Drug Safety 8 (5): 340-349. 1993 0114-5916/ 93/0005-0340/$05.00/0 Adis International Limited. All rights reserved.
Minimising the Risks of PUV A Treatment
Marinus CG. van Praag,I,2 Larissa N.L. Tseng, I A. Mieke Mommaas,2 Bart W. Boom2 and Bert Jan Vermeer2 I Department of Dermatology, Sint Franciscus Gasthuis, Rotterdam, The Netherlands 2 Department of Dermatology, University Hospital, Leiden, The Netherlands
340 Summary 341 1. Short Term Adverse Effects of PUYA Therapy 342 2. Long Term Risks of PUY A Therapy 343 2.1 Chronic Actinic Damage 343 2.2 Dyskeratotic and Precancerous Skin Tumours 343 2.3 Nonmelanoma Skin Cancer 344 2.4 Immunological Alterations 345 2.5 Ophthalmological Effects 345 3. Minimising the Long Term Risks of PUY A Therapy 347 4. Conclusion
Summary Psoralen photochemotherapy (PUY A) is a combination of orally administered psoralen and long wave ultraviolet-A radiation (UY A), and is one of the most effective forms of therapy for psoriasis. The unwanted effects ofPUYA therapy can be divided into short and long term adverse effects. The short term adverse effects include erythema, pruritus, nausea and headache. While short term adverse effects are limited and reversible after discontinuation of treatment, potential long term adverse effects such as chronic actinic skin damage, dyskeratotic and precancerous skin conditions, non melanoma skin cancer, immunological alterations and cataract formation are of greater concern.
Long term risks associated with PUY A therapy can be minimised by several measures. Careful patient selection is mandatory; for example, patients with chronic actinic damage and a history of skin cancer may bear a higher risk for the development of new cancers, and previous arsenic intake and ionising radiation also increase the risk of non melanoma skin cancers. Certain drug combinations make it possible to lower the UY A dose, which is important because of the dose-dependent increase in the incidence of SQuamous cell carcinomas in patients treated with PUY A. It has been demonstrated that 200 treatments or a total UYA dose of 1200 J/cm2 seems to be the threshold for development of non melanoma skin cancer. Shielding male genitalia during PUY A treatment is essential because of the increased risk of genital squamous cell carcinomas. Yearly dermatological examination to detect skin cancer at an early stage is highly advisable. Sunscreen use, protective clothing and avoidance of sun exposure reduce the uncontrolled dose of solar UY radiation. Other psoralens with a less carcinogenic potential can be used. UY A-opaque sunglasses during the entire period of increased photosensitivity after psoralen ingestion help avoid cataract formation.
Minimising the Risks of PUV A Treatment 341
Assignment to PUV A ought to be based on the risk-benefit ratio for the individual patient and should be limited to those who can be monitored and controlled by informed, competent and conscientious physicians.
Psoralen photochemotherapy (PUV A) is a com-bination of orally administered psoralen and long wave ultraviolet-A radiation (UV A) which brings about a therapeutically beneficial result not pro-duced by either the drug or UV A radiation alone. This particular form of therapy is currently used in the treatment of several common and uncom-mon skin diseases, such as psoriasis, atopic der-matitis, vitiligo, polymorphous light eruption and mycosis fungoides. Psoralens may be applied top-ically, but systemic administration has been proven to be more practical and easier to control, and has a lower incidence of undesirable adverse effects. Since oral PUV A now represents the most widely used form of photochemotherapy, this article will focus on this type of treatment.
Psoralens can be derived from a number of plants, such as lime, lemon, bergamot, parsley, ce-lery, fig and clove. The use of extracts or parts of plants (Ammi majus) containing natural psoralens and subsequent exposure to sunlight as a treatment for vitiligo dates back to ancient Egyptian, Indian and Chinese healers. A study of the efficacy of psoralens in the treatment of vitiligo in Egypt in the 1940s by EI Mofty (1968) and the subsequent work of Parrish et al. (1974) indicated that this group of compounds could be used to affect epi-dermal proliferative diseases, such as psoriasis, by inhibition of cellular DNA synthesis which took place following the interaction between psoralen and UV A in the skin. However, a direct disturb-ance of epidermal cell DNA function is not the only biological effect ofPUVA therapy in psoriasis, since the treatment also affects the immune system in the skin, and may in addition have a systemic ef-fect (Van Praag et al. 1991).
The psoralens used in PUV A therapy are mainly 8-methoxypsoralen (8-MOP), trimethoxypsoralen (TMP), and 5-methoxypsoralen (5-MOP). TMP is a synthetic psoralen, whereas 8-MOP and 5-MOP
are derived from plants. The most widely used de-rivative in PUVA therapy is 8-MOP, which is given orally in a dose of 0.6 mg/kg. It reaches a maxi-mum concentration in the plasma about I to 2 hours after ingestion. In most individuals it is rea-sonable to assume that 2 hours after ingestion is the best time for the skin to be exposed to UV A.
8-MOP is metabolised in the liver where it dem-onstrates a saturable first-pass effect. It has a serum half-life of approximately 1 hour and is rapidly eliminated. The short serum half-life prevents pro-longed photosensitivity. The psoralens are exten-sively metabolised, with only trace amounts being detected in the urine or bile. Urinary and faecal excretion of psoralen metabolites are 74 and 15%, respectively. 80% of the excretion has taken place in 6 to 8 hours and 90% in 12 hours. Following repeated doses of psoralens, there does not appear to be significant accumulation in the body (Gupta & Anderson 1987).
Like any effective treatment, PUV A therapy is associated with certain risks that arise from the toxic properties of 8-MOP, the action of UVA ra-diation and the molecular-biological effects of UV A on 8-MOP. Most of the known risks of PUV A therapy are UV A dose related, so that the greatest frequency of adverse effects is associated with high-exposure doses of the treatments. Unwanted effects of PUV A therapy can be divided into acute (short term) and chronic (long term) adverse effects. There is now considerable clinical experience regarding the short term safety of PUV A, but the risks of potential long term adverse effects are, as yet, not clearly determined. This article primarily focuses on long term adverse effects and discusses several approaches to minimise these risks.
1. Short Term Adverse Effects of PUVA Therapy
Acute adverse effects are related to the patient's inability to tolerate psoralens or overdosing with long wave ultraviolet radiation (UV A). The most
frequent adverse etTect due to overexposure is ery-thema (Parrish et aI. 1974). Depending on the dose, it varies from a pink to a severe violaceous ery-thema with oedema and blistering. Compared with UVB erythema, PUV A-induced erythema appears later, lasts longer and is more intense. It has an onset at 24 hours, peaks at 36 to 72 hours and re-solves over the next week, whereas UVB erythema has an onset 2 to 6 hours, peaks at 12 to 24 hours and resolves over 48 hours (Warin 1978).
There is no satisfactory treatment for PUV A-induced erythema other than cool compresses, emollient lotions and shielding of the atTected area. Topically applied corticosteroids and orally ad-ministered indomethacin may be helpful. Over-dose phenomena usually arise in areas normally not exposed to natural sunlight. Careful observa-tion of the criteria for dosimetry and the guidelines for treatment can minimise these adverse etTects to an acceptable level and do not interfere with therapy in general (W oltT et al. 1977).
Pruritus is also dose related, and usually man-ifests as a mild and transient itching. Intense, sometimes torturing pruritus may occur in severely overdosed skin (WoltT & Honigsmann 1981). Prur-itus can be treated by oral antihistamines or bland emollients. If pruritus persists, the atTected area should be shielded during therapy or exposure times should be reduced. In a few patients, painful sting-ing sensations in exposed skin occur towards the end phase of treatment. This is refractory to any treatment but usually disappears spontaneously after about 2 weeks.
Transient nausea is relatively common with 8-MOP used as a photosensitiser (Parrish et al. 1974), but occurs only rarely with 5-MOP (Tanew et al. 1988). Nausea can be minimised by taking the psoralens with food or milk, decreasing the dose of psoralen, splitting the dose so that half the dose is taken 21/2 hours and the rest 2 hours before re-ceiving UV A radiation, or using antiemetics. Only on rare occasions is nausea so severe that discon-tinuation of PUV A therapy is necessary (Gupta & Anderson 1987).
Very rare short term adverse etTects of PUVA include acne-like eruptions, polymorphous light
Drug Safety 8 (5) 1993
Table I. Long term risks of PUVA therapy
Chronic actinic skin damage (xerosis, wrinkles, freckles)
Dyskeratotic or precancerous skin conditions (actinic
keratosis, PUVA keratosis keratoacanthoma, Bowen's
Nonmelanoma skin cancer (squamous cell and basal cell carcinoma)
Immunological alterations Ophthalmological effects (cataracts)
eruption-like rashes, transient hypertrichosis of the face and nail changes, such as subungual haem-orrhages, onycholysis and pigmentations due to photo toxic reactions in the nail bed (Parrish et al. 1974). These adverse etTects disappear after dis-continuation of treatment.
Single case reports have described aggravation of underlying skin diseases including bullous pem-phigoid, lupus erythematosus, seborrhoeic derma-titis, acne VUlgaris and herpes simplex (Gupta & Anderson 1987).
2. Long Term Risks 0/ PUVA Therapy
While short term adverse effects of PUV A therapy are limited and reversible after discon-tinuation of treatment, long term adverse etTects are of greater concern. From in vitro experiments and animal studies, psoralens and UV A are known to have mutagenic (Schenley & Hsie 1981; Swan-beck & Thyresson 1974) and carcinogenic effects (Griffin 1959; Pathak et al. 1959). It is to be ex-pected that these effects are of relevance to hu-mans. Potential long term adverse etTects ofPUVA therapy include chronic actinic damage, dyskera-totic and precancerous skin tumours, nonmelan-oma skin cancer, immunological alterations and ophthalmological effects (table I). It is assumed that these adverse etTects depend on the total cumula-tive UVA dose received over a prolonged period of time and may be fully identified only after an extended period of latency (WoltT & Honigsmann 1981).
Analysis oflaboratory data in several large scale studies has not revealed significant abnormal find-
Minimising the Risks of PUY A Treatment
ings in patients receiving long term PUY A therapy (Henseler et al. 1981; Melski et al. 1977). Since psoralens can produce liver damage in laboratory animals when given in excessive doses (Wolff 1990), concern has been expressed in the past about pos-sible hepatotoxic effects in humans. However, ser-ial laboratory examinations performed over a period of several years have not revealed any sub-stantial evidence of impairment of hepatic func-tion (Wolff & H6nigsmann 1981). A slight increase in blood urea nitrogen (BUN) and creatinine has been documented in one series (Wolff 1990), but was considered insignificant; there has been no evi-dence suggesting impairment of renal function in large scale studies (Henseler et al. 1981; Melski et al. 1977). An increased incidence of antinuclear antibodies in PUV A-treated patients was reported by Bjellerup et al. (1979), but several other studies have not shown such relationships (Gschnait et al. 1980; Levin et a1.1982; Stern et al. 1979a).
2.1 Chronic Actinic Damage
Repeated phototoxic injury to the skin can be expected to result in cumulative actinic damage re-gardless of whether it is induced by sunlight, arti-ficial UV radiation or PUV A. Although the precise action spectrum of actinic damage has not been determined, epidermal changes are attributed to UVB and dermal changes to UV A because the lat-ter penetrates more deeply into the skin. Long term exposure to PUV A may thus produce changes in the skin that resemble premature aging induced by sunlight (Langner et al. 1977). Early in the course of therapy the skin becomes dry and wrinkled; these changes are fully reversible after cessation of therapy. However, with long term treatment changes become more marked, with the appear-ance of telangiectasia and disturbance of melani-sation in the form of freckling (Swart et al. 1984) and macules of hypopigmentation. This pigmen-tary phenomenon, for which the term 'PUV A-induced mottling' has been coined, is observed in a minority of patients and is mainly, but not in-variably, restricted to areas of overdosage (Gschnait et al. 1980).
Table II. Characteristics of the various skin types
Always burn, never tan II Usually burn, tan less than average (with difficulty) III Sometimes mild burn, tan about average
IV Rarely burn, tan more than average (with ease) V Moderately pigmented skin VI Heavily pigmented
As with sunlight, the development of premature aging of the skin from PUV A treatment is closely related to skin type (table II). Thus, patients with skin types I and II show most marked changes, while little or no change is seen in skin types V and VI. The dose ofPUVA is also important; when PUV A treatment was introduced, very high doses tended to be used to achieve complete clearance, and aging appeared earlier than with the present low dose regimens (Morison 1991).
2.2 Dyskeratotic and Precancerous Skin Tumours
Development of actinic keratoses is the most common precancerous condition that occurs dur-ing PUV A treatment (Stuttgen 1982). PUV A therapy may also be associated with Bowen's dis-ease and keratoacanthomas (Hofmann et a1.1979; H6nigsmann et al. 1980; Sina & Adrian 1983). Re-cently, we introduced the term 'PUV A keratosis', a newly described hyperkeratotic lesion, observed in PUV A-treated patients (Bruynzeel et al. 1991). It appeared that 'PUV A keratoses' could be re-garded as predictive lesions for an increased risk for non melanoma skin cancer, especially squa-mous cell carcinomas (Van Praag et al. 1993).
2.3 Nonmelanoma Skin Cancer
The combination of psoralen and UV A radia-tion is mutagenic in a variety of systems, including mammalian cell lines (Schenley & Hsie 1981; Swan beck & Thyresson 1974). PUV A is also car-cinogenic in mice and causes the development of
squamous cell carcinomas in the skin (Griffin 1959; Pathak et al. 1959). Therefore, it is not surprising that PUV A treatment is associated with non mel-anom...