COMPARISON OF CORROSITEX AND EPIDERM IN VITRO SKIN CORROSION SCREENING ASSAYS TO IN VIVO CORROSIVITY RESULTS.

• TR Number: 435
• Authors: Wilson, D.M., McMullin, T.T. Kan, H.L., Golden, R.M. Stott, W.T. Dow Chemical, Midland, MI, USA.

This study by researchers at Dow Chemical Co. demonstrated that Mattek’s EpiDerm human skin tissue equivalent gave near-perfect concordance with in vivo corrosivity data for a panel of 12 reference chemicals recommended in OECD Guideline 431. In this study, Dow Chemical scientists objectively compared a battery of chemicals in the EpiDerm™ (Mattek Corp., Ashland, MA) and Corrositex (InVitro International, Irvine, CA) in vitro corrosivity screens to existing in vivo data. EpiDerm consists of cultured, normal human epidermal keratinocytes in a differentiated epidermal model with organized basal, spinous and granular layers, and a stratum corneum containing intercellular lamellar lipid layers analogous to those found in vivo, Corrosivity is assessed as relative cytotoxicity at 3-min and 1-hr as the mitochondrial reduction of a tetrazolium salt. Corrositex uses a bio-barrier membrane comprised of a hydrated collagen matrix and supporting filter membrane on top of a vial filled with a chemical indicator dye, with corrosivity assessed according to breakthrough time. A panel of 12 reference chemicals recommended in OECD Guideline 431 for laboratory validation procedures was used [six corrosive: 1,2 diaminopropane; acrylic acid; 2-tert-butylphenol; potassium hydroxide; sulfuric acid; octanoic acid; and six non-corrosive: 4-amino-l,2,4-triazole; eugenol; phenethyl bromide; tetrachloroethylene; isostearic acid; and 4-methylthio-benzaldehyde], as well as 15 proprietary chemicals [nine corrosives and six non-corrosives in vivo, comprised of one alcohol, three organic acids, seven amines, two acrylic acid derivatives, and two halogenated herbicides]. EpiDerm exhibited 96% in vivo concordance (26 of 27), with 1 of 27 giving a borderline false negative for an organic acid. Corrositex exhibited 70% (19 of 27) in vivo concordance; with 3 of 27 false negatives (two organic acids and one halogenated herbicide), 2 of 27 false positives (both were amines), and 3 of 27 that didn’t qualify due to lack of reactivity with the indicator dye. Erroneous results weren’t restricted to false negatives, positives or by chemical class. One organic acid was a false negative for both in vitro assays, with the borderline EpiDerm result suggesting potential corrosivity in vivo. In conclusion, EpiDerm gave near-perfect concordance with in vivo corrosivity data.