In September, we were delighted to partner with Personal Care magazine and our testing colleagues at Cutest, to share the results of our 2 year R&D project looking at a new, more accurate in vitro test for skin mildness.

In the webinar broadcast, we explained:

  • Why existing mildness tests are not sensitive enough for today’s gentle cosmetic ingredients
  • How we optimised our in vitro (ET50) method, whilst Cutest optimised their in vivo (human patch test) method
  • The results of studies comparing the in vitro and in vivo test results of some common surfactants and 3 face masks
  • Real world applications of the test including the results of a comparison of some of the UK’s leading soaps and facial cleansers

Listen to the recording here, and for everyone who asked a question that we didn’t get chance to answer at the time, we’ve compiled our responses below. If you have any other questions about the new XtraMild test and its future applications, please don’t hesitate to get in touch.


  1. Are the facial cleansers and soaps tested at 100% in ET50 assay? What is the threshold to classify the product as Mild as per ET50 Score?

The facial cleansers and soaps are tested at 100%. All products are tested neat unless there is a specific requirement by our client to do otherwise. There are a number of classification thresholds built into the prediction model as follows:

  • ET50 <30 minutes: Strong / Severe Irritant
  • ET50 30 minutes to 3 hours 59 minutes: Moderate Irritant
  • ET50 4 hours to 11 hours 59 minutes: Moderate to Mild Irritant
  • ET50 12 hours to 23 hours 59 minutes: Very Mild
  • ET50 >24 hours: Non-Irritant

2. How was the test concentration determined for the in vitro and in vivo testing?

All products were tested neat (100%) to model real-life human exposure for products applied to the skin.

3. In some of the results, I can see % viability that’s higher than 100%. Is there any specific reason?

This is a common observation in a variety of in vitro methods. It could be partially caused by a stimulation of metabolism in the cells by the test item, since the read-out for this test looks at cell health and viability in terms of metabolic conversion of a tetrazolium salt (MTT) to a formazan product by dehydrogenase and reductase enzymes.  (Note that the test method includes a control for any direct interference between the test item and the absorbance reading of the formazan product).

4. Will a higher pH level of soap affect the ET50 value of the product?

Most soaps have a slightly alkaline pH (commonly up to pH10). Reduction of the pH closer to neutral pH7 may be expected to make the soap slightly milder and therefore result in a higher ET50 value. This is something we haven’t specifically investigated to date but further studies could address, as we know that marketing claims around low pH levels are popular.

5. What is the likely impact of the dermis on skin irritation testing as the model you used did not have a dermal component?

For the purpose of this test, a dermal component is not considered necessary as we are looking at the viability of the epidermal keratinocytes as the measured endpoint. While this is a simple endpoint, it has been shown to be very effective in picking up subtle differences between very mild ingredients and formulations, providing a fit-for-purpose test while avoiding unnecessary additional costs for our clients. The use of a full thickness skin model may provide additional value in some cases when looking at cross-talk between epidermal and dermal cell types as a more comprehensive assessment of the biology of irritancy / mildness. This may be the topic of future work if companies express interest.

6. How is eye irritation considered within this in vitro testing?

We have a similar method available using a human reconstructed cornea model, to look at irritation / mildness of ingredients or formulations to the human eye:

7. Is legally acceptable to claim “Dermatologically Tested” by testing a product only with the in vitro method you presented?

In our view, the “Dermatologically Tested” claim should be supported by clinical data (in vivo human patch test studies) and cannot be made using in vitro data as stand-alone. However, the in vitro method can be used to provide valuable supporting information for safety, efficacy and marketing purposes. It can also provide a useful preliminary step before human testing, especially for baby products.

8. What is the major challenge mild products face in terms of performance compared to traditional products?

Mild products are formulated to be gentle to the skin without losing efficacy, so they should not necessarily face any specific challenges. For some personal care products such as shampoos and shower gels, the surfactant system must be optimised for mildness while not compromising on the ability to effectively clean the skin and hair, and to feel good when used – this is a challenge faced by the formulators but new-generation mild surfactants are helping to address the issue.

9. Are you planning to study surfactant-surfactant interaction and effect of micelle size on mildness?

We have done some experiments looking at surfactant blends, and the effect of pH and salt concentration on mildness, as these are factors that can have an impact on micelles. We have not yet shared this latest data but we hope that it will be available soon as a peer-reviewed publication and we may be able to discuss specific information requests. This is a topic that may be researched further with follow-on funding and we are currently looking into sources for this.

10. Have baby wipe products been tested with this method?

We have not tested baby wipes yet, but are keen to do so as part of our ongoing work with the XtraMild test. We have tested a number of other baby products including shampoos, washes, lotions, sunscreens and oils.

11. Can you share which “Bio surfactant” (Chemical class) was included in the study?

We’re unable to share this specific information at the time of writing due to confidentiality agreements with our partners for this research. However, we do hope to publish the work in a peer reviewed journal soon and may be able to reveal more details at that time.

12. Do you expect this method to be ECVAM supported?

We would not submit this particular method to ECVAM because it’s designed primarily for claim support purposes, rather than as a regulatory safety test. Other methods using reconstructed human skin and eye models have gained regulatory acceptance in the past (eg OECD Test Guidelines 439, 392, 431), showing ECVAM’s general approval of this type of approach.

13. Is there a location within the US that can perform this test?

XCellR8 currently serves all of our global client companies, including many in the US, from our laboratory in the UK.