Several pathways eventuate in the development of phototoxic tissue damage, and for many phototoxic agents more than one pathway is responsible.
On absorption of radiation energy by the photosensitizer (P) at its ground state, formation of an excited (usually triplet) state (3P) molecule occurs. The excited state molecule may then participate in oxygen-dependent processes (i.e., photodynamic processes) via two major pathways, type I and type II reactions, both of which result in cytotoxic injury.12
The type I reaction involves transfer of an electron or a hydrogen atom to the excited state photosensitizer (3P), which results in the formation of free radicals [Eq. (92-1)]. These may then participate in an oxidation–reduction reaction that results in peroxide formation and subsequent cell damage [Eqs. (92-2) and (92-3)].
Alternatively, interaction of 3P with ground state oxygen could result in the formation of superoxide anion (O2−.), which, in turn, can be converted into highly reactive and cytotoxic hydroxyl radicals (OH·).
The type II reaction is also known as an energy transfer process. Transfer of energy to ground state oxygen results in the formation of singlet oxygen (1O2), which is highly reactive and has a lifetime of 50 ns [Eq. (92-4)]:
Cytotoxic injury occurs upon singlet oxygen-induced oxidation of amino acids and unsaturated fatty acids; interaction with the latter results in the formation of hydroperoxides, which initiate lipid and protein oxidation.
Phototoxicities induced by porphyrins,12 quinolones,13 nonsteroidal anti-inflammatory agents, tetracyclines, amitriptyline, imipramine, sulfonylureas, hydrochlorothiazide, furosemide, and chlorpromazine14 are examples of photodynamic phototoxic reactions.
Generation of Photoproducts
Exposure to radiation may result in the generation of stable photoproducts that are responsible for tissue injury. Phototoxic products have been demonstrated on irradiation of phenothiazines, chlorpromazine, tetracyclines, quinolones, and nonsteroidal anti-inflammatory agents.15
Another mechanism of phototoxicity is radiation-mediated binding of the photosensitizer to its biologic substrate. A photoaddition reaction occurs when the excited state molecule covalently binds to a ground state molecule. An example is the covalent binding of 8-methoxypsoralen to pyrimidine bases of the DNA molecules, which results in the formation of a cross-link between the DNA strands.
Mediators of inflammation and inflammatory cells participate in phototoxic tissue injury. Biologically active products of complement activation, mast cell-derived mediators, eicosanoids, proteases, and polymorphonuclear leukocytes contribute to the development of phototoxicity induced by porphyrins, demeclocycline, and chlorpromazine.16
Photodynamic therapy (PDT) involves the use of a photosensitizer and electromagnetic radiation in the presence of oxygen to treat premalignant and malignant skin conditions. In addition to generating reactive oxygen species, which results in cytotoxicity, PDT also is a potent inducer of apoptosis.12
Acute phototoxicity occurs within hours of exposure to the phototoxic agent and UV radiation. Symptoms are drug-dose and UV-dose dependent—usually asymptomatic, but at sufficient doses, the patient complains of a burning and stinging sensation on exposed areas, such as forehead, nose, V area of the neck, and dorsa of the hands (Fig. 92-1). Erythema and edema may appear within hours of exposure; in severe cases, vesicles and bullae may develop accompanied by pruritus. Protected areas, such as nasolabial folds, postauricular and submental areas, and areas covered by clothing, are spared. A notable exception to these kinetics is psoralen-induced phototoxicity, in which often the acute response first appears after 24 hours, and peaks at 48–72 hours, which is the rationale for administering psoralen plus UVA (PUVA) photochemotherapy doses 48–72 hours apart. The phototoxic response resolves with a varying degree of hyperpigmentation, which may last for months. At lower drug/UV doses, gradual tanning only, without preceding sunburn-like reaction, can be seen.
Amiodarone-induced phototoxicity. Note the erythema and slate-gray pigmentation (nose, forehead) on the sun-exposed area.
Separation of the distal nail from the nail bed, usually painful, is a manifestation of acute phototoxicity, with the nail plate serving as a lens to focus UV energy on the nail bed. It has been reported with doxycycline and other tetracyclines, fluoroquinolones, psoralens, benoxaprofen, clorazepate dipotassium, olanzapine, aripiprazole, indapamide, and quinine (Fig. 92-2).17
Distal onycholysis in a patient receiving psoralen plus ultraviolet A therapy.
Asymptomatic blue–gray pigmentation on sun-exposed areas has been associated with exposure to several agents.18,19 One percent to ten percent of patients taking amiodarone develop this side effect (Fig. 92-1). Chlorpromazine and clozapine can induce a similar change. The tricyclic antidepressants imipramine and, less commonly, desipramine have also been reported to cause slate-gray pigmentation. A drug metabolite–melanin complex has been postulated to be the cause of this alteration. Minocycline can induce blue–gray pigmentation on the face (Fig. 92-3), frequently on sites of acne scars, although similar pigmentation on forearms and shins can also occur. Chronic exposure to diltiazem, a benzothiazepine calcium channel blocker, has resulted in photodistributed, reticulated, slate-gray pigmentation. Slate-gray pigmentation seen in argyria involves the nail lunulae, mucous membranes, and sclerae. A photochemical reaction, in which silver granules are deposited in the dermis, results in these pigmentary alterations.
Minocycline-induced blue–gray pigmentation on cheeks and upper lip.
Lichenoid eruption has been reported as a form of phototoxicity, but is controversial.
The development of porphyria cutanea tarda-like cutaneous changes of skin fragility, vesicles, and subepidermal blisters is associated with several phototoxic agents (Fig. 92-4). Although histologic and immunofluorescence findings are similar to those of porphyria cutanea tarda, the porphyrin profile is normal or in the upper range of normal in these patients. Naproxen is the most commonly reported causative agent. Other drugs incriminated include amiodarone, β-lactam antibiotics, celecoxib, ciprofloxacin, cyclosporine, diflunisal, etretinate, furosemide, imatinib, nabumetone, nalidixic acid, narrowband UVB, oral contraceptives, oxaprozin, ketoprofen, mefenamic acid, the tetracyclines, tiaprofenic acid, torsemide, and voriconazole.20,21
Pseudoporphyria. Note subtle erosions on dorsum of hand and at the base of the index finger, and crusting on the knuckle.
Accelerated Photo-Induced Changes
This has been uniquely described with voriconazole, a broad spectrum antifungal agent. Immunosuppressed patients receiving voriconazole for >12 weeks can develop photosensitivity, pseudoporphyria, photoaging, lentigines, premature dermatoheliosis; in addition, squamous cell carcinoma and melanoma have been described in this group of patients who were on voriconazole for >12 months.21
Telangiectasia on sun-exposed areas has been reported with calcium channel blockers, including nifedipine, amlodipine, felodipine, and diltiazem, with the antibiotic cefotaxime, and with antidepressant venlafaxine. In some of these patients, provocation with UVA resulted in the development of telangiectasia.22
Persistence of Photosensitivity and Evolution to Chronic Actinic Dermatitis
Although phototoxicity usually resolves after discontinuation of the causative agent, there are reports of persistence of photosensitivity for many years after the cessation of exposure, which results in the development of chronic actinic dermatitis (Fig. 92-5). The condition presents with pruritus and lichenification and excoriation on sun-exposed sites; it has been reported with thiazides, quinidine, quinine, and amiodarone.23
Chronic actinic dermatitis. Note the lichenification and hyperpigmentation on sun-exposed areas, and sparing of skin folds.
Cutaneous effects of long-term, repeated phototoxic tissue injury are best exemplified by the manifestations in patients who have received long-term PUVA photochemotherapy, which is known to affect DNA. These effects include premature aging of the skin, lentigines, squamous cell and basal cell carcinomas, and melanoma. These are discussed in greater detail in Chapter 238.
Table 92-2 lists the major topical phototoxic and photosensitizing agents. It should be noted that fluorouracil and retinoids induce exaggerated UV response due to their irritant effect on the skin. Therapeutic or occupational exposures to these agents are the common route of contact.
Topical exposures to furocoumarins may occur in individuals in certain occupations (bartenders, salad chefs, gardeners) and in patients receiving topical photochemotherapy with psoralens.
Crude coal tar, although no longer commonly used in dermatologic therapy, is well documented to produce a burning and stinging sensation on exposure to UVA (“tar smarts”). In addition to phototoxicity, occupational exposure to tar is associated with increased risk of nonmelanoma skin cancers.
Table 92-3 lists the major systemic phototoxic agents.24–28 They commonly produce an exaggerated sunburn reaction but, like most phototoxins, may also induce an eczematous photoallergic response in a small percentage of users, especially after topical exposure. As a rule, the action spectra are in the UVA range; notable exceptions are the porphyrins, fluorescein, and other dyes, whose action spectra are in the visible light range.
Acute phototoxicity is characterized by individual necrotic keratinocytes and, in severe cases, epidermal necrosis (see Table 92-1). There may be epidermal spongiosis, dermal edema, and a mild infiltrate consisting of neutrophils, lymphocytes, and macrophages. Slate-gray pigmentation is associated with increased dermal melanin and dermal deposits of the drug or its metabolite.18,19 Histologic features of lichenoid eruptions are similar to those of idiopathic lichen planus; however, there may be a greater degree of spongiosis and dermal eosinophilic and plasma cell infiltrates, and a larger number of necrotic keratinocytes and cytoid bodies. In pseudoporphyria, as in porphyria cutanea tarda, there is dermal–epidermal separation at the lamina lucida and deposits of immunoglobulins at the dermal–epidermal junction and surrounding blood vessel walls.20,21
Identification and avoidance of the causative phototoxic agent are the most important steps in management. Beyond this or if the agent cannot be removed, sun avoidance is essential. Because the action spectrum for most agents is in the UVA range, high sun protection factor, broad-spectrum sunscreens containing efficient UVA filters should be used (see Chapter 223). Acute phototoxicity can be managed with topical corticosteroids and compresses; systemic corticosteroids should be reserved for only the most severely affected patients. Management of patients with slate-gray pigmentation, lichenoid eruption, pseudoporphyria, and photodistributed telangiectasia is symptomatic only, and patients should be advised that it will take months after the discontinuation of the offending agent for the condition to resolve. Patients with nonsteroidal anti-inflammatory drug-induced (NSAID-induced) pseudoporphyria who require NSAIDs should be switched to a different class of agents or to those that are less photosensitizing, such as indomethacin or sulindac.29