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

Psoralen–Ultraviolet A for Specific Dermatoses


Is Oral Psoralen–Ultraviolet A Therapy More Effective Than Topical Psoralen–Ultraviolet A Therapy in Patients With Chronic Plaque Psoriasis?. Two randomized parallel group studies compared oral 8-MOP PUVA with 8-MOP bath PUVA.[10,11] One prospective contemporaneous controlled (but not reported to be randomized) study compared oral 8-MOP PUVA with trioxsalen (TMP) bath PUVA,[12] and another compared oral 8-MOP PUVA with 8-MOP bath PUVA.[13] None of these studies detected a definite difference in efficacy between oral and bath PUVA ( Table 2 ; Fig. 1), although one did show 18% more of the small study sample population clearing with bath PUVA than with oral PUVA (Fig. 1).[11] A recently published randomized study comparing bath 8-MOP PUVA with oral 8-MOP PUVA for psoriasis did not detect a difference in efficacy between these modalities.[14] One study report included a questionnaire administered to 13 patients who had received both oral and bath PUVA. There was a roughly equal split among these patients in which form of PUVA they favoured. A recent questionnaire survey of patients referred to a U.K. phototherapy unit found a similar, roughly equal, split between patients who would choose bath PUVA and those who would choose oral PUVA.[15]

Figure 1.

Controlled study (including randomized controlled trials by Collins et al. and Cooper et al.)10,11 comparisons of oral psoralen–ultraviolet A (PUVA) with bath PUVA for psoriasis. CI, confidence interval.

Oral PUVA was not more effective than topical PUVA as a whole-body treatment for psoriasis. The evidence that exists from randomized controlled trials (RCTs) and prospective contemporaneous nonrandomized controlled studies showed that bath PUVA works at least as well as oral PUVA. It is of value to be able to offer bath PUVA as well as oral PUVA.

Recommendation [Strength of Recommendation B; Level of Evidence 1+ (see Appendix)]: All dermatology phototherapy units should offer bath PUVA as well as oral PUVA to treat psoriasis.

Is Psoralen–Ultraviolet A Therapy More Efficacious Than Conventional Oral Systemic Therapies in Patients With Chronic Plaque Psoriasis?. There have been no RCTs comparing conventional oral systemic therapies with PUVA to treat chronic plaque psoriasis. An RCT compared PUVA with placebo and showed a similar magnitude of benefit as has been shown with various conventional systemic therapies when compared with placebo (Figs 2 and 3).[16–19] Owing to differences in study methodologies, a meta-analysis comparing PUVA with conventional systemic therapies would not be appropriate. Nevertheless, the information given in these publications concerning baseline psoriasis severity suggests that PUVA was likely to be at least of similar efficacy in the relatively short term as these systemic therapies. As the risks with PUVA are lower than the risks with systemic therapy,[20,21] PUVA should usually be considered first.

Figure 2.

Randomized controlled trial comparing psoralen–ultraviolet A (PUVA) with placebo for psoriasis. CI, confidence interval.

Figure 3.

Randomized controlled trials comparing conventional systemic therapies with placebo for psoriasis. CI, confidence interval; PASI 75%, 75% reduction in Psoriasis Area and Severity Index.

Recommendation (Strength of Recommendation B; Level of Evidence 1+): PUVA should usually be offered before oral systemic therapy for patients with chronic plaque psoriasis that has not responded adequately to other therapies, including NB-UVB.

Is Psoralen–Ultraviolet A Therapy More Efficacious Than Biologics in Patients With Chronic Plaque Psoriasis?. This has not been assessed in any head-to-head comparative studies. However, a retrospective database comparison showed PUVA in the short term to be more effective than most biologicals in improving psoriasis by reducing the Psoriasis Area and Severity Score (PASI) by 75%, and even more so in reducing the PASI by 90% (Figs 4 and 5).[7]

Figure 4.

Retrospective comparison of psoralen–ultraviolet A (PUVA) therapy (n = 118 patients) with biologicals (Inzinger et al.).7 Baseline mean Psoriasis Area and Severity Index (PASI) score for PUVA was 15 and for biologicals it was 16·9. Outcome of 75% reduction in PASI (PASI 75). CI, confidence interval.

Figure 5.

Retrospective comparison of psoralen–ultraviolet A (PUVA) therapy (n = 118 patients) with biologicals (Inzinger et al.).7 Baseline mean Psoriasis Area and Severity Index (PASI) score for PUVA was 15 and for biologicals it was 16·9. Outcome of 90% reduction in PASI (PASI 90) (similar outcome measure to 'clearance' or 'minimal residual activity'). CI, confidence interval.

Some national guidelines on the use of biological therapy for psoriasis do not clearly indicate if PUVA should be used before biologics. For example, some guidelines have stated that one of the criterions for using biological therapy was 'where phototherapy and alternative standard systemic therapy are contraindicated or cannot be used due to the development of, or risk of developing, clinically important treatment related toxicity', not clarifying whether phototherapy should be taken to include PUVA or not.[22]

A randomized comparative study of adequate duration comparing PUVA with biological therapy would help guidance with prescribing PUVA before biologicals or not, taking into consideration the relative risks (RR) and efficacy. However, on current evidence it seems appropriate to continue to follow the advice in the British National Formulary,[23] where PUVA should be considered before biological therapy.

Recommendation (Strength of Recommendation C; Level of Evidence 2+): PUVA should be considered before biological therapy to treat chronic plaque psoriasis.

Is Psoralen–Ultraviolet A Therapy More Efficacious Than Narrowband Ultraviolet B in Patients With Chronic Plaque Psoriasis?. Ten comparative studies of PUVA with NB-UVB to treat psoriasis have been published (Fig. 6);[24–33] six of these studies were randomized,[26–30,33] with the rest being contemporaneous controlled studies. A study by Dayal et al. is not included in Figure 6 because insufficient data were available in the report, although it was reported that all patients (in both groups) reached 'greater than 75% clearance or complete clearance'.[32] One study found that PUVA was more effective the more severe the psoriasis was at baseline,[25] and another showed that PUVA remained more effective than NB-UVB at 6 months after completion of treatment courses.[26] In most (but not all) comparisons involving NB-UVB three times a week there was either little difference or NB-UVB was more effective, whereas in all but one of the studies comparing NB-UVB twice weekly with PUVA, PUVA was more effective (Fig. 6). A randomized study comparing NB-UVB twice weekly with NB-UVB three times weekly found the psoriasis in 11% more patients in the latter group to be clearing;[34] possibly those studies comparing twice weekly NB-UVB regimens were not comparing optimally effective NB-UVB regimens with PUVA. Only two studies showed NB-UVB to be significantly more effective than PUVA: both used a TMP bath PUVA regimen.[27,29]

Figure 6.

Controlled studies comparing psoralen–ultraviolet A therapy (PUVA) with narrowband ultraviolet B (NB-UVB) for psoriasis. CI, confidence interval; 8-MOP, 8-methyoxypsoralen; TMP, trimethylpsoralen; MRA, minimal residual activity; SPT, skin phototype.

Overall, in the U.K., PUVA appears to be more effective than NB-UVB for psoriasis. It may work better in those with more severe psoriasis, although only one study showed this to be a significant finding. PUVA can work when NB-UVB has not worked, as found in at least one of the paired comparison studies. The experience of all members of the GDG is that failure to respond adequately (either in initial clearance or maintenance of improvement after a course) to NB-UVB does not mean that PUVA will not prove adequate. An assessment of one region's data corroborates this impression: during the 5 years (2005–2009 inclusive), 128 patients in Tayside had a first course of NB-UVB followed by a first course of PUVA. Of these PUVA courses, subsequent to a course of NB-UVB, 62% (70/128) were documented to achieve 'clearance' or 'minimal residual activity', and 56% (29/52) of PUVA courses subsequent to a failed NB-UVB course were documented as achieving 'clearance' or 'minimal residual activity'.

We found no studies directly comparing PUVA with NB-UVB in children.[35] It seems likely that the relative efficacy of these treatments will be similar to that found in adults.

There are greater adverse effect concerns with PUVA than with NB-UVB, with particularly long-term skin cancer risks a concern when treating children for what is frequently a lifelong condition. Also, PUVA is more involved, requiring the taking of tablets or attendance for baths in a hospital unit. Therefore, PUVA is not a first-line phototherapy for adults but even less so for children. However, for individuals who have not adequately responded to NB-UVB it is appropriate to consider it, and for many children it is more appropriate than other options, which may include hospital admissions resulting in disruption of school and home life or systemic therapies with their adverse effect risks.

Recommendation (Strength of Recommendation B; Level of Evidence 1+): Although PUVA may occasionally be appropriate as a first-line phototherapy treatment for especially thick and/or extensive plaque psoriasis it should usually only be considered in patients with chronic plaque psoriasis if NB-UVB has not been adequately effective.


We did not find any controlled studies investigating the use of PUVA in atopic eczema. The findings of several uncontrolled studies have suggested that PUVA is an effective treatment for severe atopic eczema.[36,37] A more recent randomized crossover trial found 5-MOP plus UVA to be superior to medium-dose UVA1 in the treatment of severe atopic eczema.[38] The authors also found that reductions in an eczema severity score were observed after only 10 irradiations.

A systematic review on the use of photo(chemo)therapy in the management of atopic eczema with PUVA therapy found good-quality RCTs to be limited.[39]

Is Psoralen–Ultraviolet A Therapy More Efficacious Than Narrowband Ultraviolet B in Patients With Atopic Eczema?. We found no prospective trials comparing PUVA and NB-UVB. The efficacy of 8-MOP bath PUVA vs. NB-UVB has been studied in a half-side comparison study.[40] The authors of this small study concluded that both regimens were not detectably different in efficacy. In the absence of strong evidence favouring PUVA, NB-UVB should be the first-line phototherapy as a simpler and safer intervention for atopic eczema.

Recommendation (Strength of Recommendation D; Level of Evidence 3): PUVA should be considered in patients with atopic eczema only if NB-UVB has not been adequately effective.

Is Psoralen–Ultraviolet A Therapy More Efficacious Than Conventional Oral Systemic Therapies in Patients With Atopic Eczema?. There were no direct comparative trials comparing systemic agents with PUVA in patients with atopic eczema.

Is Psoralen–Ultraviolet A Therapy More Efficacious Than Narrowband Ultraviolet B in Children (Younger Than 16 Years) With Eczema?. While there were several retrospective studies showing efficacy of PUVA in children with atopic eczema, there were no direct comparative studies comparing PUVA with NB-UVB in children. A large study of 113 Japanese patients with severe atopic eczema treated with oral PUVA, which included 18 patients aged 12–19 years, reported that the majority of patients improved with PUVA,[36] although only 31 participants were scored for severity. The efficacy of bath PUVA therapy in severe cases of atopic dermatitis (AD) in adults has been studied in children over the age of 12 years.[41] This small study of 30 participants reported patient evaluations with an overall patient satisfaction score of 8·8 on a scale of 0–10.

The most convincing evidence for the use of PUVA in children with atopic eczema came from a report by Sheehan et al., who found that 32 of 39 children were able to achieve remission.[42] Similar results have been reported for the use of UVB in children with atopic eczema.[43]

Cutaneous T-cell Lymphoma

PUVA remains a major therapeutic modality in the treatment of cutaneous T-cell lymphoma (CTCL). Its use is in the treatment of the most common form of CTCL, MF, where it remains the major therapy for plaque-stage disease. PUVA phototoxicity has been shown to target selectively neoplastic T lymphocytes in the skin.[44–48]

How Does Psoralen–Ultraviolet A Therapy Compare With Other Types of Phototherapy in Cytotoxic T-cell Lymphoma?. Narrowband Ultraviolet B (TL-01) Compared With Psoralen–Ultraviolet A Therapy: There are no double-blinded, controlled comparison trials of PUVA vs. NB-UVB for the treatment of early-stage CTCL, and most data are from retrospective studies.

In a retrospective analysis of 56 patients with early-stage (IA and IB) disease, 81% (17 of 21) of patients treated with NB-UVB achieved a complete remission compared with 71% (25 of 35) of those treated with PUVA.[49] In a second retrospective study of 40 patients with early-stage disease, PUVA and NB-UVB were found to be equally effective, with NB-UVB resulting in a 50% complete remission rate and a 33% partial response rate compared with PUVA response rates of 64% and 21%, respectively.[50] A third prospective study also showed no statistically significant difference in the degree of response obtained with the use of NB-UVB or PUVA in the treatment of 20 patients with early stage MF.[51] Almost complete remission occurred in 70% of participants with either treatment. Remission durations with NB-UVB were assessed in one study which showed that relapse occurred in a mean of 6 months.[52] NB-UVB appears to be as effective as PUVA in patch stage CTCL; remission durations with NB-UVB and PUVA have not been directly compared.

PUVA has long been the first-line phototherapy for plaque-stage MF. PUVA and NB-UVB may work in MF through direct effects on abnormal lymphocytes. If this is part of how it works, it seems unlikely that NB-UVB should penetrate sufficiently to clear plaque-stage MF, as well as PUVA. This is supported by case series of treatment of patch-stage MF, which found that histologically incomplete deep clearance was associated with poorer response.[53] Also, a recent review of how patients who had been receiving PUVA responded when switched to NB-UVB due to a psoralen shortage confirmed that those with more severe disease (that study did not report explicitly on the differences between patch- and plaque-stage MF) were less likely to improve when switched to NB-UVB.[54]

Other Types of Phototherapy: There are no comparative studies of BB-UVB vs. PUVA in CTCL. As with NB-UVB, the role of BB-UVB in CTCL appears to be in early-stage disease.[55,56]

There are no comparative studies of UVA1 vs. PUVA. There are limited data regarding UVA1 therapy in CTCL, and this suggests that it may potentially be a useful treatment in patch-stage disease,[57] and perhaps in more advanced disease as well.[58]

Recommendation (Strength of Recommendation D; Level of Evidence 3): For patch-stage CTCL, NB-UVB is as effective as PUVA and is the treatment of choice.

Recommendation (Strength of Recommendation D; Level of Evidence 4): For plaque-stage CTCL, PUVA is the treatment of choice.

When Should Psoralen–Ultraviolet A be Used in Cytotoxic T-cell Lymphoma?. PUVA is very effective in clearing lesions of early-stage CTCL, that is, patches and thin plaques. Its effect on infiltrative thick lesions and tumours is more controversial.[59] As patch-stage disease is so responsive to NB-UVB, the role of PUVA is primarily in the treatment of plaque-stage disease. However, one must be aware and use caution in this setting, knowing that there may be an increased risk of other skin tumours if immunosuppressive agents are required later in the disease process (S.J. Whittaker, personal communication). It is not always possible to predict in advance who might require systemic immunosuppressive therapies later and progressing to next-line therapies earlier rather than PUVA is often not appropriate. The issues should be fully discussed with patients.

PUVA was first reported as a treatment for MF in the 1970s. In 1976, Gilchrest et al. reported nine patients with MF at a spectrum of stages from plaque stage to erythroderma who had failed to respond to conventional therapies.[60] Two patients were added to this series and reported on in 1979.[47] A course of PUVA, ranging from 16 to 28 months, induced complete clearing in 10 of 11 patients. Long-term outcome was difficult to evaluate as three patients died within the first year and six further patients were lost to follow-up. Roenigk treated 12 patients with plaque-stage MF, a combination of plaques and tumours, or erythrodermic disease.[44] Complete response was observed in the seven patients with plaque-stage disease, with partial response in the four patients with both plaques and tumours, and a poor response in the erythrodermic patient. The mean number of treatments was just 17.

A 15-year prospective study compared PUVA with topical therapies. Eighty-two patients with MF (68 patients at stage I, seven at stage II, six at stage III and one at stage IV) were studied.[61] It was found that 95% (including all stages of disease) of patients had some level of response to PUVA: 65% went into complete remission, 31% had a partial response and 4% showed no response. The medium time to best response for all stages was 3 months. Patients in early-stage disease responded best: 79% of patients with stage 1A disease achieved complete remission, 59% with stage IB, 83% with stage IIA and 33% with stage III.

A smaller case series has shown clinical improvement in six patients with stage II or III MF after PUVA.[62] Remission after PUVA can be long lasting: a single course of PUVA led to remission for up to 79 months in five of nine stage IA patients, and in 10 of 26 stage IB patients.[63]

There has been one randomized crossover study comparing PUVA with extracorporeal photophoresis (ECP) in stage IB disease in patients who had a detectable peripheral T-cell clone. In 16 patients with stage IB disease, skin disease activity scores fell more following 3 months of PUVA than following 6 months of ECP. Neither treatment achieved clearance of T-cell clones from peripheral blood.[64]

Flexural sites ('sanctuary sites') often fail to respond completely – in both patch and plaque stages – and the duration of response varies.[65]

Trial evidence of the role of PUVA in late-stage disease is more limited but nevertheless suggestive that PUVA, as a monotherapy, is not an effective therapy in late-stage disease. This is thought to be because of the failure of UVA to expose significant reservoirs of malignant cells in the lymph nodes and circulation.[44,55,61]

In studies for PUVA, treatment schedules vary from twice to four times a week with varied increment protocols. In the U.K., commonly used schedules are twice or three times weekly treatments until disease clearance or best partial response.[65] Unlike in most other conditions treated with phototherapy, maintenance therapy to prevent disease relapse in patients with quickly recurrent disease is still common practice.[66] However, there is no consensus on this [and a trial comparing maintenance with no maintenance in PUVA for MF is ongoing ( registration NCT01686594)] and the benefits of maintenance therapy are still uncertain. In the recently published BAD skin lymphoma guidelines, Whittaker et al. concluded that maintenance PUVA therapy should be avoided and the cumulative lifetime PUVA exposure should be limited (1200 J cm−2 of UVA and/or 250 sessions).

Recommendation (Strength of Recommendation B; Level of Evidence 1+): PUVA is the first-line treatment for plaque-stage MF.

Recommendation [Strength of Recommendation D (Good Practice Point); Level of Evidence 4]: Maintenance therapy may be considered to prevent relapse in quickly recurrent disease.

When Should Psoralen–Ultraviolet A be Used With Other Therapies?. In practice, PUVA is often used with other therapies rather than as monotherapy. The choice of therapy will be determined by the disease stage.

Psoralen–Ultraviolet A With Interferon: In early-stage disease the combination of PUVA and interferon (IFN) has been shown to be well tolerated and effective. In a phase II clinical trial, 89 patients with stage IA–IIA disease treated for 14 months with a combination of PUVA and low-dose IFN demonstrated an overall response rate of 98% with complete remission in 84%. The study reported an overall relapse rate of 37% with a median time to relapse of 46 months. Of the relapses, all were local recurrences that further responded to further low-dose IFN therapy.[67]

An RCT comparing PUVA vs. PUVA and IFN-α in early-stage MF found higher remission rates and longer progression-free survival with combination therapy than PUVA monotherapy.[68] In other trials, similar results have been found.[69]

Psoralen–Ultraviolet A With Retinoids and Rexinoids: The retinoids acitretin and isotretinoin, and the rexinoid bexarotene have been shown to be effective in combination with PUVA, and may reduce the total cumulative PUVA dose needed to induce and sustain remission. A nonrandomized trial of patients with plaque-stage MF found that combination therapy of isotretinoin or etretinate with PUVA achieved faster remission, with reduced doses of PUVA required. However, the overall remission rate was not increased.[70]

Although data are limited, small studies suggest that PUVA–bexarotene may be an effective combination.[71,72] A larger, phase III RCT (n = 93) reported no significant difference in response rate or response duration between treating with PUVA alone vs. PUVA and bexarotene in stage IB–IIA.[73] However, there was a trend towards fewer PUVA sessions and a lower dose required to achieve complete clinical response in the combination arm but, owing to insufficient power, this did not achieve statistical significance.

Psoralen–Ultraviolet A Following Total Skin Electron Beam Therapy: A retrospective review found, compared with TSEB alone and with TSEB followed by other forms of adjuvant therapy, that PUVA adjuvant therapy following total skin electron beam therapy (TSEB) increased 5-year survival in patients with early-stage disease.[74] There was a 14% relapse rate at 100 months for patients who had received adjuvant PUVA therapy compared with a 49% relapse rate in those who had not. However, TSEB is not usually appropriate in early disease and that study did not show a difference in survival (as opposed to disease clearance) with TSEB followed by PUVA.

Recommendation (Strength of Recommendation B; Level of Evidence 1+): In the treatment of early-stage CTCL, combination therapy with PUVA and IFN or retinoids/rexinoids should be considered if the response to monotherapy is slow.


Is Psoralen–Ultraviolet A More Efficacious Than Narrowband Ultraviolet B in Patients With Vitiligo?. No; NB-UVB is at least as effective as PUVA in treating vitiligo (Fig. 7).[6,75,76] Also, the match of repigmentation to normal skin colour is better with NB-UVB than with PUVA,[6] and NB-UVB is more effective in inducing repigmentation in unstable vitiligo than PUVA.[76,77] These studies involved widespread vitiligo vulgaris; there is no available good quality evidence comparing PUVA and NB-UVB for other patterns of vitiligo, such as segmental vitiligo.

Figure 7.

Controlled studies comparing psoralen–ultraviolet A (PUVA) with narrowband ultraviolet B (NB-UVB) for vitiligo. TMP, trimethylpsoralen; 8-MOP, 8-methyoxypsoralen; 5-MOP, 5-methyoxypsoralen; CI, confidence interval.

Recommendation (Strength of Recommendation A; Level of Evidence 1+): PUVA should only be considered for widespread vitiligo if NB-UVB has not shown to be adequately effective.


Owing to the relative rarity of some idiopathic photodermatoses and paucity of evidence in this area, the literature search was extended to include all papers from 1966. Both oral and topical PUVA have been reported to be used in the following idiopathic photodermatoses: polymorphic light eruption (PLE), chronic actinic dermatitis (CAD), solar urticaria (SU), erythropoietic protoporphyria (EPP) and actinic prurigo (AP). Most papers were found for PLE (n = 15), followed by SU (n = 8), CAD (n = 7), AP (n = 2) and EPP (n = 2).

In the Treatment of Photodermatoses, What is the Efficacy and Safety of Psoralen–Ultraviolet A Compared With Ultraviolet B?. Polymorphic light eruption: In PLE, the reported efficacy of PUVA was a 65–100% photoprotection rate.[78–89] There were five comparative studies with UVB,[78,79,81,83,84] including two RCTs.[81,83]

The only RCT comparing PUVA with NB-UVB did not detect any significant difference in efficacy. In this trial, 12 patients treated with oral 5-MOP or 8-MOP PUVA were compared with 13 patients treated with UVB and placebo tablets [first dose was 70% minimal phototoxic dose (MPD)/minimal erythema dose (MED); 20% increments; three times weekly UV exposures for 5 weeks]. The study found no significant differences between the two forms of phototherapy in reducing the number of episodes of PLE or restriction of outdoor activity.[81]

Two studies compared the efficacy of PUVA with BB-UVB.[83,84] In an RCT, oral 8-MOP PUVA (n = 13) was compared with BB-UVB plus placebo tablets (n = 13) and low-dose UVA plus placebo tablets (n = 12).[83] Using self-assessment of treatment efficacy, 92% of patients considered PUVA successful compared with 62% with BB-UVB. These findings were supported by Addo and Sharma, who reported complete remission in 89% of patients treated with PUVA compared with 69% treated with BB-UVB.[84]

Man et al. reported a 10-year retrospective review of 170 patients with moderate-to-severe PLE who attended for PUVA and/or UVB phototherapy.[78] Eight patients received PUVA, 128 NB-UVB, five BB-UVB and 29 patients who failed to respond satisfactorily to NB-UVB were given PUVA the following year. The patients were followed up in autumn or the following spring and self-assessments were made of the severity and frequency of PLE episodes. Two hundred and eighty-one courses of UVB and 99 courses of PUVA were evaluated. At follow-up, 88% of those who received PUVA and 89% who received UVB reported good or moderate improvement, and of the 29 who received both PUVA and NB-UVB, 12 favoured PUVA, four preferred NB-UVB and five liked both equally.[78]

Mastalier et al. presented 14-year retrospective data on 79 patients treated with phototherapy; 17 patients with oral 5-MOP or topical 8-MOP PUVA, 56 with BB-UVB and six with UVA alone. The patients were assessed on the first summer after phototherapy for prevention of PLE episodes. The photoprotection rate was complete/partial remission in 65% for PUVA, 82% for BB-UVB and 83% for UVA. However, the authors acknowledged that PUVA was given to patients with more severe symptoms, while BB-UVB and UVA alone were given to patients with milder symptoms.[79]

The latter two studies were retrospective chart review studies, with smaller numbers of patients receiving PUVA compared with UVB; moreover, patients who had more severe PLE, or who had previously failed UVB, tended to receive PUVA.[78,79]

Safety of Psoralen–Ultraviolet A Compared With Ultraviolet B in PLE: There is an inherent risk of provoking the underlying condition in treating any photodermatosis. There was limited comparative data between the two forms of phototherapy to ascertain which form is more likely to provoke the eruption. Pooling the incidence of adverse events from the small number of PLE studies, the side-effects of rash provocation, erythema and pruritus were found to be common in both forms of phototherapy, although more common with UVB than with PUVA ( Table 3 ). However, as the number of patients in each cohort was small and the severity of the adverse events not directly comparable, the overall percentages should be regarded with caution.

In the treatment of PLE, the side-effects of rash provocation, erythema and pruritus were found to be common in both forms of phototherapy. There was no appreciable difference in the efficacy of PUVA and UVB.

Photodermatoses Other Than Polymorphic Light Eruption: No comparative studies with UVB were identified for the other photodermatoses.

When Should Psoralen–Ultraviolet A be Used in Preference to Other Therapies for Photodermatoses?. Photodermatoses are chronic conditions requiring ongoing treatment. Photoprotective measures and symptomatic treatment may be adequate in milder cases; however, in more severely affected patients, second-line treatments are often required. First-line treatment may include topical corticosteroids, antihistamines and β-carotene (for EPP), and is not within the scope of this guideline. Second-line treatment may include phototherapy. NB-UVB is generally considered before PUVA because it has the following advantages: lower risk of photocarcinogenesis, no risk of nausea or other side-effects associated with the ingestion of MOP and increased convenience as there is no need to take tablets or use eye protection post-treatment.

Polymorphic Light Eruption: There is good evidence of efficacy with UVB compared with PUVA.[81] There are no other direct comparative studies with other modalities of treatment, which include systemic corticosteroid, azathioprine, ciclosporin, hydroxychloroquine, β-carotene, nicotinamide, omega-3 fatty acids, antioxidants and Escherichia coli infiltrate.[90–104] Systemic corticosteroid has been demonstrated to reduce the severity of PLE when used as a short course on sunny holidays.[105] Systemic immunosuppression was found to be effective in a small number of case reports. Efficacy of treatment with hydroxychloroquine, β-carotene, nicotinamide, omega-3 fatty acids, antioxidants, and E. coli infiltrate has not been well established in RCTs.

Recommendation (Strength of Recommendation D; Level of Evidence 4): In the treatment of PLE, NB-UVB should be considered before PUVA. However, PUVA should be considered if NB-UVB has failed or has previously triggered the eruption, or if there are other practical issues. PUVA should be considered before other systemic treatments.

Chronic Actinic Dermatitis: We found no comparative studies using PUVA in the treatment of CAD. There are a limited number of case reports and small case series reporting the use of PUVA in CAD.[106–112]

The largest case series (only four patients) reported the use of PUVA with topical corticosteroid cover and inpatient supervision in UV-protected rooms (first dose was 0·25 J cm−2, with increments of 0·25–2·50 J cm−2, twice a week; maintenance at 10 J cm−2 fortnightly, then monthly). After each of the first six exposures, topical betamethasone or hydrocortisone butyrate was applied immediately to the trunk and 1% hydrocortisone to the face. In two of the patients who were phototested, there was resultant normalization of MEDs to UVB and UVA, and reported normal sun tolerance at 5-year follow-up, with continued monthly maintenance treatment.[108,109]

Treatment with PUVA covered with oral prednisolone and initial inpatient supervision has been reported in three case reports.[111,112] Prednisolone (60–100 mg) was administered daily initially and tapered over 4 weeks, while PUVA was given 3–4 times a week initially and tapered to once or twice weekly. Maintenance PUVA was given for 6–18 months. One patient with a borderline-normal MED to UVB and a low MED to UVA when retested had normal UVA response 4 months after stopping treatment in winter. All reported improvement of sun tolerance during maintenance treatment.

Successful use of ciclosporin, despite concerns about increasing risks of carcinogenesis using it with PUVA, at 5 mg kg−1 daily in combination with PUVA and initial prednisolone cover, has also been reported in case reports.[106,110]

Mycophenolate mofetil 1 g twice daily was reportedly effective when administered in combination with PUVA and initial prednisolone cover in two cases, followed by maintenance PUVA once weekly.[107]

One tertiary referral unit in the U.K. routinely treats patients with CAD with PUVA with success (T. Garibaldinos, personal communication). Prednisolone (20–30 mg) is taken on the day of phototherapy; small-dose increments of 0·05 J cm−2 are given with each UVA exposure; the course is given three times a week, then reduced to twice weekly and then once weekly. Inpatient supervision is not generally required.

Recommendation (Strength of Recommendation D; Level of Evidence 4): In the treatment of CAD, PUVA therapy should be considered in a specialist unit experienced in managing this disease, with full knowledge of the individual patient's action spectrum. Special precautions, including inpatient supervision and topical/oral corticosteroid cover, may be required.

Idiopathic Solar Urticaria: We found no comparative studies with PUVA in the treatment of SU. As would be anticipated in this rare disorder, a limited number of studies report the use of PUVA or other treatment options in SU.

PUVA has been used in small case series as monotherapy (four patients in one study and six in another),[84,113] in combination with intravenous immunoglobulins (IVIg; single case reports) and with plasmapheresis.[114–116] As monotherapy, PUVA was started at either the MPD, based on an erythema reaction without urticaria, or 0·25 J cm−2, whichever was lower.[84] Treatments were given three times weekly, with gradually increasing areas of skin exposed to UVA. At the end of treatment, a fourfold increase in the minimum urticarial dose (MUD) and improved sun tolerance were reported. In the other case series (n = 6), whole-body PUVA was given with a starting dose of 80% MUD, three times weekly, for 4–8 weeks, with a resultant increase in MUD and sun tolerance (from < 15 min pretreatment to 2–8 h post-treatment).[113] Thus, there is some evidence that PUVA may be beneficial in SU.

Antihistamines are regarded as the standard therapy in SU (other than photoprotection); RCTs are not available in SU itself and it is reported that a substantial proportion of these patients receive only modest benefit.[117–119] High-dose H1 antihistamines are frequently prescribed, in-line with published research and recommendations made in general for other chronic urticarias.[120]

NB-UVB has also been reported to be helpful in SU [Calzavara-Pinton et al. (n = 39), Wolf et al. (n = 1), Collins and Ferguson (n = 1)].[121–123] In an open trial using two protocols of NB-UVB in patients both with UVB-sensitive (n = 29) and UVB-insensitive SU (n = 10),[121] NB-UVB was well tolerated; 20% of both groups reported SU relapse at 3 months. There are no comparative trials between NB-UVB and PUVA.

A small number of case reports of combination therapy of PUVA with plasmapheresis,[115] and with IVIg,[114] reported an increase in sun tolerance.

There was limited evidence from small case series on the use of UVA, either administered on its own or for pre-PUVA desensitization.[124–127] This has been used in patients who have a very low MUD for UVA, where PUVA was thought to be unsafe. Increased MUDs were reported in all cases (n = 14), and improved sun tolerance in eight of 10 patients, but long-term follow-up was lacking. UVA alone was observed to have the advantage over PUVA of reduced risk of an acute phototoxic reaction and long-term side-effects such as skin cancers, and increased convenience. It was suggested to be worthwhile considering if patients demonstrate a higher MUD with UVA alone compared with PUVA or NB-UVB.[125] Maintenance treatment may be required, although long-term follow-up is lacking to ascertain the frequency and UVA dosage required.

Recommendation (Strength of Recommendation D; Level of Evidence 4): In the management of SU (after full assessment, including definition of the action spectrum), PUVA can be considered. The treatment should be carried out with full knowledge of the patient's action spectrum, in a specialist unit experienced in managing this disease.

Erythropoietic Protoporphyria: We did not find any comparative trials with PUVA in the treatment of EPP.

There was limited evidence in the use of PUVA in EPP.[128,129] Roelandts reported a therapeutic benefit of PUVA in patients who had not responded sufficiently to β-carotene, but did not include data on the number of patients treated or outcomes.[128] Oral 8-MOP was given and whole-body irradiation was administered three times weekly, with 20% increments twice weekly; in total, 26–30 irradiations were given. Another case report showed a threefold increase in MED to UVA with twice-weekly treatment with oral TMP and UVA in comparison with a fourfold increase in MED to UVA with previous β-carotene therapy.[129]

PUVA cannot currently be recommended for the treatment of EPP, owing to lack of evidence and comparative trials with β-carotene or UVB, and the long-term risks associated with the need for annual treatment on a lifelong basis.

UVB has been reported in small open studies to be effective in EPP (six patients in one study; one patient in another; and 12 in yet another).[123,130,131]

Recommendation (Strength of Recommendation D; Level of Evidence 4): PUVA is rarely appropriate in EPP, in which NB-UVB is the phototherapy of first choice.

Actinic Prurigo: We found no comparative trials of thalidomide or immunosuppressants with PUVA in the treatment of AP.

There was limited evidence of the use of PUVA in AP. Treatment with PUVA 2–3 times weekly for 15–20 weeks induced normal sun tolerance and increased the MED to UVA immediately post-PUVA, but clinical photosensitivity recurred after 4–6 months' follow-up.[132,133]

UVB has been reported, in a small open study, to be effective in AP (n = 6).[123]

Recommendation (Strength of Recommendation D; Level of Evidence 4): NB-UVB may be a safer therapeutic option in terms of phototherapy-associated carcinogenic risk in patients with AP, particularly in children, and should be considered before PUVA.

What Special Precautions Should be Undertaken During Psoralen–Ultraviolet A Therapy of Photodermatoses?. Photodermatoses run a high risk of provocation with the different forms of phototherapy. Special precautions are required and are described below for the better-reported photodermatoses PLE, CAD and SU. In temperate countries, PUVA is usually administered in early spring. The benefit conferred by phototherapy is diminished or lost several weeks postphototherapy, and post-treatment advice generally includes continued natural sunlight exposure, if tolerable, on an individual basis, to keep resistance to provocation for the rest of summer. As PUVA generally needs to be repeated, the longer-term risk of skin carcinogenesis needs to be weighed against the therapeutic benefit, and annual desensitization is not usually recommended.

Polymorphic Light Eruption: The regimens vary between phototherapy centres; most administer 8-MOP, although 5-MOP, and oral and bath TMP have all been used.[78–89] In most of these studies, 8-MOP PUVA was given three times weekly for 12–20 treatments. Currently in the U.K., a twice-weekly regimen is standard. There are no studies comparing the efficacy and safety of twice- and three times-weekly regimens.

Although there were no studies on the timing of PUVA therapy during the year, this is likely to be an important consideration, particularly in temperate climes. If administered too early in the year, the photoprotective effect may have subsided by mid-summer; administered too late, and the patient may have already suffered an eruption and PUVA may increase the risk of provocation or further aggravation.

The risk of provoking PLE is high, particularly with the first few PUVA exposures. At least one episode was induced during 12–50% of PUVA treatment courses,[78,80–83,86,87,134] which is a little lower than the rates with UVB (48–62%).[78,81] There is evidence that PUVA is as likely as UVB to cause provocation of PLE. Provocation episodes can be managed with potent topical steroid and subsequently lower dose increments, omitting one or two treatments if particularly severe.[78,80,81,86,87] To prevent provocation, one group administered oral prednisolone (40–50 mg) for the first 2 weeks of phototherapy,[85] while another reported routine prophylactic application of a potent topical steroid after each exposure in UVB phototherapy.[78]

Pruritus was also reported in 18–33% of patients,[81,82] with one group managing this with oral corticosteroid.[82]

Post-treatment advice generally includes continued natural sunlight exposure, ranging from 2 h weekly to 'cautious exposure, with sunscreens for extended outdoor stay' to 'expose freely to sun'.[78,83,85,88]

Chronic Actinic Dermatitis: PUVA phototherapy is generally undertaken under close supervision under cover of topical or systemic corticosteroid, as discussed above.[108,109,111,112] Maintenance treatment may be required. Annual repeated courses can be considered but the benefit needs to be balanced against the long-term risk of skin carcinogenicity.

Solar Urticaria: Treatment of SU with phototherapy can potentially result in provocation, syncope and anaphylaxis. UVA is often, but not always, a wavelength where there is sensitivity in this disease. The choice of PUVA for SU as a therapy, and the protocol used, will be governed by the action spectrum as determined by monochromator phototesting.

Thus, it is important to determine both the action spectrum of the disorder, in a specialized phototesting unit, and the MUD before starting phototherapy, preferably with the UV source to be used for treatment. The initiating dose should be lower than the MUD. In patients with a very low MUD, UVA alone or pre-PUVA UVA has been used.[124–127] The amount of photoprotection will subside, and maintenance treatment with PUVA or UVA has been reported.

Key Points. When using prophylactic phototherapy to treat photodermatoses, the risk of rash provocation is high. The use of topical/oral corticosteroids may be used prophylactically or therapeutically in PLE or CAD.

Phototherapy may be given in spring, or before a sunny holiday, before an expected increased natural UV exposure.

Particularly with CAD and SU, the action spectrum should be determined prior to commencement of phototherapy; PUVA should only be carried out in an experienced specialized unit.

With AP and EPP, UVB should be considered before PUVA.

Post-treatment advice of cautious, continued sunlight exposure should be given to maintain the photoprotective effect, but this is dependent on individual tolerance and photoprotection is still needed in very sunny conditions.

Hand and Foot Dermatoses

Local PUVA, using oral and topical psoralens, has been widely used to treat hand and foot dermatoses over the last 25 years and is generally considered an effective therapy. Oral psoralen is favoured by some patients for convenience (less time spent in a hospital treatment unit) and because there are lower risks of phototoxic reactions. Topical psoralen is favoured by other patients who prefer not to take oral medication and to avoid the potential risks of systemic side-effects and/or drug interactions. In addition, the use of eye protection outside the UV irradiation period is not required. Different forms of topical psoralen are in use, including soaks, paint, cream, ointment, emulsion and gel. Treatment regimens also vary not only with regard to methods of psoralen application, but also in the time interval between psoralen application and UVA irradiation. It has previously been shown in a small study that, in unaffected skin of the palms and soles, there is a lag time of approximately 40 min between soaking in psoralen solution and maximum UVA sensitivity in the palmar and plantar skin of healthy volunteers.[135] As a result of this, and earlier reports that suggested efficacy in the hand and foot dermatoses of topical psoralens with a 1–2-h delay between psoralen application and UVA exposure but no efficacy in studies with psoralen application immediately before UVA exposure, previous U.K. guidelines have recommended a delay of at least 30 min between immersion and irradiation.[136]

The hand and foot dermatoses comprise three main conditions: eczema, psoriasis and palmoplantar pustulosis.

What is the Efficacy of Psoralen–Ultraviolet A Therapy in Patients With and and Foot Dermatoses Compared With Placebo or Other Active Treatments?. Palmoplantar Eczema: Oral PUVA has been compared with UVB in a prospective contemporaneous controlled study of hand eczema (n = 35).[137] This showed both treatments to be effective but PUVA was superior to UVB, although the relapse rate was high. A subsequent smaller study (n = 20) carried out by the same group using a similar design showed similar results.[138] Uncontrolled studies with oral PUVA have shown significant improvement or clearance in 81–86% of patients with hand and foot eczema,[139,140] with similarly good results in smaller case series.[141,142]

Uncontrolled studies of topical PUVA have suggested this to be effective treatment with clearance or considerable improvement reported in 58–81% of dyshidrotic eczema and 50–67% of hyperkeratotic eczema.[143–154] However, evidence from comparative studies was less convincing. A left–right, within-patient study (n = 15) with 8-MOP paint vs. placebo in dyshidrotic eczema found no difference between the two groups, and no patient achieved clearance.[155] The lack of efficacy was supported by a small controlled study (n = 6) with 20% 8-MOP gel in which no patient cleared.[156]

A significant difference between topical PUVA compared with UVB has not been shown. In a nonrandomized, within-patient study (n = 13) of trioxsalen soaks compared with UVB, the reduction of severity scores of 25% and 39%, respectively, was not clinically significant.[157] A further randomized study (n = 15) with 8-MOP PUVA paint and NB-UVB in dyshidrotic eczema showed a reduction in mean severity scores of 75% and 75%, and clearance rates of 8% and 17%, respectively.[158]

Topical 8-MOP cream was compared with high-dose UVA1 in a nonrandomized, within-patient study with dyshidrotic hand eczema (n = 27).[159] No patient obtained clearance.

Topical PUVA paint compared with superficial X-ray showed significant improvement in both groups in a randomized double-blind trial (n = 25) of dyshidrotic hand eczema. At 6 weeks the scores for radiotherapy were significantly better than for PUVA but this was not maintained at follow-up.[160]

For foot eczema, topical PUVA has been compared in a randomized trial to topical PUVA combined with iontophoresis and to potent topical steroid (n = 48) over an 8-week treatment period and at follow-up 8 weeks later. There was no significant difference in eczema and Dermatology Life Quality Index (DLQI) scores between the three groups.[161]

In summary, in hand eczema, evidence from two prospective contemporaneous controlled studies have shown PUVA using oral psoralen to be effective and superior to UVB. Evidence from RCTs and prospective contemporaneous controlled studies has not demonstrated topical PUVA to be more effective than placebo, UVB, UVA1 or superficial X-ray therapy. However, some uncontrolled studies have shown topical PUVA to be associated with good degrees of improvement. All comparative studies lacked the power to detect what might still be important differences between treatments.

Recommendation (Strength of Recommendation B; Level of Evidence 2++): PUVA, using oral psoralen, should be considered as a treatment for hand and foot eczema. Although the evidence for topical PUVA in palmoplantar eczema is weak, lack of proof of efficacy does not prove no efficacy and in open discussion with patients it will sometimes be appropriate to consider topical PUVA.

Palmoplantar Psoriasis: We found no comparative studies comparing oral PUVA with placebo or other active treatments, although a small retrospective study (n = 12)[162] and case series (n = 5)[141] report improvement or clearance.

A large retrospective study (n = 92) compared oral PUVA with NB-UVB and concluded they were equally effective. However, assessment and outcome measures were unclear and no statistical analysis was carried out.[163]

Uncontrolled studies have reported clearance or significant improvement occurring in 58–87% of patients treated with topical psoralen.[143,146–148,150,151,164–167]

In a randomized, left–right, within-patient study in patients with hand and foot dermatoses (psoriasis, n = 11), topical PUVA (20% 8-MOP gel) was compared with UVA, and PUVA was suggested to be more effective overall, but the results were not analysed separately according to the different dermatoses investigated.[156] No patient was reported as being cleared in either group.

Topical PUVA was compared with UVB in two small randomized trials. A left–right, inpatient, observer-blinded study compared PUVA cream with 308-nm UVB excimer laser (n = 10).[168] Both treatments showed similar efficacy, with a mean reduction in modified PASI score of 64% with PUVA and 64% with excimer laser. In a further randomized study (n = 25), local PUVA with 8-MOP paint was compared with NB-UVB, with a reduction in severity scores of 85% compared with 61% with NB-UVB, and clearance rates of 24% and 0%, respectively.[169] The authors of a Cochrane systematic review of NB-UVB phototherapy vs. BB-UVB or PUVA for psoriasis identified only this study for the indication, and concluded no significant difference between treatments.[170]

In summary, for palmoplantar psoriasis there is some randomized study evidence for topical PUVA; however, these studies were underpowered. Oral PUVA for this indication has not been studied adequately.

Recommendations (Strength of Recommendation C; Level of Evidence 2++): PUVA using topical or oral psoralen should be considered as a treatment for palmoplantar psoriasis.

Palmoplantar Pustulosis: A randomized, within-patient study compared oral PUVA with no treatment (n = 22).[171] This showed improvement in 100% of PUVA-treated sites compared with 59% of untreated sites, with 55% and 0% clearance rates, respectively. In a prospective contemporaneous controlled study (n = 14), improvement was recorded in 64% of participants receiving oral PUVA and 14% with no treatment.[172] The clearance rates were 21% and 0%, respectively.

A double-blinded, randomized, left–right, within-patient study compared topical PUVA with placebo (n = 27).[173] The study detected no difference in improvement with topical PUVA compared with placebo. The lack of efficacy of topical PUVA was further supported by smaller randomized studies (10 and five patients, respectively),[156,174] with clearance rates of only 0–10%.

Two RCTs have compared PUVA with oral retinoids. One study (n = 84) compared two forms of topical PUVA, oral PUVA and etretinate.[175] A total of 12% (4/33) treated with topical methoxsalen cream cleared compared with 70% (14/20) in the etretinate group. None cleared in the oral PUVA and trioxsalen soak group. This suggests etretinate to be more efficacious, but response to PUVA, and, in particular, to oral PUVA, was much lower than in other published studies. Another study comparing oral PUVA to etretinate showed clearance in 21% (3/12) receiving PUVA vs. 17% (3/18) receiving etretinate.[172]

The authors of a Cochrane systematic review of interventions for palmoplantar pustulosis, which included the controlled studies detailed above, concluded that oral PUVA is an effective intervention for this indication, although the combination of retinoids plus oral PUVA is more effective. No proven benefit has been demonstrated with topical PUVA and no definite benefit of retinoid as a monotherapy over PUVA was established.[176]

Recommendations (Strength of Recommendation C; Level of Evidence 2++): Oral PUVA should be considered as a treatment for palmoplantar pustulosis.

What is the Efficacy of Oral Psoralen–Ultraviolet A Therapy Compared With Topical Psoralen–Ultraviolet A Therapy in Patients With Hand and Foot Dermatoses?. Palmoplantar eczema: Two RCTs have compared oral and topical PUVA. van Coevorden et al. performed an open-label RCT (n = 158) comparing oral methoxsalen (safer for home use than topical psoralens) using a home UVA unit with topical PUVA using trioxsalen soaks in a hospital setting over a 10-week period.[177] The mean hand eczema scores showed a significant reduction of 41% in the oral PUVA group compared with 31% in the topical PUVA group, with no statistically significant difference between the two groups. This improvement was maintained at an 8-week follow-up.

A small RCT compared oral 8-MOP with 8-MOP soak in dyshidrotic eczema (n = 15) and hyperkeratotic eczema (n = 12). This study did not detect a difference in therapeutic efficacy, or that dyshidrotic eczema improved to a greater extent than hyperkeratotic eczema; however, the study was underpowered and does not justify these conclusions.[178]

Palmoplantar Psoriasis: A retrospective review by Hawk and Grice of oral and topical 8-MOP PUVA in the treatment of patients with hand and foot dermatoses included 18 patients with psoriasis.[179] Fifty per cent (7/14) of patients who received topical PUVA and 50% (2/4) oral PUVA cleared. A small, within-patient, randomized, right–left comparison study compared PUVA using oral oxsoralen with PUVA using 8-MOP soaks in patients with palmoplantar dermatoses but included only three patients with psoriasis. Oral PUVA was reported to be more effective.[180]

Palmoplantar Pustulosis: Hawk and Grice's retrospective review, which included 15 patients with palmoplantar pustulosis, showed clearance in 67% (4/6) with oral 8-MOP and 33% (3/9) with topical 8-MOP emulsion.[179]

A small, right–left, within-patient comparative study of oral oxsoralen vs. 8-MOP soaks in patients with hand and foot dermatoses included five patients with pustulosis. Oral PUVA was reported to be more effective.[180]

The RCT by Lassus et al. of two forms of topical and oral PUVA showed a clearance rate of 8% (4/51) with topical PUVA compared with 0% (0/13) with oral PUVA.[175] PUVA as a whole was disappointing in that study and no difference between the methods of delivering it was detected. A Cochrane review of interventions for palmoplantar pustulosis found only this one direct comparison of topical with oral PUVA for this indication.[176]

Recommendation [Strength of Recommendation D (Good Practice Point); Level of Evidence 4]: Oral PUVA should usually be considered as the first-line PUVA treatment for patients with palmoplantar dermatoses.

 What is the Efficacy of Psoralen–Ultraviolet A Therapy Alone Compared With Psoralen–Ultraviolet A and Adjuvant Therapies in the Hand and Foot Dermatoses?. Palmoplantar Eczema: For foot eczema, topical PUVA has been compared, in a randomized trial, to topical PUVA combined with iontophoresis over an 8-week treatment period and at follow-up 8 weeks later. There was no significant difference in eczema and DLQI scores.[161]

Palmoplantar Psoriasis: There was only one RCT comparing combination therapy with oral PUVA and etretinate vs. PUVA and placebo but this included only three patients with psoriasis.[181] No other relevant studies were identified.

Palmoplantar Pustulosis: Three RCTs examined combination therapy with oral PUVA and retinoids compared with PUVA alone. The RCT by Rosén et al. showed clearance in 61% (14/23) of patients receiving combination therapy and 21% (three of 14) receiving PUVA alone.[172] A further RCT comparing oral PUVA–etretinate vs. oral PUVA alone showed clearance in 100% (8/8) and 44% (four of nine) of patients, respectively.[181] An RCT using 8-MOP PUVA lotion found 60% (six of 10) of patients clearing with topical PUVA–etretinate compared with 10% (one of 10) receiving topical PUVA alone.[174]

A Cochrane review of the interventions for palmoplantar pustulosis concluded there was evidence for an increased efficacy of retinoids combined with PUVA compared with PUVA used alone.[176]

Recommendation (Strength of Recommendation 1+; Level of Evidence A): Unless there are contraindications, the combination of oral PUVA with oral retinoids should be considered as a treatment for palmoplantar pustulosis.