British Association of Dermatologists and British Photodermatology Group Guidelines for the Safe and Effective Use of Psoralen–Ultraviolet A Therapy 2015

T.C. Ling; T.H. Clayton; J. Crawley; L.S. Exton; V. Goulden; S. Ibbotson; K. McKenna; M.F. Mohd Mustapa; L.E. Rhodes; R. Sarkany; R.S. Dawe


The British Journal of Dermatology. 2016;174(1):24-55. 

In This Article

Adverse Reactions to Psoralen–Ultraviolet A Therapy


Incidence of Acute Adverse Events. Both systemic and topical PUVA are generally well tolerated and the acute adverse effects of PUVA are reasonably well documented. The most common adverse effects are erythema and pruritus, and nausea associated with oral PUVA.[135,182] In one review the rates of acute adverse effects for systemic, bath, and hand and foot PUVA were 1·3% (n = 299 treatments), 1·3% (n = 1675 treatments) and 0·8% (n = 836 treatments), respectively.[183] Importantly, only one severe adverse event (SAE; severe nausea, vomiting and collapse) with oral PUVA was documented in that study (0·3% of treatments), no SAEs were recorded for bath or localized PUVA, and none was attributed to operator error. However, in one review of severe PUVA burns, all four patients had received an accidental overdose of UVA.[184]

In early studies of oral PUVA, reported acute adverse events were uncommon, temporary and generally mild. Melski et al. analysed the adverse events from 41 000 oral PUVA treatments given to 1308 patients.[185] The incidence of adverse events was as follows: erythema, 9·8%; pruritus, 14·0%; nausea, 3·2%; headaches, 2·0%; dizziness, 1·5%. Erythema was more likely to occur in lighter skin phototypes and on a three-times-weekly regimen, and resulted in missed treatments in only 1·1%. Pruritus settled, with emollients and antihistamines, in approximately two-thirds of cases. Nausea settled with antiemetics in one-third of cases. The adverse events rarely led to missed treatments. No clinically significant changes in laboratory screening or changes on ophthalmological examination were observed in that study. In later large series (n = 3175), the incidence of adverse events reported were higher: erythema, 32·4%; pruritus, 25·6%; nausea, 13·5%; headaches, 2·0%; Köbner reaction in psoriasis, 2·0%.[186,187] PUVA treatment was interrupted owing to erythema in 6·8%, and only uncommonly discontinued. Nausea with 8-MOP, but not with 5-MOP, is a relatively common occurrence and rarely requires discontinuation of treatment.[188]

Erythema: A report in healthy volunteers demonstrated that oral 8-MOP PUVA erythema was evident at 24 h and did not peak until 96 h or later in 75% of participants, based on MPD.[189] Studies of the time course of PUVA erythema following topical delivery of 8-MOP also confirmed the delayed erythemal response, with a broad erythemal peak between 96 and 144 h.[190–192] Oral 5-MOP PUVA erythema shows a similar time course to that of 8-MOP.[193] Thus, as erythema following both topical and systemic 8-MOP PUVA peaks at 96 h or later in the majority of patients, this supports the use of 96-h MPD assessment times and treatment intervals of no less than 2 or 3 days. In practice, many centres use an MPD assessment time of 72 h, and no studies have been undertaken to evaluate whether there are differences in efficacy or adverse effects between treatment regimens based on a 72-h vs. a 96-h MPD assessment. Until such a study has been undertaken, MPD assessment at either 72 or 96 h is appropriate. Most U.K. centres use a twice-weekly regimen for both oral and topical PUVA.

The UVA dose should be delivered within 10 min of an 8-MOP bath as photosensitivity reduces quickly, although there may be a more prolonged effect over 40 min if palmar and plantar skin are being treated.[136,194] The temperature of the bath for PUVA delivery is also important, with photosensitization being optimal at 37 °C and reduced at lower temperatures.[195]

Unusual patterns of erythema and burning have been reported, particularly with TMP bath PUVA, owing to the powder formulation of TMP and the variable distribution within the bath water as it forms a microcrystallized suspension.[196] Frank blisters can also occur either owing to UVA overexposure or to psoralen dose, for example as seen if patients stand in a particular way in the cabinet.[197] Particular care is needed in patients with vitiligo;[198] in one atopic patient with vitiligo, accidental PUVA burning resulted in subsequent development of nodular prurigo at affected sites.[199]

Thus, depending on the methods for reporting erythema, the incidence of PUVA erythema during treatment is between 10% and 32%, requiring treatment interruption in between 1% and 7% of cases. Patients of lighter skin phototype and those treated with more frequent treatment regimens are at greatest risk of developing erythema during treatment. Early recognition of cumulative erythema is important and omission of treatment until this has resolved and adjustment of dose increments may be required. Topical emollients and steroids may offer symptomatic relief.

Pruritus: There seems to be a similar incidence of pruritus following oral and bath PUVA, occurring in 10–40% of patients.[135,185,187] The incidence of pruritus is reported to be lower with 5-MOP (43·6%) than with 8-MOP (71·8%).[196] There do not appear to be specific predictors for the likelihood of this adverse effect occurring.

Nausea: Nausea is relatively common, occurring in about 3–13% of patients.[185,187,200] This can be helped by taking the psoralen with a light meal or by using an antiemetic. If nausea is significant then 5-MOP or topical psoralen can be used.[200] In one study the incidence of nausea was reported to be much lower with 5-MOP (7·7%) than with 8-MOP (51%).[196]

Induction of Photodermatoses: Precipitation or aggravation of photodermatoses that are associated with abnormal UVA photosensitivity can occur with PUVA, with reports of CAD and PLE exacerbation having been observed.[81,201] The rate of inducing PLE was estimated to be 50% compared with 62% with NB-UVB in one study.[81] Theoretically, exacerbation or induction of lupus can occur, and there are case reports of the coincidence of PUVA with the development of lupus.[202–204]

Drug Phototoxicity: In general, drug phototoxicity is not a major problem because of the overwhelming effect of psoralen photosensitization, and as long as MPD assessment is performed on the patient's drug regimen, this can be largely avoided.[200] However, pseudoporphyria can occur, as with some of the other phototoxic drugs such as the fluoroquinolones, nonsteroidals, tetracyclines and diuretics, and is more likely to occur after minor trauma and on acral sites. Furthermore, enhancement of PUVA phototoxicity has been reported after ingestion of celery and vegetable broths during PUVA therapy and after taking Rutaceae extracts and celery soup.[205,206] Reported phototoxicity rates seem to be similar for oral and bath 8-MOP PUVA,[135] although clinical experience suggests that severe phototoxic reactions are more likely with bath PUVA. In general, avoiding psoralen-containing plants for at least 2 h before treatment, is advised.

Systemic psoralen can interfere with liver metabolism, which results in many potential drug interactions. Psoralens cause liver enzyme inhibition, which, in turn, causes a potential serum increase of the following: warfarin, anticholinergics, antipsychotics, nonsteroidal anti-inflammatories, theophylline and caffeine. More frequent monitoring is advised if a patient is on warfarin. The effects of caffeine are augmented by PUVA,[207–209] and can occasionally cause headaches and 'jitteriness' on treatment days. Advice to reduce caffeinated beverage intake on these days may be required. Systemic psoralen can also cause liver enzyme induction, which may reduce serum concentration of ciclosporin, chemotherapy agents, azole antifungals, macrolides, tricyclics, antidepressants and antipsychotics, benzodiazepines, statins, calcium channel blockers, protease inhibitors and oral contraceptives. Co-administration of potent liver enzyme inducers such as phenytoin and carbamazepine may increase metabolism of psoralen and result in a reduced response to PUVA. The risk of potential drug interactions further underscores the importance of MPD testing to ensure patient safety and also an adequate amount of psoralen in the skin at the correct time to allow the wanted phototoxic responses.

Hepatotoxicity: There are case reports of hepatitis following 5-MOP and 8-MOP.[210–215] Importantly, two of the four cases arose in patients who had methotrexate (MTX)-induced liver damage, and the remaining cases resolved once the drug was stopped. In one study of 162 patients receiving PUVA, only three had transient elevations of transaminases, which all reverted to normal after treatment was stopped.[216] The rarity of these cases, that have occurred with bath, as well as oral, PUVA, support the view that routine monitoring of liver function tests is not required. The mechanism is unclear and may be idiosyncratic.[217]

Pain: Severe skin pain with PUVA is uncommon and was seen in 4% (eight of 210) of patients in one series treated with oral PUVA.[218] It is characterized by persistent severe prickling, burning or dysaesthesia, which can last from minutes to hours and persist for weeks or months. The risk of developing PUVA pain is unpredictable and does not appear to be related to the patient's skin phototype, PUVA sensitivity or induction of PUVA phototoxicity. An underlying neurogenic mechanism appears most likely.[219–221] Treatment with analgesics, topical anaesthetics and topical or systemic steroids usually has minimal effect. Capsaicin, gabapentin, phenytoin and low-frequency electrotherapy can potentially be of benefit.[220–224] Once PUVA pain has developed it is then a relative contraindication to further PUVA treatment, as pain is likely to recur.

Miscellaneous: Allergic and photocontact allergic dermatitis to psoralens have been reported to affect 0·8% (three of 371) of patients treated with topical PUVA.[225] Type I anaphylaxis to both 5-MOP and 8-MOP has also been observed in isolated cases,[226,227] and asthma may be aggravated.[228–231]

Hyperpigmentation is usually a result of repeated PUVA treatments but may occur with the concurrent use of topical vitamin D analogues at treated lesional sites,[232,233] and with concomitant treatment with isotretinoin.[234]

Triggering of herpes simplex virus can occur with PUVA and use of high-factor sunscreen on the lips and prophylactic aciclovir for susceptible subjects is recommended. Eczema herpeticum has been reported in a patient with AD treated with PUVA.[235] Folliculitis can occur and can be managed with applying emollients following the downward direction of the hair.

Onycholysis following ingestion of phototoxic drugs is well documented. Onycholysis following ingestion of psoralens and natural sunlight has been observed and has also been reported following PUVA photochemotherapy.[236] Other associated adverse events include lunular changes, subungual haemorrhage and nail pigmentation.[187,237]

Clinically significant consistent changes in laboratory parameters are rarely observed. However, severe hyperlipidaemia following PUVA treatment for acute skin graft-versus-host disease has been reported.[238]

Other rare acute side-effects of PUVA include bullous pemphigoid,[239–241] lichen planus,[242] lichen planus pemphigoides,[243] polymyositis,[244] influenza-like symptoms,[245] lymphomatoid papulosis in a patient with MF,[246] and neurological disorders such as insomnia, nervousness, headache, migraine, dyssomnia, depression and dizziness.[247,248] The development of hypertrichosis, acne-like eruptions, milia and seborrhoeic-like facial dermatitis can also uncommonly occur, and these changes are usually reversible on stopping treatment.[187,249] An intertriginous, asymptomatic, self-resolving, maculopapular rash was reported in 8% of participants undergoing 5-MOP PUVA and this resolved spontaneously, despite continuation of PUVA.[188] In one report, three patients with MF developed new lesions at sites previously considered to be clear, early in the PUVA course. This might have been due to unmasking of subclinical lesions due to inflammation.[250]

Key Points. Relatively common acute adverse effects of PUVA include erythema, pruritus, nausea and polymorphic light eruption.

Uncommon but important acute adverse effects of PUVA include PUVA pain, idiosyncratic hepatitis and psoralen allergy.

Long Term

Psoralen–Ultraviolet A Therapy-induced Skin Cancer. PUVA photochemotherapy is mutagenic, carcinogenic and immunosuppressive.[251–260] Skin cancer is a well-recognized side-effect of PUVA and it is well established that the risk of squamous cell carcinoma (SCC) increases in a dose-dependent manner.[261–266] Additional risk factors include exposure to co-carcinogenic therapeutic agents such as UVB therapy, MTX, ciclosporin and X-ray radiation/arsenic.[267–274] The human papilloma virus has also been suggested as a co-carcinogen and has been detected in PUVA-associated nonmelanoma skin cancers (NMSC), especially types associated with epidermodysplasia verruciformis, particularly types 5, 14 and 20.[275–278] More recently, the U.S.A. 16-centre, follow-up study has reported an increased incidence of melanoma in patients treated with PUVA,[279,280] but it remains uncertain that this was a causative association and this association has not been reported by other groups.[265,273,281,282] There is no evidence of either increased incidence of NMSC or melanoma in patients with psoriasis treated by topical PUVA.[283,284] However, the number of patients with high PUVA exposure was low in these studies and many were treated with bath TMP PUVA. It seems likely that any differential risk of carcinogenicity relates more to differences in psoralens than different routes of administration, so it seems reasonable, on the basis of current knowledge, to believe that the risks per phototoxic exposure should be the same for oral 8-MOP PUVA as for bath 8-MOP PUVA.[283] This is a reason why the (mainly historical) use of cumulative UVA dose with PUVA when considering action limits is less useful than considering number of treatments as cumulative UVA dose will always tend to be less with topical PUVA even when the number of phototoxic exposures is the same.

Psoralen–Ultraviolet A Therapy and Nonmelanoma Skin Cancer: The two studies of most value in assessing NMSC risks in humans treated with PUVA were a North American study and a large Swedish study.[263,264,281] The North American study involved the use of maintenance PUVA, and only for one indication (psoriasis) in a fashion rarely used in Europe; the Swedish study used PUVA with a methodology close to that which is used in Europe and for a wide variety of indications. The North American study of necessity (in the absence of good cancer registry data) involved regular follow-up of the patients treated with PUVA without similar follow-up of a control group, whereas the Swedish study relied on cancer registry linkage to identify skin cancers both in patients and in the control population, so comparing better like with like. However, it is still probable that the patients (by definition, having dermatological diseases) were more likely to be diagnosed (ascertainment bias) and diagnosed earlier (lead-time bias) with skin cancers. The North American study demonstrated that patients receiving > 200 treatments have about 30 times the risk of developing a new NMSC per year than the general population.[263] An almost identical risk was found in the highest-risk group (men who had received ≥ 200 PUVA treatments) in the Swedish study.[264] When reporting, RR are typically quoted: it is important to also consider absolute risks. The RR among the highest-risk patients in the Swedish study of 30·7 meant 10 of 180 patients.[264] A meta-analysis reported that the incidence of SCC among patients exposed to high-dose PUVA was 14-fold higher than among patients with low-dose exposure.[285] The risk of skin cancer is persistent, despite discontinuation of PUVA.[286] Other risk factors for patients developing NMSC include having a lighter skin phototype, having skin tumours prior to PUVA and age at starting PUVA, the presence of PUVA keratoses and lentigines.[266,287–289]

PUVA-associated NMSCs show a reversal of the normal ratio of SCC to basal cell carcinoma (BCC) with a marked increase in SCC and these occur in relatively nonsun-exposed sites.[269] Genital SCC in males treated with PUVA has been reported to be substantially increased (95·7 times that of the general population);[290,291] however, a lower incidence was reported in the Swedish study.[264] A retrospective study from France of a cancer registry over 20 years identified only one case, out of 48 cases, of genital SCC in someone with a history of intensive PUVA. In addition, there were no reported cases of genital SCC from 5400 patients who had received PUVA, despite a lack of genital protection during PUVA exposures.[292] Despite reports that SCC occurring in patients receiving PUVA are usually well differentiated and nonaggressive,[263,265,274,287] these tumours can rarely metastasize.[269]

Lim and Stern found high levels of UVB exposure (at least 300 treatments) to be associated with a modest but significant increase in SCC (1·37 times) and BCC (1·45 times) in the U.S.A. PUVA cohort.[267] A high level of exposure to MTX (total of ≥ 4 years of use) has also been found to be a significant independent risk factor in the latter cohort, with an RR of 2·1 for high vs. low, or no exposure.[269] Previously, a synergistic carcinogenic potential between MTX and PUVA had been suggested.[268] Other data from the U.S.A. cohort showed that the risk of SCC in users of ciclosporin was about three times higher than in patients who never used ciclosporin.[270] Patients receiving PUVA who have previously received arsenic or X-ray therapy have an increased incidence of skin cancer than those who have not had such therapy.[271,273,274]

Melanoma: Patients who receive high doses of PUVA often develop lentigines that can show cytologically atypical melanocytes.[293] A fivefold increase in the annual incidence of malignant melanoma in the 16-centre PUVA cohort study, using PUVA to treat psoriasis, was initially reported in 1997.[279] Further follow-up showed a ninefold increase.[280] Melanoma developed after 15 years and was more common in patients of skin types I/II and who had received at least 250 exposures. There have been a number of case reports of melanoma occurring following PUVA,[294–302] but an increased incidence of melanoma has not been observed in European follow-up studies.[265,273,282,303] So, while there is no doubt that an association between PUVA and melanoma was detected in this one study it is uncertain that PUVA causes melanoma. The association could have been because PUVA, as used in that study, did cause melanoma but, equally, the association could have been because of confounding (e.g. patients with psoriasis responsive to PUVA seeking sunlight exposure) or have been due to ascertainment bias (melanomas, and many were early melanomas, some of which might not have progressed, more likely to be diagnosed in a prospectively followed group than in the general population).[304] When it was first reported, the potential risk of melanoma induced by PUVA raised questions as to whether PUVA should be contraindicated in those with a history of > 200 PUVA treatments or those with a personal or family history of melanoma.[303,305]

Noncutaneous Cancer Risk: As PUVA alters immune function, especially that of lymphocytes, there has been continued concern for the potential development of cancer, particularly lymphoma, with long-term use of this therapy. This is particularly relevant in patients with psoriasis, as an association with internal cancer has been found in large cohort studies,[306,307] which includes oral tumours and tumours of the pharynx, lung, liver, pancreas, liver and bladder, and lymphoma.[308–310] However, previous studies have not demonstrated a consistent increase in cancer risk in the central nervous system, thyroid gland, respiratory system, pancreas or kidney.[264,281,311]

Although there are several case reports of leukaemia developing during PUVA therapy,[312–315] Stern and Lange and Stern et al. reported no increase in lymphoma or leukaemia in a PUVA cohort, in either 1988 or 1997;[263,279] however, more recent analysis has found the incidence of lymphoma to be increased in this group in those exposed to high levels of MTX (RR 4·39), that is, 36 months of treatment or longer, but not in those exposed to low levels.[316] A study from Finland showed a 2·2-fold increase in non-Hodgkin lymphoma in patients hospitalized for psoriasis.[309]

Noncancerous Chronic Cutaneous Side-effects of Psoralen–Ultraviolet A Therapy: Keratoses are typically localized to nonsun-exposed skin; risk factors for their development include increased age, male sex and high cumulative UVA dose.[288] They are reported to occur in 46% of patients receiving high-dose (> 2000 J cm−2) PUVA and show cytological atypia.[266] Disseminated superficial actinic porokeratosis has been reported in association with both oral and topical PUVA.[317–320]

PUVA-induced lentigines are distinct irregular, stellate, darkly pigmented macular lesions and may show cytological atypia.[289,293] Incidence varies from 10% to 53%,[289,321–324] occurring most frequently in those patients exposed to high doses of PUVA and those of skin type I/II.[266,289,321,323]

PUVA-induced mottling consists of hyper- and hypopigmentation associated with atrophy that is mainly localized to areas of overdose.[187,325] Hyperpigmentation has been reported at the lesional sites of patients where calcipotriol ointment has been applied during bath PUVA.[233] Transient hyperpigmentation has also been reported around psoriatic plaques in patients treated with calcipotriol ointment and PUVA.[234] Acquired dermal melanocytosis typically affecting the back of Japanese patients treated with PUVA has been reported.[326,327]

PUVA can result in dose-related premature ageing of the skin, manifesting as wrinkling, xerosis, loss of elasticity, freckling, telangiectasia, mottled pigmentation, yellowing of the skin and comedones. More marked changes are seen in those of skin type I/II. Both epidermal and dermal changes have been reported with chronic PUVA exposure, including skin atrophy, focal epidermal and elastosis.[328,329]

Free 8-MOP can be detected in human lenses for at least 12 h after 8-MOP dosing.[330] Previous guidelines from the BPG recommend that protective eyewear be worn for 12–24 h after PUVA and for 24 h in high-risk individuals, for example patients with atopic eczema, children or those with pre-existing cataracts.[331] Previous clinical studies (involving patients advised on eye protection), including the 16-centre prospective cohort study of American patients with psoriasis,[332–334] have not detected an association between the use of PUVA and cataract development. A follow-up study of 82 patients who declined to wear UVA-blocking sunglasses after PUVA treatments over 2–4 years found no evidence of the development of cataracts.[335]

Key Points. PUVA is associated with a significant dose-dependent risk of SCC, and PUVA keratoses, pigmentary changes and photoageing.

PUVA has not been consistently shown to be associated with melanoma, internal malignancy or cataracts.

How Do We Identify Those Who Are Susceptible to the Side-effects of Psoralen–Ultraviolet A Therapy?

To prevent side-effects being encountered, particularly skin cancer, careful patient selection is essential. PUVA should be used with care (generally only if the alternatives are treatments carrying greater risks, such as systemic immunosuppressives) in children, owing to the risk of causing UV cutaneous damage at a young age.

There is concern that PUVA therapy may lead to worsening of HIV status or may increase risk of skin cancer in people with HIV,[336–338] particularly as SCC and melanoma have been reported to be more aggressive in HIV.[339,340] The risk of Merkel cell carcinoma of the skin was increased by 100 times in the U.S.A. PUVA cohort,[341] and is also increased in patients with HIV.[342] However, studies to date of PUVA use in patients with HIV have not demonstrated a deterioration in HIV status or immune function,[343–345] and it has been suggested that, in fact, PUVA might be preferable to UVB therapy in patients infected with HIV.[346]

Patients who are at risk of ocular toxicity are those with pre-existing cataracts, patients who have AD or who are aphakic.

Risk Factors Associated With Psoralen–Ultraviolet A Therapy-induced Skin Cancer. Risk factors include a history of the following: sun-reactive skin phototype I/II; previous history of skin cancer; personal/family history of skin cancer or dysplastic naevus syndrome; previous exposure to ionizing radiation, arsenic or excessive sunbed exposure; history of xeroderma pigmentosum or other genetic disorders associated with skin cancer; immunosuppressive drug therapy; HIV. Look out for the following on examination: multiple freckles/moles; dysplastic naevi/solar keratoses/SCC in situ (Bowen disease); solar elastosis; skin cancer.

How Can Side-effects Be Prevented in Patients Receiving Psoralen–Ultraviolet A Therapy?

Important measures to reduce skin cancer risk include the shielding of high-risk areas on the genitalia and face, education of patients regarding sun-protective measures with sunscreen use and protective clothing, and monitoring for premalignant or malignant skin lesions. Patients who have received > 150–200 PUVA treatments should be offered an annual skin examination to ensure no premalignant or malignant skin lesions have developed,[187,331] and as a component to education of patients and their primary care doctors that they might be at increased risk of skin cancer as a result of medical treatment. Patients should be advised to wear protective UVA-blocking glasses from the time they ingest psoralen until 12 h following PUVA therapy (24 h for high-risk individuals, e.g. patients with atopic eczema, children or those with pre-existing cataracts). Sunglasses or opaque UV protective glasses should be used,[347,348] and these can be tested for suitability.[347,349] Uncoloured glasses are, if there are any uncertainties about their ability to stop UVA transmission, more suitable than tinted glasses as the latter result in dilation of the pupil, potentially allowing more UVA to reach the lens.[350] Protective eyewear should be of sufficient size to reduce peripheral UV exposure, and additional side protection is recommended. Most contact lenses have little or no UVA protection and are not recommended.[351]

UVA-blocking goggles should be worn during PUVA therapy. As studies have not shown an increased incidence of cataracts, it has been suggested that an examination by an ophthalmologist should usually be considered only if the patient is at increased risk of cataracts,[331] that is, patients with pre-existing cataracts, who have atopic eczema or who are aphakic.

As the risk of PUVA-induced skin cancer is related to the cumulative exposure, attempts should be made to reduce this exposure. This can be done in a number of ways, including more efficient dosimetry using less intense PUVA treatment regimens,[352,353] the avoidance of maintenance treatment, following guidelines on action limit the number of treatment exposures (e.g. 150–200 treatments),[331] consideration of periods of breaks from PUVA with rotational therapy,[354] and the use of combination therapy with retinoids or topical vitamin D analogues.[355] Retinoids not only increase the efficacy of PUVA, but also have a skin cancer prophylactic action.[356,357] Concurrent use of PUVA with ciclosporin should be avoided as it can significantly accelerate skin cancer development in patients receiving such treatment.[270] Topical vitamin D analogues and tazarotene have also been used as dose-sparing agents combined with PUVA.[358,359] There may also be a role of potentially safer photosensitizers such as trimethylangelicin.[360] Photochemotherapy using potentially safer wavelengths of radiation, for example NB-UVB, requires further research.[361,362] Agents that may protect against PUVA-induced photochemical damage, for example green tea or Polypodium leucotomos extract, may decrease long-term carcinogenesis.[363,364]

Recommendation (Strength of Recommendation D; Level of Evidence 2+): Concurrent use of PUVA with ciclosporin should be avoided. Post-PUVA ciclosporin is associated with increased risks of NMSC and should be avoided when possible. However, previous use of ciclosporin should not preclude consideration for PUVA.

How Are the Side-effects Managed in Patients Receiving Psoralen–Ultraviolet A Therapy?

Premalignant and malignant skin lesions are treated in the same way as patients not receiving PUVA. PUVA must be discontinued if neoplastic lesions develop and alternative therapy should be considered, but ciclosporin should be avoided.[270] Introduction of acitretin can be helpful in the management of patients with multiple keratoses and skin cancer following PUVA.[356,357,365] These individuals will require careful follow-up as they are at increased risk of further skin cancers developing with time.

Psoralen–Ultraviolet A Therapy and Pregnancy

As PUVA therapy is mutagenic there is concern regarding potential teratogenicity of this treatment. Oral psoralen is associated with reduced birth rate and teratogenicity in animal studies,[366–368] but this is not found in humans.[369–371] One study observed a marked increase in infants with low birth weights when pregnancy occurred after treatment.[370] It was thought that this may be an effect of the underlying disease rather than the treatment itself. Recently, it has been reported that pregnant women with severe psoriasis have a 1·4-fold increased risk of giving birth to infants with low birth weights.[372] It has been suggested that local topical PUVA may be relatively safe in pregnancy as it does not give rise to detectable levels of psoralen.[373] It was recommended that PUVA therapy should be avoided during pregnancy whenever possible as it is mutagenic.[369,374]

There is no evidence that PUVA is a significant teratogen (level of evidence 2++).

Recommendation (Strength of Recommendation D; Level of Evidence 4): It is recommended that women should avoid conception during PUVA therapy and that, if despite this advice, pregnancy does occur, PUVA should be discontinued.