Ability of the GARDskin assay to Predict Skin Sensitization Response in the Guinea Pig Maximization Test
Joint poster with Risk Science Consortium,
Presented at the 2022 SOT
Rose-Marie Jenvert1, Alexandra Zambriczki Lee2, Ronald P Brown2
1SenzaGen, Lund, Sweden, 2Risk Science Consortium, LLC, Arnold, MD USA
Conclusion
- The GARD®skin assay is able to predict skin sensitization potential in humans with a level of accuracy that is equal to or exceeds that of GPMT and the LLNA.
- As a result, the GARDskin assay serves as a promising alternative to assess the skin sensitization potential of medical devices.
Abstract
The preclinical safety assessment of medical devices typically involves an evaluation of the skin sensitization potential of the device. The GARDskin assay is being proposed as an in vitro alternative to the animal-based tests, Local Lymph Node Assay (LLNA) and Guinea Pig Maximization Test (GPMT), that are typically used to assess the skin sensitization potential of medical devices. The ability of the GARDskin assay to replace LLNA for prediction of skin sensitization response has been evaluated (e.g., Johansson et al., 2019) but since GARDskin has also been proposed as an alternative to the GPMT, it is important to compare the concordance of the prediction of the GARDskin assay with the in vivo response obtained in both of the animal-based tests. Based on the results of the GARDskin assay for 122 compounds, this in vitro assay shows a high concordance with the predicted results of the LLNA (87.5%); however, the concordance for results obtained in the GPMT is much lower (71.2%). The concordance of the GARDskin assay and the GPMT is impacted by the relatively high number of false positive results (15 out of 73) compared to the false positives seen in the comparison between GARDskin and LLNA (2 out of 80). The high number of false positives found when comparing the results from GARDskin and the GPMT results from the inaccurate characterization of the human skin sensitization potential of these compounds by the GPMT. Therefore, the low concordance between the GARDskin assay and the GPMT is due largely to inaccurate predictions of human skin sensitization potential by the GPMT and not by shortcomings of the GARDskin assay. Notably, the GARDskin assay (88.7% accuracy) outperforms the GPMT (83.0% accuracy) in the ability to predict the human sensitization response of compounds in this dataset. The results of this project show that the GARDskin assay is able to predict skin sensitization potential with a level of accuracy that is equal to or exceeds that of the currently accepted animal-based tests, suggesting that the GARDskin assay can serve as a promising alternative to the GPMT and the LLNA, and provide a more human relevant result for assessment of the skin sensitization potential of medical devices.
