Buying reconstructed human epidermis: models, validation, kit sizes and calendars

The short answer: six reconstructed human epidermis models are commercially available and named in the current OECD skin irritation guideline — EpiDerm SIT, SkinEthic RHE, LabCyte EPI-MODEL24 SIT, epiCS, Skin+ and KeraSkin SIT. EpiSkin, the original validated reference method, has not been commercially available since December 2024, though data generated with it remains covered by Mutual Acceptance of Data.

The second thing to know: almost nobody publishes a price. One supplier in the category — Phenion — publishes a complete list down to the medium and the Franz cell adapter. Everyone else quotes.

The third thing, and the one that actually determines whether your study runs on time: this is a living product with a four-day shelf life including transit, shipped on a fixed weekly calendar, and released on QC that completes after it has already left the building.

What is available, and what holds which validation

Every method included in TG 439 was required to meet a minimum predictive capacity of 80 percent sensitivity, 70 percent specificity and 75 percent accuracy following independent peer review. That is the floor, and it applies to all six available models equally. Mutual Acceptance of Data is guaranteed for methods validated against the performance standards and reviewed and adopted by OECD — which all six are.

The practical consequence is stronger than most buyers assume: the guideline states that the test methods it includes can be used indiscriminately to address countries’ requirements for in vitro skin irritation test results. For a regulatory endpoint, the choice between these six is not a regulatory-acceptance choice. It is a choice about price, availability, format, shipping geography and local technical support.

The EpiSkin withdrawal is the most consequential recent change

EpiSkin was the Validated Reference Method: the method whose data defined the specificity and sensitivity values of the original performance standards, and which together with EpiDerm was used to develop the guideline itself. It stopped being commercially available in December 2024.

Two things follow. Historical EpiSkin data remains fully valid and accepted — the guideline says so explicitly, and confirms the performance standards remain valid after the model’s removal. But if your validated in-house method, your SOP or your submitted study protocol names EpiSkin, you have a transfer to plan, not a purchase to make. EPISKIN continues to supply SkinEthic RHE, which is the natural first call.

The verified EpiDerm operating specification

MatTek publishes the most complete technical specification in the category, so it is the useful reference point even if you buy elsewhere. From the 2025 EpiDerm Technical Specifications:

Kit sizes and formats across suppliers

Format is a real price and protocol variable, not a packaging detail. The surface area determines the treatment dose, and the treatment doses are specified per model in the OECD guideline.

SkinEthic RHE has the widest format range by a clear margin — five surface areas from 0.11 cm² to 4 cm². If your assay needs a large surface for extraction or a 96-well footprint for throughput, format availability may decide the supplier before anything else does.

The J-TEC EPI-MODEL 6D deserves a mention for a specific use case. A shortened six-day culture produces a thinner cornified layer and consequently greater cytotoxicity sensitivity. That is deliberately not the standard model and is not the TG 439 method — but for detecting weak effects that a fully cornified model absorbs, it is a purpose-built tool.

The only complete public price list in the category

Phenion publishes its full price list as a downloadable PDF, valid from 1 January 2024. It is the only such document we found from any supplier in this market, and it is worth reproducing because it shows what a fully transparent tissue price structure looks like.

Published volume discounts: above €3,000 = 5 percent; above €6,000 = 7 percent; above €10,000 = 9 percent. Production site: Henkel, Düsseldorf.

Two things this table teaches that no quote-only supplier will show you.

The media variants carry a defined premium. Standard ALI medium is €75 per 250 mL; the phenol red-free, hydrocortisone-free and antibiotic-free variants are €110 each. That is a 47 percent premium on the medium for removing an agent. When a quote-only supplier offers you a media variant, you now know roughly what shape that premium takes.

Do not read these as RhE prices. Phenion sells full-thickness models — dermis plus epidermis. These are not TG 439 methods, Phenion does not present them as such, and substituting them for an RhE model in a regulatory context is not an option. The FT LONG-LIFE at €190 answers a different question entirely: it is culturable for at least 50 days after arrival, against EpiDerm’s four-day shelf life and roughly two weeks of good morphology retention. Fifty days of usable culture is a different experimental capability, not a cheaper version of the same thing.

Shipping and QC on arrival: the operational part

This is where RhE purchases actually go wrong, and it is entirely predictable.

The calendar sets your experiment date

EpiDerm ships every Monday, arrives Tuesday morning in the US, Tuesday to Thursday internationally, and has a four-day shelf life at 4 °C including transit. Not four days from arrival — four days total. Ship Monday, arrive Tuesday morning, and you have until roughly Friday.

MatTek’s own guidance is to use tissues consistently on the same day — Tuesday afternoon, or Wednesday morning after overnight storage — for best reproducibility. The consequence: plan your experiment backwards from a Monday shipment, and treat the purchasing decision and the experiment date as one decision. Ordering on a Tuesday gains you nothing against ordering the previous Friday.

International buyers should add the geography penalty explicitly. Tuesday-to-Thursday delivery against a four-day shelf life can leave a one-day usable window. Confirm the transit time to your specific address before committing to a study schedule.

QC completes after the tissue has shipped

This surprises people every time. EpiDerm lot release testing — the ET-50 acceptance range of 4.77 to 8.72 hours — happens after the tissues are already in transit to you. ET-50 failures are notified by Wednesday 5 p.m.; sterility failures within 8 days of shipment. Which is to say, after you have run the experiment.

MatTek’s replacement policy is fair: customers are notified and tissues replaced without charge where appropriate. But a lot failure costs you a week of schedule regardless of who pays for the tissue. If you are on a submission deadline, that is the risk to plan around, and the mitigation is scheduling slack, not a better supplier.

Acceptance testing on arrival

Do this on every delivery, and do it the same way every time:

1. Check the temperature indicator and the packaging before opening. These ship at 4 °C on medium-supplemented agarose gels, not on ice packs at random.
2. Record the arrival date and time, because your usable window is measured against the ship date, not the arrival.
3. Inspect every insert visually for gel damage, tissue detachment, dryness or discolouration. Photograph anything anomalous immediately — a claim is far easier to make on the day.
4. Transfer to fresh maintenance medium per the supplier protocol, and note the time.
5. Run your own negative and positive controls in the same plate as the test article. With post-shipment lot release, your controls are the only same-day evidence you will have that the tissue behaved.
6. Keep the lot number against the experiment record. MatTek assigns lot numbers per production week with letters distinguishing kits within a lot, so the lot identifies both the batch and the week.
7. Log the ET-50 or sterility notification when it arrives, against the experiment it applies to.

Why almost nobody publishes an RhE price

Three structural reasons, and each is usable in a negotiation.

1. The product is a production slot, not a unit. Tissue is grown in weekly lots. What you can buy depends on what is being grown that week, so publishing a price would expose capacity allocation.
2. Academic and commercial pricing genuinely differ. Distributors of these models commonly note that academic pricing is available on request. Where a supplier maintains a two-tier structure, publishing one number breaks the other.
3. Format multiplicity. Five surface areas at one supplier, four media variants at another. There is no single number to publish.

The implication for the buyer: you are expected to ask, and the first number is not necessarily the only number. State your institution type, your annual volume and your format in the first email.

Specification checklist for an RhE quote

1. Model and supplier, named explicitly. Not “an RhE model”.
2. Which OECD guideline you are running — TG 439, TG 431, TG 442D, or ISO 10993-23 for medical device extracts. This determines the protocol and sometimes the model variant.
3. Format and surface area. 0.11 cm² HTS to 4 cm² insert is a factor of 36 in surface area and a different price point at every step.
4. Kit size and number of kits, and what a partial or larger order does to the unit price.
5. Media variant — standard, phenol red-free, antibiotic-free, anti-fungal-free or hydrocortisone-free. Agents are removed at least three days before shipment, so this affects lead time as well as price.
6. Maintenance medium and quantity. A two-week EpiDerm protocol at 5.0 mL every other day is a meaningful consumable line, and it is not in the tissue price.
7. Delivery date and day, against the supplier’s fixed weekly cadence.
8. Destination, because international transit eats into a four-day shelf life.
9. Donor, if it matters — neonatal foreskin is the default; adult breast skin keratinocytes are available from at least one supplier.
10. Academic, non-profit or commercial, stated up front.
11. Committed annual volume, which is the main lever a buyer has on a quote-only product.
12. Whether you need the lot’s ET-50 result, and when it will be available relative to your experiment.

The honest decision rule

If the endpoint is regulatory — a REACH dossier, a cosmetics safety assessment, a medical device submission — buy one of the six models named in the guideline, and choose between them on price, format, availability and shipping geography rather than on validation, because all six carry the same regulatory standing.

If the endpoint is exploratory — formulation screening, mechanism work, method development — the validation premium buys you nothing, and a model with a published price and a longer culture window may serve the experiment better. Phenion’s FT LONG-LIFE at €190 with 50 days of culture is a genuine alternative for anyone whose question is about long-term exposure rather than acute irritation classification.

If your SOP names EpiSkin, start the transfer now rather than at the next study.

Where to go next

The price and spec comparison covers what is priced publicly and what is not in more detail. The head-to-head model comparison covers barrier function, ET-50 and the validation literature. The 3D tissue model supplier comparison puts RhE in the context of the wider living-tissue market, and engineered tissue explains why epithelial tissue ships as a finished product when contractile tissue does not.

How we can help

RhE sourcing fails on calendars, not on price. Weekly production, a four-day shelf life including transit and a fixed shipping day mean the purchase decision and the experiment date are the same decision, and post-shipment lot release means a failure costs you a week whoever pays for it.

We put your specification to the suppliers that hold the validation you need, chase the format and media-variant details that make quotes comparable, consolidate tissue, maintenance medium and shipping onto one order so nothing arrives without its medium, and confirm transit time to your actual address against the shelf life before anyone commits. If a published-price model would serve your endpoint, we will tell you that instead of running a quote cycle.