the skin the disease is rarer still: a 2026 review found 26 known cases of primary cutaneous myoepithelial carcinoma²⁶.

6How it is diagnosed

A panel of stains under the microscope, usually keratin with S-100¹². In one series most tumors had first been given another diagnosis¹⁰.

Diagnosis is made under the microscope and confirmed with immunostains. Most myoepithelial tumors stain for both epithelial markers (keratins or EMA) and S-100 protein. GFAP and p63 are often positive, and a subset of malignant tumors lose INI1¹². No single stain is decisive, so pathologists read a panel.

The chart shows how often each marker was positive in a soft-tissue series of 101 tumors, 40 of them malignant¹⁵.

Salivary tumors show a similar pattern. In 25 salivary carcinomas, keratin AE1/AE3 and S-100 were positive in all, calponin in 75% and smooth muscle actin in 50%¹⁰. In a 51-case series from a tertiary cancer center, calponin was positive in 98% and S-100 in 82%¹⁴.

7Gene fusions

Two genetic families, EWSR1 and PLAG1, that arise in different places and behave differently¹¹.

Since 2008, genetics has changed how the disease is understood. Many tumors carry a fusion, where two genes are joined and the hybrid drives growth. Which fusion a tumor has tracks with where it arises, how old the patient is and how it behaves¹¹.

A 2026 methylation study of 52 fusion-positive skin, soft-tissue and bone tumors concluded that tumors with EWSR1 or FUS fusions are epigenetically unrelated to the PLAG1-driven tumors of skin and salivary gland, and that their malignant forms are better classified as sarcomas than carcinomas¹¹. A separate 2026 sequencing study of 27 patients placed the disease closest to salivary adenoid cystic carcinoma, found IGF1R over-expressed with or without a fusion, and made its sequencing data available to researchers¹⁷. The two studies point in different directions, and the question is open.

8Survival and outcomes

Across all sites, about two in three patients were alive at five years and about half at ten⁴.

Out of 100 people diagnosed

Survival across all body sites in the U.S. National Cancer Database: 336 patients, 2004 to 2020⁴. The abstract does not say whether deaths from other causes are counted.

In the pooled literature, median overall survival was 167 months, about 14 years³. The table sets out what each major series reported.

Deaths from the disease, series by series

Share of followed patients who had died of the disease at last follow-up. Follow-up was a few years in every series, so later deaths are not counted.

9Treatment: what studies report

Surgery with clear margins first, radiotherapy sometimes, and no standard drug treatment yet³⁵.

This section reports what studies found. It is not advice, and nobody should change treatment on the strength of a web page.

• Surgery is the mainstay. Complete resection with clear margins was associated with better progression-free and overall survival³, and complete resection is the standard for localized soft-tissue disease¹³. In the National Cancer Database, 77.9% of patients had surgery⁴.
• Radiotherapy after surgery reduced local and regional recurrence in the pooled analysis³. A SEER analysis linked it to better survival in high-grade tumors²⁴. In major salivary gland cases the link appeared only after statistical adjustment²⁵. 36.6% of patients in the National Cancer Database received it⁴.
• Chemotherapy has limited evidence. In 24 adults with soft-tissue disease, nine received systemic treatment, one had a partial response (to doxorubicin), and median progression-free survival on first-line chemotherapy was 9.3 months¹³.
• Targeted treatment is at the case-report stage. One group used RNA expression analysis to choose a targeted drug for a four-year-old patient³⁴. A 2026 sequencing study proposed IGF1R as a target worth exploring¹⁷.
• Mohs surgery has been reported for a recurrent cutaneous myoepithelioma³⁵.

There are no established therapeutic standards for the disease⁵. Authors of the adult soft-tissue series concluded that more effective systemic therapies are needed and that enrollment in clinical trials should be encouraged¹³.

10In the skin

A 2026 review found 26 reported cases of primary skin carcinoma²⁶.

The earliest paper in this catalog, from 1957, describes a malignant myoepithelioma of the sweat glands²². The first series of skin tumors in the catalog came in 2003 and 2004. One had 20 tumors, two of them carcinomas, and its authors wrote that the idea of myoepithelial tumors of the skin was still debated¹⁹. The other had 14²⁰. In the 14-case series most tumors were on the limbs, the median age was 22.5, and one of the 14 was a carcinoma²⁰.

Skin tumors split genetically. Of the tumors that form ducts (mixed tumors, also called chondroid syringomas), half carried PLAG1 rearrangements in one study, as salivary mixed tumors do²⁸. EWSR1 rearrangement occurs in a subset of cutaneous myoepithelial tumors³⁶, and EWSR1 fusions with PBX1 or PBX3 are enriched in skin and bone¹¹.

Primary cutaneous myoepithelial carcinoma has a high rate of local recurrence and metastasis, and a 2026 review counted 26 known cases²⁶. A 2025 report describes a case in the axilla with lymph-node metastasis³⁷.

This catalog flags 281 works as closest to the skin form: myoepithelial carcinoma placed in the skin by its title or abstract, and malignant mixed tumor of the skin²¹. Browse them in the directory.

11In children

Frequently malignant in children, and often aggressive¹²³³.

About 20% of soft-tissue myoepithelial tumors occur in children, and in children they are frequently malignant¹². In 29 pediatric soft-tissue carcinomas, ages ran from newborn to 17, with a median of nine. Of 23 children with follow-up, 52% developed metastases and 43% died of the disease, at a median of nine months³³.

Genetics separates the childhood disease. EWSR1::KLF15 tumors occur mostly in children under five, and age under 25 predicted worse progression-free survival in a pooled analysis of 185 fusion-positive cases¹¹. The patient group cureMEC describes the pediatric form as fusion-driven and reports building preclinical models³⁸.

12What is not known

A reference should say where the knowledge runs out. These are the questions the papers themselves call open.

• Which drugs work. There are no established therapeutic standards⁵. In what its authors call the largest published cohort of adult soft-tissue disease, nine patients received drug treatment¹³.
• Who should have radiotherapy. The pooled analysis calls the role of adjuvant therapy unclear³, and a SEER study calls the role of radiation unknown while reporting a benefit in high-grade tumors²⁴.
• Whether it is one disease. One 2026 study argues that malignant tumors with EWSR1 or FUS fusions are sarcomas, and that PLAG1-rearranged skin tumors belong with the salivary ones¹¹. Another finds the disease closest to salivary carcinomas¹⁷.
• How to grade it. Grading is described as controversial²³.
• How the skin form behaves. Everything known comes from 26 single case reports, which point to a high rate of recurrence and metastasis²⁶.
• Who gets it. Reports conflict. SEER found the same incidence in men and women¹. The National Cancer Database found that 70.5% of patients were women⁴.
• Why it starts. None of the series cited on this page identifies a cause.
• Who is paying to find out. A search of U.S. federal research grants found no project with the disease in its title³⁹.

13Timeline of the research

The main steps in the research, by year.

14The literature in numbers

Everything on this page rests on a catalog built from four indexes: PubMed, Europe PMC, OpenAlex and Semantic Scholar²¹. Of 10,435 works, 1,366 name the disease in their title or abstract. The rest cover neighboring tumors or mention the disease only in passing. All of it comes from free public sources.

Papers per year

Works naming myoepithelial carcinoma in title or abstract, by year of publication. The count for 2026 is incomplete.

What kind of evidence

671 of the 1,366 are flagged as case reports, which is 49% of the literature. Most of what is known about this disease comes from descriptions of one patient at a time.

Languages

The disease has been written about in 17 languages. 115 of the 1,366 papers are not in English.

Journals that publish it most

How much is open

15Most cited papers

The ten papers naming the disease in their title that other researchers cite most, by OpenAlex and Europe PMC citation counts²¹. Citation counts favor older papers.

16Clinical trials

ClinicalTrials.gov holds 13 records in which the word myoepithelial appears anywhere⁴¹. A record is listed here if it matches, so some mention the disease only as one of many eligible tumor types. The catalog also holds 514 trials for salivary gland cancers and skin adnexal tumors, which may accept this diagnosis without naming it.

Status is as recorded on 11 November 2026. Check the registry for current status and eligibility.

17Researchers and organizations

cureMEC: The Myoepithelial Carcinoma Project is a patient-driven initiative launched in 2022. It reports building biospecimen collections, molecular data and preclinical models⁵, in collaboration with the Rare Cancer Research Foundation³⁸.

The authors below appear most often on PubMed-indexed papers naming the disease²¹. Counts are by name as printed, so they can merge or split people.

Shared data: a 2026 study deposited exome and RNA sequencing from 27 patients in the EGA repository¹⁷.

18For researchers: codes, data and funding

Standard codes

How the main vocabularies name and code the disease⁸. The ICD-O-3 code is the one a registry study used⁴. Use these to search registries, biobanks and genomic portals. MeSH has no heading for the carcinoma alone: it is indexed under Myoepithelioma.

Public datasets

NCBI holds 19 studies that mention myoepithelial tumors, of which 8 name the carcinoma⁴². Exome and RNA sequencing from 27 patients is in the EGA archive¹⁷.

Federal funding

NIH RePORTER lists 4 funded projects whose title, abstract or terms mention the disease or myoepithelioma. None has it in the title³⁹.

This page's data

The full catalog, without abstracts: papers.json, papers.csv, trials.json, datasets.json, terminology.json.

19Common questions

20Glossary

Words that appear in pathology reports and in the papers cited here.

Adnexal

Belonging to the skin's appendages: the sweat glands, oil glands and hair follicles.

Carcinoma

A cancer that arises from epithelial cells, the cells that line surfaces and form glands.

Chondroid syringoma

The name for a mixed tumor when it arises in the skin.

FISH

Fluorescence in situ hybridization. A laboratory test that shows whether a gene has been broken or moved.

Gene fusion

Two genes joined abnormally into one hybrid gene. Written with a double colon, as in EWSR1::PBX1.

Grade

How abnormal the cells look and how fast they are dividing. High grade means more aggressive.

Immunostain

A stain that shows which proteins a tumor's cells make. Pathologists use a panel of them to identify the cell type.

INI1

A protein, also called SMARCB1, that restrains cell growth. Some myoepithelial carcinomas have lost it¹².

Local recurrence

The tumor returning at or near the place it was removed from.

Margin

The rim of normal tissue removed around a tumor. A clear or negative margin has no tumor cells at its edge. An R0 resection is complete removal with clear margins.

Median

The middle value. With a median survival of ten years, half of the patients lived longer than ten years and half did not.

Metastasis

Spread of the cancer to another part of the body.

Mitotic rate

The number of dividing cells counted under the microscope. A higher count means faster growth.

Mixed tumor

A tumor with ducts as well as myoepithelial cells. Called pleomorphic adenoma in the salivary glands.

Mohs surgery

A way of removing a skin cancer one thin layer at a time, checking each layer under the microscope before taking the next.

Myoepithelial cell

A contractile cell that sits around a gland and squeezes its secretions out.

Myoepithelioma

The benign tumor of myoepithelial cells.

Overall survival

The share of patients alive at a given time, whatever the cause of any deaths.

Progression-free survival

The time during which the disease does not grow or spread.

Sarcoma

A cancer of connective tissue such as bone, muscle or fat.

Surveillance

Scheduled examinations and scans after treatment, to catch a return early.

Tumor board

A meeting at which surgeons, oncologists, radiologists and pathologists review a case together.

21Directory of papers

Every paper that names the disease, with its authors, journal, year and a link to the source. Filter by body site, by type of paper, or search by word.

Newest papers

22Method and limits

• Open sources only. Everything on this site was gathered from free public sources: PubMed, Europe PMC, OpenAlex, Semantic Scholar, ClinicalTrials.gov and the other public databases named on each page. No paywalled article was accessed. Of the 1,366 papers that name the disease, 208 are held in full text from PubMed Central, 432 more are open access elsewhere, and 726 are behind a paywall. For those 726, this site knows only what the public indexes show: the title, the authors and, where there is one, the abstract. If you wrote one of them and can share it, please write.
• Sources. PubMed, Europe PMC, OpenAlex and Semantic Scholar were searched for the disease and its neighbors, and ClinicalTrials.gov for trials. Records were merged by PubMed ID, DOI and PMC ID²¹.
• Sorting is by keyword. A paper counts as naming the disease if its title or abstract contains "myoepithelial carcinoma" or "malignant myoepithelioma". Body-site tags and case-report flags are also keyword rules. Expect some misfiled papers.
• Figures are quoted from abstracts. Each number on this page is followed by its source. Where a paper's abstract and full text differ, the full text is right.
• Not yet covered. Subscription indexes, non-English databases, conference posters without a DOI and trial registries outside the United States. The counts here are a floor.
• Read by a model. The case registry, the treatment table, the trial list and the pathology tables were extracted from abstracts by an AI model told to record only what the text states. They are marked as such on their pages.
• Checked against the sources. On 11 November 2026 a model checked 141 cited claims against their sources: 65 against the full paper and 76 against the abstract. Thirteen were wrong or not supported by the source cited, and were corrected. A claim confirmed only against an abstract may still differ from the full paper.
• Not yet reviewed by a clinician. If you treat or study this disease and would read it, please write.
• Not medical advice. This is a reference to published research, compiled by a patient. It cannot tell you about your own case. Decisions belong with you and your doctors.

Last updated 11 November 2026. Data: papers.json · papers.csv · trials.json.

23Corrections and additions

24Why I built this

By Peter Saddington, diagnosed September 2021

On 20 September 2021 a doctor phoned to tell me I had cancer. When I asked how I got it, he said it was bad luck, and that there were few studies. Over the next weeks I met five more doctors. Every one of them said the word “rare.” One told me he had never heard of it. None could point to a standard of care⁹.

I was told there was no cure, a 40% chance it would spread or return, and that the average patient lived 3.5 years. I asked for data. What I could find on my own in 2021 was one prognosis paragraph last updated in 2016 and a study of 24 patients⁹. I came away believing that most people with this disease die of it, and soon.

I lived under that number for three and a half years. On the day I was statistically supposed to be dead, I wrote about it⁹. Then I did the research again, properly.

What went wrong

I do not think any one doctor lied to me. This disease turns up in about one person in every 7.7 million each year¹. Most doctors will never see a case, and none has the hours to read 1,366 papers. So they reach for the nearest small study, and the patient hears one sentence. I was given incomplete information at the moment it mattered most, and it shaped the care I thought I should ask for.

What changed

In 2021 I searched by hand and found fragments. In 2026 I did it again with AI. In a day it pulled 10,435 works from four open research indexes, sorted them, and traced every figure on this page back to its source²¹. A patient could not have done that alone in 2021. Now anyone can read the result.

None of this makes it a mild disease. It comes back. Mine has, on my jawline, hairline, leg, back and finger, and I am scanned every three months¹⁸. The skin form I have has fewer than 30 reported cases²⁶, and the largest group in the big datasets is salivary gland cases³. Group numbers cannot tell any one person what will happen. But “about three and a half years” and “half of patients alive at ten years” are different lives, and a patient deserves the second number when it is the true one.

I built this so the next person does not start from a phone call and a paragraph from 2016. If that is you, start with the first-steps guide.

25References

Sources for the numbered markers on this page. All 1,497 papers are listed in the directory.