Dr Carol Treasure is Co-Founder and Managing Director of animal-free testing laboratory XCellR8.

For a long time, sensitisation was considered the elusive “holy grail” of regulatory in vitro safety testing. Why? Because sensitisation is usually a much more serious effect than, for example, irritation. While irritation tends to be a short-term, reversible, local reaction in the skin, sensitisation follows a more complex pathway involving an immune response. Once a consumer is sensitised to a particular cleaning chemical or cosmetic ingredient, they will typically be sensitised for life – so it’s an extremely important part of safety assessment.

Are we really at a scientific impasse?

By its very nature, the complex pathway of skin sensitisation was considered notoriously difficult to model in vitro: how could such an intricate, multi-step, physiological pathway be accurately represented in a single test? The critics of in vitro testing would cite this as a perfect example of why animal experiments were needed – a whole animal, they said, could never be replaced effectively. This argument became so entrenched in discussions across the chemical industry that, surprisingly, (and as recently as 2017), it was being stated at industry meetings that we are still stuck at this scientific impasse, and that no in vitro skin sensitisation tests were available.

Happily, this is simply not the case.

Three in vitro skin sensitisation tests now available

Thanks to the significant efforts of multiple scientific groups over the past decade, we now have a portfolio of three in vitro skin sensitisation tests with direct relevance to humans. Together, they model the key events in the sensitisation pathway – and they have been fully adopted at a regulatory level in the form of OECD Test Guidelines.

New methods for potency prediction

It’s so important to highlight that this testing strategy is now available as a regulatory option, while being mindful that, like any approach to safety assessment, limitations must be realistically considered.

For example, potency predictions are currently not possible in a regulatory context using in vitro methods, which only address a “yes / no” classification for hazard identification purposes (is the chemical a sensitiser or not?) Efforts are underway to ensure that in vitro methods for potency assessment are available ASAP, and new methods – such as the human genomic screen GARD – are already being validated for potency prediction. Meanwhile, in cases where the in vitro tests have yielded an inconclusive result, or where potency data is specifically required, regulators may insist on the Local Lymph Node Assay (LLNA) as a follow up.

The LLNA (TG429) is a mouse-based test in which sensitisers produce a proliferation of lymphocytes (“T cells”) in the lymph nodes of the ear, draining the site of chemical application. It has been widely used for skin sensitisation assessment and has become the “gold standard” against which new in vitro methods have been validated. However, this approach is flawed in the sense that the LLNA itself is not always an accurate predictor of human skin sensitisation potential.

So, if the new, human-relevant in vitro test states a different result from the traditional mouse-based method – which one is correct? Traditional animal tests were never validated in the same rigorous fashion as today’s in vitro methodologies. Therefore, does using a flawed system as a benchmark give us a skewed picture? As an industry, we need to come up with better validation approaches that compare directly to human data wherever possible.

The three regulatory in vitro tests for skin sensitisation explained


The three tests measure human responses to a chemical at three points in the AOP, described as key events.

Direct Peptide Reactivity Assay (DPRA) (OECD Test Guideline 442c)

Represents Key Event 1 – the “molecular initiating event” of binding to skin proteins

KeratinoSens™ (OECD Test Guideline 442d)

Represents Key Event 2 – activation of skin cells (keratinocytes) in the epidermis

h-CLAT (OECD Test Guideline 442e)

Represents Key Event 3 – activation of the local immune response via dendritic cells in the epidermis

For REACH compliance, these in vitro methods must now be used as the default strategy. Current guidance recommends a “2 out of 3” approach, whereby 2 positive results would result in classification of a chemical as a skin sensitiser.