Exploration of the GARD® applicability domain – Sensitization assessment of UVCBs

U. I. Torstensdotter Mattson, C. Humfrey, O. Larne, H. Johansson, L. Sweet
SenzaGen, Lund, Sweden, Lubrizol, Derbyshire, United Kingdom, Lubrizol, Ohio, United States of America

The GARD – Genomic Allergen Rapid Detection – platform is a state of the art in vitro assay for assessment of chemical sensitizers. The GARD®skin assay is a powerful tool for assessment of chemical sensitizers, with a predictive accuracy of 94%. In this study, four UVCB test items, provided by Lubrizol and selected based on existing in vivo data (internal Lubrizol data), were evaluated. Sensitizing hazard was assessed using the GARD®skin assay, and the GARD®potency assay further subcategorized the sensitizers into strong (1A) or weak (1B) sensitizers according to GHS/CLP classification. Here we show the importance of using appropriate vehicles in order to predict a correct classification of Test items.


A UVCB Test item with poor water and DMSO solubility was assessed using a mixture of vehicles with different polarity indexes (DMF and Glycerol 1:1). This experimental vehicle mixture classified the UVCB as a skin sensitizer, being consistent with the in vivo data. This case study demonstrates the broadening of the applicability domain of the GARD -assay when assessing UVCBs.

Poster dowload

The use of GARDskin for sensitization evaluation of cosmetic ingredients and ‘real-life’ mixtures

Renato Ivan de Ávila, Tim Lindbergh, Malin Lindstedt and Marize Campos Valadares
Lab. of Education and Research in Pharmacology and Cellular Toxicology, Faculty of Pharmacy, Federal University of Goiás, Goiânia, Goiás, Brazil.
Department of Immunotechnology, Medicon Village, Lund University, Lund, Sweden.


Genomic fingerprints in dendritic cells after chemical exposure is a recent strategy in in vitro techniques for skin sensitization hazard. Within this perspective, Genomic Allergen Rapid Detection (GARDskinTM), an assay based on a support vector machine (SVM) model, was developed for identifying contact allergens using a myeloid cell line as a surrogate for dendritic cells. Predictive system behind the GARDskin™ consists on the transcriptional quantitative analysis of 200 genes, referred as the GARDskin™ prediction signature. Mechanistically, GARDskin™ is linked to key event 3 “Activation of DCs”, as defined by the Adverse Outcome Pathways for skin sensitization published in 2012 by OECD (https://read.oecd-ilibrary.org/environment/the-adverse-outcome-pathway-for-skin-sensitisation-initiated-by-covalent-binding-to-proteins_9789264221444-en#page1)


Information declared on the label and lawsone and PDD levels found in ten commercial henna-based hair coloring cosmetics are show in Table 1. Since all products analyzed were declared as henna cosmetics by the manufactures, the presence of LAW, the main active phytochemical of henna, was then expected in all samples. However, HPLC analysis showed no LAW level in the product nº 2, suggesting falsification. Furthermore, the presence of PPD was declared on the products nº 2 and 8 only by the manufactures. However, this substance was detected in all products, suggesting undisclosed adulteration.

The use of GARDskin for sensitization evaluation of cosmetic ingredients and ‘real-life’ mixtures

The In Vitro assessment of Respiratory Sensitisation Potential of Electronic Cigarette Liquids

Matthew Stevenson, Lukasz Czekala, Liam Simms, Nicole Tschierske, Henrik Johansson, Tanvir Walele
Imperial Tobacco Ltd, Reemtsma Cigarettenfabriken GmbH, an Imperial Brands PLC Company, SenzaGen AB, Fontem Ventures B.V., an Imperial Brands PLC Company,


Introduction and Objectives

There is a general consensus amongst the scientific and public health community that e-cigarettes constitute a less harmful source of nicotine than combustible cigarettes, and that flavours play a critical role in attracting and retaining smokers into the vaping category. Due to the dynamic nature of innovation with e-cigarettes new assays are required to quickly determine the subtle biological response of these products for product stewardship activities. The size of this task is considerable as recent estimates state that more than 8,000 e-liquid flavours are on the market (Hartung, 2016). One particular toxicological endpoint which is of interest for the Stewardship of e-liquids, is Respiratory Sensitisation.
Respiratory sensitization (RS) is an allergic type I hypersensitivity reaction of the upper and lower respiratory tract caused by an immune response triggered by low molecular weight compounds or other environmental proteins. Clinical symptoms of RS include asthmatic attacks, bronchoconstriction and wheezing upon repeated exposure to the same compound. However, respiratory sensitisers are rare, with around 100 well characterised substances described in the literature.
It is Fontem Ventures policy to screen all novel e-liquid ingredients for Respiratory sensitising activities using published literature and in silico techniques. However, there is a need for alternative techniques to fill data gaps and add to a weight of evidence. Several in vitro assays have been described and validated to assess skin sensitisation, however for respiratory sensitization there are no validated predictive assays. It is of note that not all skin sensitizers are also respiratory sensitizers. In 2015, Basketter and Kimber concluded that “…airborne fragrance materials, including skin sensitising fragrance materials, do not pose a risk of the induction or elicitation of allergic reactions consequent upon exposure via the respiratory tract”. Therefore, it is critical that any assays developed to determine the sensitising properties of a chemical can distinguish between dermal and respiratory activity.
The objective of this study was to assess experimental and commercial e-liquids in GARDair™; an assay which claims to detect respiratory sensitisers.
GARDair measures the genomic biomarker signature of a human myeloid leukemia-derived cell line exposed to test substances; making this technology in keeping with the 3Rs (Reduce, Replace and Refine) and Toxicity Testing in the 21st Century principles. Gene expression analysis is performed using Affymetrix microarray technology and a prediction model is used to classify each sample according to its respiratory sensitizing potential.



• From the Benchmark Control data it was estimated that GARDair™ had a sensitivity and specificity of 71% and 100% respectively; with an overall predictive accuracy estimated as 89%.
• Extensive validation of this assay is ongoing, however, the lack of well characterised Chemical Respiratory Sensitisers may limit this.
• None of the experimental or commercial samples were classified as respiratory sensitisers.
• Further exploration of this assay is required, particularly its ability to detect low concentrations of sensitiser in complex mixtures and to ensure that the e-liquid matrix does not interfere with the detection of activity.

The In Vitro assessment of Respiratory Sensitisation Potential of Electronic Cigarette Liquids_British ToxSocCong 2018 Imperial

Integrated hazard identification of chemical sensitizers using in vitro and in silico readouts – A comparative evaluation of predictive performance

By Lhasa

Donna Macmillan, Henrik Johansson, Olivia Larne, Malin Lindstedt
1. Granary Wharf House, 2 Canal Wharf, Leeds, LS11 5PS
2. SenzaGen, Lund, Sweden
3. Lund University, Lund, Sweden



There has been a significant drive to reduce, refine and replace animal models for the prediction of skin sensitization. This is in part due to the implementation of EU regulation 1223/20091 which prohibits the sale and marketing of any cosmetics and cosmetic ingredients which have been tested on animals, alongside REACH2 and CLP3 regulations which state that non-animal methods must be exhausted prior to considering the use of animal tests. The use and availability of non-animal methods is ever-increasing and 3 assays have been validated by the OECD thus far; the in chemico DPRA, the in vitro KeratinoSens™ and the in vitro h-CLAT. A number of other assays are undergoing OECD validation, including the GARDskin assay (Genomic Allergen Rapid Detection), a dendritic cell-based assay which identifies skin sensitizers from 200 genomic biomarkers4. However, it is generally accepted that no single non-animal method can be used as a standalone approach to replace animal models such as the murine local lymph node assay (LLNA). The focus has instead turned to combining multiple in chemico/in vitro/in silico assays and/or molecular descriptors to derive a more accurate assessment of hazard or risk, known as integrated testing strategies (ITS)5. The GARDskin assay has demonstrated high predictivity and has been reported as ready to use in an ITS6, therefore, it was decided to investigate the effect on performance when GARD was used in combination with Derek Nexus – and to compare these results against Derek with the DPRA, KeratinoSens™ and h-CLAT.



Using Derek skin sensitization predictions in combination with in chemico/in vitro assay results has a beneficial effect when predicting the LLNA outcome. GARDskin in particular performs extremely well when used with Derek in a conservative call approach. Human sensitization is more challenging to predict and GARDskin performs less well for this compared to predicting the LLNA – attributed to the small number of chemicals with both GARDskin and human data (n = 57), in addition to the positive bias in the GARD dataset (70%). However, the addition of Derek predictions clearly improve assay performance. Future work will focus on repeating this analysis on a larger, more balanced dataset.

Poster at Lhasa website


Performance of the GARD assay in a blind Cosmetics Europe study

Johansson H, Gradin R, Forreryd A, Agemark M, Zeller K, Malmborg-Hager A1, Johansson A, Larne O, Van Vliet E3, Borrebaeck C, Lindstedt M,
SenzaGen AB, Lund, Sweden, Department of Immunotechnology, Lund University, Lund, Sweden, Cosmetics Europe –The Personal Care Association, Brussels, Belgium


Chemical hypersensitivity is an immunological response to foreign substances. Primarily, these give rise to the clinical symptoms known as allergic contact dermatitis. To mitigate risks associated with consumer products, chemicals are screened for sensitizing effects. Historically, such predictive screenings have been performed using animal models, but industrial and regulatory authorities now demand animal-free methods for the assessment of sensitization. This is a global development spreading across industries and markets. To meet this demand, the Genomic Allergen Rapid Detection (GARD) assay has been developed. Here, we present novel data reconfirming the performance and accuracy of GARD.


The performance of GARD is highly accurate
83% (72 chemicals)
86% (127 chemicals)
83% (72 chemicals)
Correlated GARD predictions and potency classifications