Background: Cutaneous photosensitivity is a recognized adverse effect of photosensitizing drugs, arising from the interaction between the drug and ultraviolet radiation (UVR) or visible light. Drug-induced photosensitivity comprises phototoxic and photoallergic reactions, distinguished by their underlying pathophysiological mechanisms. In this context, phototoxicity may occur when photosensitizing medications enhance the skin’s susceptibility to light-induced mechanisms potentially leading to carcinogenesis. The IARC Monographs programme integrates evidence of cancer in humans and cancer in experimental animals with mechanistic evidence, which is evaluated using the “10 key characteristics (KC) of carcinogens” framework. However, “phototoxicity” has not yet been formally considered a KC. In November 2024, three drugs were classified as carcinogenic to humans (Group 1): hydrochlorothiazide, voriconazole and tacrolimus during IARC Monographs Meeting 137. There was sufficient evidence that hydrochlorothiazide and voriconazole cause squamous cell carcinoma (SCC) of the skin in humans, and limited evidence that tacrolimus causes SCC of the skin.
Objectives and methods: We aim to describe how phototoxic properties of hydrochlorothiazide, voriconazole and tacrolimus were evaluated through the KCs framework by reviewing the information from IARC Monographs Volume 137. We analysed specific phototoxic endpoints and experimental test systems and how they were considered in the evaluation of the mechanistic evidence.
Results: Hydrochlorothiazide displayed phototoxicity endpoints related to KC3 “Alters DNA repair or causes genomic instability”, KC5 “Induces oxidative stress”, KC10 “Alters cell proliferation, cell death or nutrient supply”, KC2 “Genotoxicity” and KC6 “Induces chronic inflammation”. Although hydrochlorothiazide is phototoxic, the overall mechanistic evidence was evaluated as limited: few data showed that hydrochlorothiazide could enhance UVR-induced oxidative endpoints (KC5) in human keratinocytes, chronic photodermatitis in exposed humans, UVR-induced inflammatory responses in mice (KC6), and UVR-induced double-strand breaks in human keratinocytes (KC2), as well as increase UVA-induced cellular proliferation (KC10). Voriconazole also displayed phototoxicity endpoints related to KC3, KC5, and KC10. In fact, the major circulating metabolite, voriconazole-N-oxide is highly phototoxic. Notably, consistent evidence was observed for KC5 in human primary cells and in a human living skin-equivalent model exposed to UVR. Strong evidence for KC10 was also identified in exposed humans, based on the disease end-point actinic keratosis, a pathological UVR-associated skin condition and recognized precursor to SCC. Phototoxicity-related mechanisms associated with these KCs contributed to the strong mechanistic evidence for voriconazole. Few data showed that tacrolimus could impair UVB-induced DNA damage repair (KC3) and apoptosis (KC10) in human epidermal keratinocytes and enhance cell proliferation in a mouse tumour model.
Conclusion/implications: Mechanistic evidence from IARC Monographs Volume 137 showed that voriconazole and hydrochlorothiazide induced or enhanced specific KCs-related endpoints, when the experimental test system was co-exposed with UVR, thus confirming the relevance of these mechanisms in the development of photocarcinogenesis. Co-exposure studies that include UVR in different test systems are necessary to identify the carcinogenic hazards of potential phototoxic agents. Refined literature search-terms to identify potential phototoxic effects will be implemented into the KC framework.