RASopathic comedone‐like or cystic lesions induced by vemurafenib: a model of skin lesions similar but not identical to those induced by dioxins MADISH. (20th July 2018)

• Record Type: Journal Article
• Authors: Kaya, G.; Saxer‐Sekulic, N.; Kaya, A.; Sorg, O.; Boespflug, A.; Thomas, L.; Saurat, J.H.
• Is Part Of: Journal of the European Academy of Dermatology and Venereology. Volume 32:Number 8(2018)
• Page Start: 1368
• Page End: 1372
• Publication Date: 2018-07-20
• Subjects: Dermatology -- Periodicals; Sexually transmitted diseases -- Periodicals; 616.5
• DOI: 10.1111/jdv.14945
• Languages: English
• ISSNs: 0926-9959

Abstract

Abstract: Background: Patients treated with vemurafenib for metastatic melanoma often develop skin lesions similar to those observed after exposure to dioxin‐like compounds. We previously called these lesions MADISH (metabolizing acquired dioxin‐induced skin hamartoma) when analysing a case of acute dioxin poisoning. Objective: We performed a clinical trial aimed at comparing the skin lesions observed under vemurafenib treatment with MADISH in order to bring to light a possible crosstalk between vemurafenib and dioxin pathways. Methods: In this case series study, we explored the histological aspect of skin lesions in 10 cases treated with vemurafenib for malignant melanoma. We also analysed the ability of vemurafenib and tyrosine kinase inhibitors to induce dioxin–AhR pathway. Results: All patients had skin lesions diagnosed as 'non‐inflammatory acneiform eruption' by dermatologists. These were predominantly facial with notable retroauricular involvement and clinically compatible with chloracne/MADISH when assessed by dioxin expert. Histological analysis showed mostly comedone‐like lesions and dermal cysts containing epithelial wall with basal or lateral epithelial projections and lamellar keratinization and alterations of remaining sebaceous glands. The expression of CYP1A1, a gene highly induced following dioxin exposure, was not observed in these lesions. Vemurafenib and the tyrosine kinase inhibitors erlotinib and gefitinib did not induce CYP1A1 activity. Discussion: Although the skin lesions under vemurafenib treatment were morphologically similar to MADISH, the absence of CYP1A1 expression in dermal cysts of patients and the absence of CYP1A1 activation by vemurafenib led us consider that these skin lesions were different from true MADISH and not mediated by a crosstalk of AhR signalling, but rather to a hyperactivation of PI3K‐Akt pathway as a consequence of vemurafenib treatment. A strong expression of CYP1A1 in the epithelial wall of dermal cysts must be required, parallel to the morphology of the lesions, to make the diagnosis of MADISH, the hallmark of an exposure to dioxin‐like/chloracnegen compounds. Abstract : Linked article: This article is commented on by H.F. Merk, pp. 1233–1234 in this issue. To view this article visithttps://doi.org/10.1111/jdv.15148