Drug photoallergy

Drugs are one of the representative exogenous agents that cause photosensitive dermatitis. Both phototoxic and photoallergic mechanisms exist in photosensitivity to exogenous agents. While the phototoxic reaction is mediated mainly by reactive oxygen species, the photoallergic reaction is induced and elicited by immunological consequences. Two hypotheses have been put forward to explain the formation of photoallergen: prohapten and photohapten. The vast majority of clinically photoallergic drugs are photohapten rather than prohapten. Clinically, photocontact dermatitis and drug photosensitivity are the two major disorders caused by topical and systemic exogenous photosensitizers, respectively. The main cause of photocontact dermatitis is nonsteroidal anti‐inflammatory drugs. In drug photosensitivity, various causative agents have been reported and are recently represented by hydrochlorothiazide, quinolones, piroxicam, and flutamide. Orally administered drugs diffuse from the blood to the epidermis, and keratinocytes are photoderivatized with a given drug upon ultraviolet (UV) A irradiation, leading to photoantigen formation and cytokine production. In parallel, dendritic cells become photohapten‐bearing, T‐cell–sensitizing cells. Considering the mechanisms of photoallergy to chemicals, several in vitro assessments have been proposed to detect the photoallergenicity. Finally, a recent observation with newly marketed drugs has demonstrated that drugs may function as immunomodulators and induce photosensitivity as typically seen in anti‐CCR4 antibody.

immunomodulatory drugs can induce photosensitivity is also mentioned.

| Two types of photosensitivity to chemicals
Photosensitive materials have two properties, phototoxicity and photoallergenicity. The phototoxic reaction eventually results in a cellular cytotoxicity, while the photoallergic reaction is induced and elicited by immunological consequences involving various immunocompetent cells and molecules. [6][7][8] Each photosensitive chemical has different dominancy to phototoxicity or photoallergenicity. For example, psoralen and porphyrin derivatives are strong phototoxic agents with scarce photoallergenicity and thus used for photochemotherapy or photodynamic therapy with few photoallergic adverse effects. 9 By contrast, ketoprofen and fluoroquinolones (FQs) are causative agents for photocontact dermatitis and drug photoallergy, respectively. 7 It is noted, however, that all photoallergic chemicals have a phototoxic property because the photoallergic reaction requires the initial phototoxic step ( Figure 1) in which photosensitizers bind to protein via the formation of reactive oxygen species (ROS). 1 Historically, it was believed that most cases of photocontact dermatitis and drug photoallergy are induced by the phototoxic reaction, and the incidence of the photoallergic reaction is low.
However, recent clinical studies have suggested that the photoallergic type is rather common. 7 This misunderstanding seems to be caused by easy evaluation of phototoxicity and difficult assessment of photoallergenicity.
The action spectrum (the provocative light wavelength) of these two types of photosensitivity is mostly ultraviolet A light (UVA; 320-400 nm). 1 Ultraviolet B light (UVB; 290-320 nm) rarely evokes the diseases, as represented by photosensitivity to drugs, such as sulfanilamide, 10 ranitidine, 11 and bicalutamide. 12 Photoaugmentation by UVA and UVB is occasionally seen in some drugs. 13

| Phototoxicity
Phototoxicity is mainly caused by generation of ROS. 13,14 Singlet oxygen is most important for chemical phototoxicity and the generally termed type II photodynamic reaction. 14 The target molecules of phototoxic chemicals include proteins or amino acids, lipids, and DNA, 13,14 and their alterations lead to cellular damage or even cellular death ( Figure 1). Therefore, cellular cytotoxicity has been used as a classical method to evaluate phototoxicity. Both necrosis and apoptosis occur in cells phototreated with chemicals and UV. 15 Various cells have been utilized for cytotoxicity assessments, including erythrocytes, fibroblasts, keratinocytes, macrophages, lymphocytes, and even fungi, but the reduction in neutral red uptake (NRU) in phototreated fibroblasts (3T3) has been the standard assessment. 16 Phototoxicity can also be evaluated using target molecules, and such tests include protein (histidine, lysine, and cysteine) degradation, lipid oxidation, and plasmid DNA-breaking activity. 13 In addition, the binding capacity of chemicals to protein upon exposure to UV is a phototoxicity test. 17  determinant. Thus, it is now thought that photobinding of agents with protein represents a photoallergic potency of a given chemical (Figure 1).

| Photoallergenicity
Photoallergy is a well-organized immunological reaction. The pathogenesis of contact dermatitis and drug hypersensitivity is based on the hapten hypothesis 18 : A hapten binds covalently to protein, and the resulting conjugate can be recognized as immunogenic determinants. Likewise, photosensitive materials have a haptenic moiety.
Two hypotheses have been put forward to explain the formation of photoallergens ( Figure 2). The initially proposed one is the prohapten, which is converted to a complete hapten by UV irradiation, and exhibits an eczematous eruption consisting of erythema, papules/vesicles, and occasionally bullae, at the skin sites where a photocontactant is applied. 21 The action spectrum of this photosensitivity is mainly UVA. The sensitivity is divided into two, phototoxic and photoallergic, types. 21 Recent attention to phototoxic materials has decreased the incidence of the phototoxic type of photocontact dermatitis. Therefore, the incidence of photoallergy is now thought to be higher than that of phototoxicity.
Various agents have been reported to evoke allergic photocontact dermatitis. Historically, halogenated salicylanilide, such as 3,3 0 4 0 ,5-tetrachlorosalicylanilide (TCSA), and related compounds, which were contained in soaps/detergents and used as topical antimicrobial agents, yielded a large number of patients with photocontact dermatitis. 22 Elimination of these germicides from the market reduced the frequency of the patients. Perfumes, such as musk ambrette 23 and 6-methylcoumarin, and sunscreen agents, especially benzophenone-3 (oxybenzone), 24 had been causative thereafter.
Recently emerging causative agents of photocontact dermatitis are topical nonsteroidal, anti-inflammatory drugs (NSAIDs), such as ketoprofen, 24,25 suprofen, 26 dexketoprofen, 27,28 and piketoprofen. 29 Diclofenac rarely induces photosensitivity. 30 Benzydamine, a nonaspirin-like anti-inflammatory topical agent, provokes photocontact dermatitis on the skin and lips. 31 Sunscreens are still very important photoallergens in cosmetics. 32 In this regard, not only benzophenone and para-aminobenzoic acid (PABA) derivatives, which are now rarely Drug photosensitivity is one of the adverse reactions of systemically administered drugs 35 and is clinically recognized as skin eruptions on sun-exposed areas, including cheeks, nose, forehead, posterior nuchal area, V area of neck, dorsal aspect of hands, extensor surface of forearms, and lower legs. The action spectrum is usually UVA, although UVB may exceptionally induce the sensitivity or augment the level of UVA-induced sensitivity. 13 It should be noted that the absorption spectrum and the action spectrum are same in phototoxicity; however, the action spectrum is shifted from the absorption spectrum to longer wave range in photoallergy.
Drug photosensitivity usually shows erythematous eruption and lichenoid eruption, and occasionally bullous eruption and leukomelanoderma. The erythematous eruption is the common type of drug photoallergy and may have scaling on the surface. The lichenoid eruption is occasionally similar to lichen planus. 36 This type is clinically characterized by erythematous but dark-colored papules and histologically by CD8 + T-cell infiltration in the upper dermis and attacking keratinocytes. 37 Leukomelanoderma displays a unique clinical appearance of a mixture of pigmentation and depigmentation and occurs in dark-colored individuals such as Japanese. In some patients with the erythematous, lichenoid, and bullous eruptions, biopsied specimens exhibit infiltration of eosinophils as well as lymphocytes. 38 Various drugs have been reported to induce photosensitivity, including quinolones as represented by fluoroquinolones (FQs), 7,39,40 afloqualone (AQ), 17 48 and ketoprofen 42,49 is also detected in this model.
Murine allergic photocontact dermatitis to TCSA is genetically controlled and determined mainly by the major histocompatibility complex (MHC). 6 On the one hand, mice with H-2 b,d alleles are high responders, whereas the H-2 k haplotype is closely associated with low responders. 6 On the other hand, in allergic photocontact dermatitis to ketoprofen, H-2 k is associated with high responders and H-2 b,d with low responders. 42  constituents is the initial step of drug photoallergy (Figure 3). In photocontact dermatitis, a chemical is applied to the skin from the outside.
Meanwhile, in drug photosensitivity, a systemically administered drug diffuses to the epidermis from the blood. Protein is covalently bound to a photodegradated site of photohapten to form an allergic photohapten-protein complex. 7 Lysine is a preferential amino acid to allow binding to FQs, 40 but other amino acids possibly afford the binding sites.
UVA is the action spectrum of photoderivatization of protein with photosensitizers. 17 This is in accordance with the historical notion that in the case of photoallergic reaction to exogenous agents, the action spectrum is shifted from its absorption wavelength to a longer wavelength. Thus, even if the absorption spectrum of a given material is UVB, its action spectrum falls in UVA wave range.  In addition, uptake of drugs by keratinocytes and irradiation with UVA would produce ROS, leading to activation of antioxidant response element.

| Elicitation of sensitivity by sensitized T cells
Upon challenge with the same chemical plus UV as induction, skin eruption is elicited by sensitized T cells. An adaptive transfer study using immune T cells showed that transfer of CD4 + T cells induced ketoprofen photosensitivity in na€ ıve recipient mice, but transfer of both CD4 + and CD8 + T cells produced the full-blown sensitivity reaction. 42 Murine photoallergic contact dermatitis to TCSA involves both positive and regulatory immunologic pathways. The suppressive pathway is mediated by IL-10-producing Th2 cells, 8  Guidance on the photosafety testing of medical products was established by the regulatory agencies in the United States and EU in the early 2000s. ICH S10 guidelines on photosafety evaluation reached step 5 of the ICH process in 2014, describing detailed photosafety assessment strategies. However, the current ICH S10 guideline "photosafety evaluation of pharmaceuticals" is intended to derisk the photoirritation of new drug candidates, and the risk management on photoallergy and photogenotoxicity is currently out of scope because of limited best practice.
There have been various in vitro phototoxicity tests. Cytotoxic assays are common and evaluated using fibroblasts (3T3), erythrocytes, Candida albicans, macrophages, lymphocytes, and keratinocytes.
3T3 NRU phototoxicity test was adopted by OECD guideline in 2004.
ROS assay was adopted in the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) S10 guideline (2013). Plasmid DNA-breaking activity was proposed as a sensitive method. 57 The phototoxicity of chemicals also can be assessed by their activities to bind to protein and amino acids, and resultant reduction in certain amino acids. It is notable that the ability to bind to proteins/amino acids also indicates photoallergenicity of chemicals.

| Photoallergenicity assessments
As for photoallergy, several in vitro methods have been proposed, but none of them has yet been accepted for prediction tests (Table 1). In vitro assessments of photoallergic potentials of chemicals reflect one or two steps of the sensitivity. The major TOKURA | 53 assessments include chemical-protein binding, subsequent signal transduction, and outcome of cell function.
Assessments to utilize the initial steps of the sensitivity represent phototoxicity as well as photoallergenicity. Therefore, it is difficult to discriminate photoallergenicity from phototoxicity with these meth- In photo-DPRA, cysteine, lysine, and histidine are representative candidates to afford binding sites to sensitizers under UV irradiation.
Reduction in these amino acids after treatment of proteins with chemical and UVA suggests its photoallergenicity.
Antioxidant response element (ARE) assay is used to test sensitizers or photosensitizers, and it targets guideline. Uptake of a chemical by keratinocytes and irradiation with UVA would produce ROS, leading to activation of ARE. Therefore, photosensitizer plus UVA activates Keap1-Nrf2-ARE pathway in keratinocytes ( Figure 4).

| CONCLUSIONS
The most important issue in drug photoallergy is its diagnosis and identification of causative drugs. Photopatch test, clinically used for the diagnosis, may be false negative, because some drugs are trapped in the stratum corneum. We have therefore attempted to establish in vitro tests to diagnose drug photoallergy and used a modified lymphocyte stimulation test using drug-photomodified cells. 46 This response reflects the proliferative response of T cells to a photohaptenic moiety of chemical. There are a large number of drugs causative for photoallergy, and even newly marketed drugs could evoke photoallergy.
Attention should be paid to new types of photosensitivity, which are represented by the immune-related adverse effect of mogamulizumab. 61 Our recent observation on voriconazole photocarcinogenesis further suggests that a prodrug and its metabolite play different roles in conjunction with UV and construct a photodisordered condition. 62 Finally, recently marketed drugs, such as pirfenidone, 63 show that a long-term phototoxic reaction possibly leads to a photoallergic response.

CONF LICT OF I NTEREST
The author declares no conflict of interest.