DOI: 10.1111/j.1468-3083.2012.04519.

JEADV

REVIEW ARTICLE

Efcacy of Psoralen UV-A therapy vs. Narrowband UV-B therapy in chronic plaque psoriasis: a systematic literature review
tre, S. Aractingi, E. Archier,,* S. Devaux, E. Castela, A. Gallini, F. Aubin,** M. Le Ma H. Bachelez, B. Cribier, P. Joly,*** D. Jullien, L. Misery, C. Paul, J.-P. Ortonne, M.-A. Richard,
INSERM CRO2 and Dermatology Department, Aix-Marseille Univ, UMR 911, Timone Hospital, Marseille Dermatology Department, Paul Sabatier University, Toulouse Dermatology Department, Nice University, LArchet II Hospital, Nice UMR 1027 INSERM, Paul Sabatier University, Toulouse University, EA3181, IFR133, University Hospital, Besanc ** Dermatology Department, Franche-Comte on Dermatologist, Caen Dermatology Department, Tenon Hospital, APHP and Paris 6 University, Paris Dermatology Department, Saint-Louis Hospital, Paris Dermatology Department, University Hospital, Strasbourg *** Dermatology Department, Charles Nicolle Hospital, INSERM U 905, University of Rouen Dermatology Department, Edouard Herriot Hospital, Lyon, and Dermatology Department, University Hospital, Brest, France *Correspondence: E. Archier. E-mail: elodie.archier@live.fr

Abstract
Background Oral 8-methoxypsoralen-UV-A (PUVA) and Narrowband UV-B (NB-UVB or UVB TL-01) are well established treatments for chronic plaque psoriasis but there is limited evidence regarding their respective efcacy. Objectives To prepare for evidence-based recommendations concerning the practical use of oral 8-methoxypsoralen-UV-A and Narrowband UV-B in psoriasis, a systematic review to assess respective response rates, remission duration and predictive factors of efcacy was performed. Methods A systematic search was carried out in PubMed, Cochrane and Embase databases, using the key words Psoriasis, UVB therapy, UVA therapy for the period from 1980 to December 2010. Results The initial literature search identied 773 articles. The nal selection included 29 randomized controlled trials: 18 were about the efcacy of PUVA, eight about the efcacy of NB-UVB and three directly compared PUVA vs. NB-UVB. The response rate dened by 75% or more improvement in PASI was 80% with PUVA vs. 70% with NB-UVB. The meta-analysis of the three comparative studies found a higher probability of remission at 6 months with PUVA than with NB-UVB [OR = 2.73 (95% CI 1.19-6.27), P = 0.02]. The choice of initial dose, according to skin type, the minimal erythemal dose or minimal phototoxic dose, incremental regimen and periodicity of the sessions did not appear to be predictive factors of efcacy for PUVA or NB-UVB. Despite methodological limitations in trials, the number of sessions needed for psoriasis clearance appeared to be lower with PUVA than with NB-UVB (approx. 17 vs. 25, respectively). Conclusion PUVA and NB-UVB are both effective therapies in treatment of psoriasis. Our results suggest that compared with NB-UVB, PUVA tends to clear psoriasis more reliably, with fewer sessions, and provides with longer lasting clearance. However, the long-term safety of PUVA, especially its cutaneous carcinogenic risk, and the easier administration procedure often lead dermatologists to prefer NB-UVB as rst line phototherapy treatment in plaque type psoriasis. Received: 10 February 2012; Accepted: 20 February 2012

Funding sources
Abbott France provided nancial support for publication but took no further part in the project. The authors have no nancial interest in the subject matter or materials discussed in the manuscript.

JEADV 2012, 26 (Suppl. 3), 1121

2012 The Authors Journal of the European Academy of Dermatology and Venereology 2012 European Academy of Dermatology and Venereology

12

Archier et al.

Conicts of interest
All the authors have been paid consultants of Abbott. In addition CP has been investigator and consultant for Janssen-Cilag, Leo, Novartis and Pzer. HB has been paid for consulting activities for Centocor, Janssen-Cilag, Leo Pharma, Novartis, Pzer and Schering-Plough. BC has been paid for consulting activities for Pzer, for redaction activities by Leo Pharma and Janssen Cilag, and speaker for Pzer, Leo Pharma and Schering Plough. DJ has been consultant for Merck, Janssen-Cilag, Novartis, Pzer and Schering-Plough MSD. JPO has been investigator, speaker and advisor for Schering-Plough MSD, Abbott, Merck Serono, Centocor, Pzer, Janssen Cilag, Pierre Fabre, Galderma, Leo Pharma, Meda. LM has been a paid consultant of Novartis, Janssen-Cilag, Leo Pharma, Pzer and Pierre Fabre. MAR has been investigator and consultant for Janssen-Cilag, Novartis, Pzer.

Introduction
Psoriasis is a chronic inammatory disease predominantly targeting the skin, the pathogenesis of which involves dermal and epidermal inammation as well as abnormalities in the proliferation and differentiation of keratinocytes.1 It affects 23% of the population worldwide. Oral 8-methoxypsoralen-UV-A (PUVA) and narrowband UVB (NB-UVB) are well-established and generally effective treatments for chronic plaque psoriasis, with therapeutic efcacy supposedly dependent on dose and frequency of irradiation. In 1974, Parrish et al. rst introduced the combination of UVA (320400 nm) with orally applied methoxsalen for the treatment of psoriasis, introducing the acronym PUVA.2 Yet, despite its high efcacy and relatively low incidence of acute side effects, the use of PUVA as a therapy for psoriasis has decreased lately. One of the reasons has been the introduction of narrowband UVB, which has gradually replaced PUVA as the rst-line treatment for chronic plaque psoriasis. In the early 1980s, a uorescent lamp emitting narrowband UV-B between 311 nm and 313 nm (referred to as the TL-01 lamp) was developed to improve the efcacy of phototherapy.3,4 The rationale for manufacturing such a lamp was derived from action spectra studies for psoriasis phototherapy in which longer wavelengths in the UV-B region were indicated to have the best ratio of antipsoriatic to erythemogenic activity. Otherwise, patients often prefer NB-UVB to PUVA because there is no need to wear protective eyewear or to take tablets, and they do not experience nausea. NB-UVB also has the advantage of being suitable for use during pregnancy and it can be administered in children. Otherwise, the ideal regimen, which would minimize both the number of exposures and the cumulative UV dose, has still to be dened; in other words, the respective efcacy of the two treatment modalities needs to be better assessed. To prepare for evidence-based recommendations concerning the respective efcacy of phototherapy using PUVA or NB-UVB in the treatment of chronic plaque psoriasis, we performed a systematic review of the literature. The reviewing process leading to these recommendations is described in details in the same issue of this journal.5

Materials and methods

A systematic review of all studies investigating PUVA and NB-UVB phototherapies in adult psoriatic patients, published between 1980 and December 2010, was made by two of the authors (EA, MAR). The Cochrane, Embase and Medline databases were systematically searched, using a combination of Medical Subject Headings (Mesh): psoriasis (Majr) AND UVB therapy (Majr) AND UVA therapy (Majr). The bibliographic search was limited to articles on human subjects over 19 years of age, articles published in English or French and only randomized controlled trials (RCT) were kept. Using pre-dened inclusion and exclusion criteria, all the references obtained in the three databases were screened. The selection of relevant studies was initially made by reviewing title and abstract, excluding manuscripts not addressing the previously mentioned topics, then, for the remaining studies, by reading the complete article. Final consensus was reached through discussion between all the members of the board of experts. Wherever possible, the odds ratio (OR) and corresponding 95% condence interval (CI) were calculated from the available data (AG) if these parameters were not already given in the paper. Where appropriate, meta-analyses of treatment effects were carried out using random effect models. Calculations were performed using the Data analysis and Statistical Software STATA 9.1 (StataCorp LP, College Station, Texas, USA) and the Cochrane Software Review Manager (RevMan) Version 5.0 (Cochrane Nordic Centre, Copenhagen, Denmark).

Results
The primary purpose of this systematic literature review was to compare the respective efcacy of PUVA and of NB-UVB in psoriasis. Furthermore, the criteria retained from all the available studies for determining efcacy were variable and heterogeneous. In almost all studies, treatment efcacy was dened as clearance (complete resolution of plaques), sometimes as clearance or MRA i.e. minimal residual activity, affecting less than 1% of the body surface area. In other studies, a satisfactory response to therapy was dened as a 90% reduction of baseline Psoriasis area and severity index (PASI) (PASI 90), or as a 75% reduction in baseline PASI (PASI

JEADV 2012, 26 (Suppl. 3), 1121

2012 The Authors Journal of the European Academy of Dermatology and Venereology 2012 European Academy of Dermatology and Venereology

Efcacy of Psoralen UV-A therapy vs. Narrowband UV-B therapy

13

Potentially relevant studies published between 1980 and 2010 identified by electronic search (cochrane library, pubmed and embase), n = 773

- Articles did not meet eligibility criteria were excluded, n = 521 -Duplicated excluded, n = 149 Potentially relevant articles eligible for more detailed evaluation, n = 103 - Not original articles were excluded, n = 73 - Duplicated excluded, n = 1 29 articles selected

Figure 1 Flow chart of study selection process.

75). In the oldest studies, the SEI (scaling, erythema, inltration) score was used, in others, clearance was dened as a PGA (Physician Global Assessment) score of 0. Moreover, 3 parameters from phototherapy protocols that could inuence treatment efcacy were studied when available: - Starting UV dose, determined according to the MED (minimal erythemal dose) for NB-UVB or the MPD (minimal phototoxic dose) for PUVA, or according to the patients skin type (dened as Fitzpatricks). - UV incremental regimen according to skin type, MED MPD or a xed increment. - Number of sessions delivered per patient per week. We also assessed the total number of sessions and the cumulative UV dose needed for clearance for both PUVA and NB-UVB. We identied a total of 773 eligible articles by searching the Medline, Cochrane and Embase databases. Of these, 670 were eliminated by reading the title or the abstract and 74 after reading the article. A total of 29 randomized controlled trials were nally selected.634 The detailed owchart describing both selection process and reasons for exclusion is displayed in Fig. 1. Only three of the selected articles directly compared the respective efcacy of PUVA and NB-UVB in psoriasis.3234 To increase the power to detect potential differences we combined the results from randomized controlled studies with those extracted from therapeutic trials in which the control arm used PUVA or NB-UVB (Table 1) and from studies evaluating a relevant parameter (initial dose, increments) (Table 2). A total of 18 publications concerning NB-UVB,623 and eight concerning PUVA were nally included in the analysis.2431
Efcacy of NB-UVB in chronic plaque psoriasis

We rst examined the inuence of the method used to establish the initial dose on clearance rate: 16 studies used the MED to deter-

mine the initial dose of NB-UVB phototherapy,615,19,20,23,3234 four used skin type 17,18,21,22 and three studies used a xed dose of NB-UVB.11,16,18 Fig. 2 represents the efcacy of NB-UVB according to the different strategies to determine initial NB-UVB dose, for the 12 studies with data available for clearance rate (Fig. 2). There was no clear difference in efcacy between the different initial dosing strategies. However, condence intervals were often large, owing to the small sample size in most studies. We then examined the inuence of dose-increment strategies on clearance rate. Only ve studies were available. In two RCT, published by Evers et al. 6 and Kleinpenning et al.,7 the same cohort of 109 patients was studied. These patients were randomized to receive either NB-UVB irradiation with a suberythematogenic regimen (54 patients, initial dose = 35% of MED, increased at each irradiation session by 20%) or an erythematogenic regimen (55 patients, initial dose = 70% of MED, increased at each irradiation session by 40%), three times weekly. There was no signicant difference in the number of patients who achieved clearance between both randomization schemes (74.5% vs. 66.7%). However, the patients receiving the erythematogenic regimen needed signicantly fewer irradiation sessions to achieve clearance than the patients receiving the suberythematogenic regimen (20.6 6.9 sessions vs. 24.1 6.1 sessions, respectively) with no signicant differences in cumulative UVB dose. There was no signicant difference between median PASI scores achieved after 10 weeks of treatment in the study by Altiner et al., comparing 15 patients who underwent a weekly dose increment (once in three treatments) to 15 patients who had a daily dose increment.8 Selvaag et al.18 in a 20 patients randomized controlled left-right comparison study, compared, NB UVB with xed dose increments vs. NB UVB doses determined individually according to skin pigmentation quantied by skin reectance. PASI showed faster improvement during the rst 2 weeks of treatment on the body side receiving individualized treatment compared with the other side (% of PASI reduction = 47 vs. 43% respectively, P < 0.02). However, there was no signicant difference in the reduction in PASI between body sides at the end of the study. In an intraindividually controlled, paired comparison of a low and high dose incremental regimen study published by Wainwright et al.,19 including 20 patients, the reduction in median psoriasis SEI scores was similar with both regimens. The median cumulative UVB dose to clearance was lower on the low increment side than the high increment side (P < 0.005) but the median number of treatments required was marginally higher. However, the median duration of UV treatment (53.5 days) was identical under both regimens for all patients except one. Only two studies 9,10 evaluated the optimal treatment frequency for NB-UVB needed to achieve complete clearance in psoriasis. In the rst study published by Cameron et al.,9 113 patients received whole body NB-UVB phototherapy either twice or three times weekly. Forty patients in the twice weekly group reached clearance or MRA, as did 44 in the three times weekly group (P = 0.21). It

JEADV 2012, 26 (Suppl. 3), 1121

2012 The Authors Journal of the European Academy of Dermatology and Venereology 2012 European Academy of Dermatology and Venereology

14

Archier et al.

Table 1 Randomized controlled trials including psoriatic patients, trials in which the control arm used PUVA or NB-UVB
Study (author and year of publication) NB-UVB Mahajan 201016 Methotrexate (0.5 mg kg) NB-UVB (3x week) vs. Placebo NB-UVB (3x week) NB-UVB 3x week+ LCD (coal tar) 2x d vs. NB-UVB 3x week NB-UVB 3x week+ alefacept IM 1x week vs. NB UVB 3x week+ placebo IM 1x week NB-UVB 3x week vs. Selective UVB 3x week NB-UVB 4x week vs. Cream PUVA therapy 4x week vs. NB-UVB + cream PUVA for elbows, knees, and sacral regions NB-UVB 3x week vs. Short contact dithranol vs. Dithranol inpatient Spa water alone 5x week vs. NB-UVB alone 5x week vs. Combination spa water +NB-UVB 5x week Low-dose NB-UVB 3x week (incremental doses) vs. Low-dose NB-UVB 3x week (incremental doses) + calcipotriol 50 lg g 2x d on the affected skin. NB-UVB + acitretine 50 mg j vs. NB-UVB + placebo vs. acitretine alone PUVA 3x week vs. UVA + placebo 3x week Bath PUVA vs. Oral PUVA cross over for 10 patients Etretinate alone vs. Etretinate + PUVA (3x week): 4 weeks Retinoides then 6 weeks PUVA. vs. Etretinate + PUVA (3x week): 10 weeks Retinoides and 6 weeks concomitant PUVA vs. PUVA alone (3x week) n = 20 Design No of patients analyzed in the phototherapy control arm

Bagel 200917

n = 12

Jacobe 200812

n=7

Kirke 200713

n = 50

Grundmann 200414

n = 10

Swinkels 200415

n = 78

aute -Labre ` ze 200121 Le

n = 21

Brands 199922

n = 28

Lowe 199123

n = 16

PUVA Sivanesan 200929 n = 30

Cooper 200031

n = 17

Lauharanta 198128

n = 20

NB-UVB, Narrowband UV-B; PUVA, oral 8-methoxypsoralen-UV-A.

took 1.5 (95% CI = 1.31.7) times longer to reach clearance MRA with twice-weekly UVB therapy than with three times weekly, with a geometric mean of 88 vs. 58 days (P < 0.0001).

In a randomized half-body within-patient paired study including 21 patients, Dawe et al.10 compared three and ve times weekly sessions of a NB-UVB regimen. They did not nd statistical

JEADV 2012, 26 (Suppl. 3), 1121

2012 The Authors Journal of the European Academy of Dermatology and Venereology 2012 European Academy of Dermatology and Venereology

Efcacy of Psoralen UV-A therapy vs. Narrowband UV-B therapy

15

Table 2 Randomized controlled trials including psoriatic patients, studying a relevant parameter concerning NB-UVB or PUVA
Study (author and year of publication) NB-UVB Initial dose Dawe 201111 NB-UVB xed starting dose = 140 mJ cm2 vs. NB-UVB 70% of MED vs. NB-UVB 50% of MED NB-UVB near erythemogenic regimen (35% of the MED) on one half of the body vs. NB-UVB far erythemogenic regimen (70% of the MED) on the other half NB-UVB, 70% MED, 40% incremental doses, 3x week (erythematogenic regimen) vs. NB-UVB, 35% MED, 20% incremental doses, 3x week (suberythematogenic regimen) NB-UVB 3x week, weekly dose increments vs. NB-UVB 3x week daily dose increments NB-UVB, conventional NB-UVB regimen with xed dose increments on one side of the body vs. NB-UVB, doses according to measurements of skin pigmentation quantied by the skin reectance technique on the other side NB-UVB 3x week, high increment dose regimen on one side vs. NB-UVB 3x week, low increment dose regimen on the other side NB-UVB 2x week NB-UVB 3x week NB-UVB 3x week on one side of the body vs. NB-UVB 5x week on the other side n = 210 Design No of patients analyzed

Hofer 199820

n = 13

Increment rates Kleinpenning 20097 Evers 20096 Altiner 20068 n = 109

n = 30

Selvaag 200018

n = 20

Wainwright199819

n = 20

Treatment frequency Cameron 20029 Dawe 199810 n = 113 n = 21

PUVA Initial dose Kirby 199926 PUVA with initial dose determined by skin type vs. PUVA with initial dose determined by MPD PUVA 2x week: high dose regimen based on MPD with percentage incremental UVA increases on one side of the body vs. PUVA 2x week: low dose regimen based on skin type with xed incremental UVA increases on the other side PUVA 2x week with initial dose determined by MPD vs. PUVA 3x week with initial dose determined by skin type Group 1: PUVA 0.5 MPD 4x week vs. 1 MPD 2x week Group 2: PUVA 0.5 MPD 2x week vs. 1 MPD 2x week Group 3: PUVA 0.75 MPD 2x week vs. 0.5 MPD 2x week PUVA 2x week on one side vs. PUVA 3x week on the other side n = 85

Collins 199624

n = 37

Buckley 199525

n = 83

Legat 200430

n = 18

Treatment frequency Valbuena 200727 n = 28

NB-UVB, Narrowband UV-B; PUVA, oral 8-methoxypsoralen-UV-A; MED, Minimal Erythemal Dose; MPD, Minimal Phototoxic Dose.

differences in terms of clearance between the two arms, but time to clearance or to MRA was slightly longer with the three weekly than with the ve weekly regimen (40 vs. 35 days, P = 0.007)

It was not possible to extract with accuracy from the references the average number of sessions needed for psoriasis clearance. According to the available data, the average number of sessions

JEADV 2012, 26 (Suppl. 3), 1121

2012 The Authors Journal of the European Academy of Dermatology and Venereology 2012 European Academy of Dermatology and Venereology

16

Archier et al.

Study (ref)
Yones 2006 32 Gordon 1999 34 Markham 2003 33 Kleipenning 2009 7 Kleipenning 2009 7 Cameron 2002 9 Cameron 2002 9 Kirke 2007 13 Jacobe 2008 12 Grundmann 2004 14 Swinkels 2004 15 Dawe 2011 11 Dawe 2011 11 Bagel 2009 17 Mahajan 2010 16 Dawe 2011 11

N
34 51 29 55 54 58 55 50 7 10 78 70 70 12 Fixed dose 30 70 0 10 20 30 40 50 60 70 80 90 100 Dose according to MED

Dose according to skin type

Patients cleared or with marked improvement (%)

N: number of patients enrolled in each arm of the study MED: minimal erythemal dose

Figure 2 Clearance rates of NB-UVB in psoriasis according to initial dose determination.

Study (ref)
Yones 2006 32 Gordon 1999 34 Markham 2003 33 Kleipenning 2009 7 Kleipenning 2009 7 Cameron 2002 9 Cameron 2002 9 Kirke 2007 13 Jacobe 2008 12 Grundmann 2004 14 Swinkels 2004 15 Dawe 2011 11 Dawe 2011 11 Bagel 2009
17

N
34 51 29 55 54 58 55 50 7 10 78 70 70 12 Fixed dose 30 70 0 10 20 30 40 50 60 70 Dose according to MED

Dose according to skin type

Mahajan 2010 16 Dawe 2011 11

Number of sessions needed to achieve clearance*

* When available, median and interquartile interval or mean and 95% condence interval. N: number of patients enrolled in each arm of the study MED: minimal erythemal dose

Figure 3 Number of NB-UVB sessions needed to clear psoriasis according to initial dose determination.

needed for clearance with NB-UVB seemed to be about 25 (Fig. 3). Due to the data quality issues, we were not able to estimate the cumulative UVB dose needed for clearing psoriasis. The global efcacy of NB-UVB as dened by psoriasis clearance is good regardless of the protocol used, with a clearance rate of about 70% (Fig. 2).
Efcacy of PUVA in treating chronic plaque psoriasis

By examining the inuence of the initial UVA dose on clearance rate, we found nine studies using MPD,2427,3034 four

using UVA dose according to skin type 2426,29 and one a xed UVA dose for all patients.28 The efcacy of PUVA in psoriasis according to initial dosing strategy, in the eight studies reporting clearance rate is shown in Fig. 4. The initial dosing strategy did not appear to inuence the efcacy of PUVA. Similarly to NB-UVB phototherapy studies, the condence intervals are often large, owing to the small number of patients in most studies. Three studies directly compared MPD and dosing according to skin type for optimal initial dose of PUVA selection.2426

JEADV 2012, 26 (Suppl. 3), 1121

2012 The Authors Journal of the European Academy of Dermatology and Venereology 2012 European Academy of Dermatology and Venereology

Efcacy of Psoralen UV-A therapy vs. Narrowband UV-B therapy

17

Study (ref)

Buckley 1995a 25 Yones 2006a 32 Gordon 1999a 34 Markham 2003a 33 Kirby 1999a 26 Collins 1996a 24 Sivanesan 2009b 29 Kirby 1999a 26 Collins 1996a 24 Buckley 1995a 25 Lahauranta 1981b 28

42 45 49 25 44 37 30 41 37 41 20 0

Dose according to MED

Dose according to skin type

Fixed dose 10 20 30 40 50 60 70 Clearance rate (%)* 80 90 100

* Clearance was dened as: a clearance (without precision) b PASI75 N: number of patients enrolled in each arm of the study MED: minimal erythemal dose

Figure 4 Clearance rates of PUVA in psoriasis according to initial dose determination.

In the study by Kirby et al.,26 80 patients were treated with PUVA twice weekly, 40 in the skin type group and 40 in the MPD group. Signicantly more patients in the skin type group cleared than in the MPD group: 38 (95%) vs. 27 patients (67.5%) respectively (P < 0.05). Furthermore, the cumulative UVA dose required for clearance was lower in the skin type group {60.7 J cm2 [inter quartile range (IQR) = 10.6221.2] for the skin type group vs. 75.0 J cm2 (IQR = 16.4287.5) for the MPD group, P = 0.08}. Time to clearance was shorter in the MPD group 44.5 days (IQR = 2197) in comparison to the skin type group [66 days (33105), P = 0.04]. Buckley et al.25 compared twice-weekly MPD-PUVA with 3 times weekly skin-type selected PUVA regimens in 83 patients. They did not nd any difference between the two protocols in terms of efcacy or number of sessions needed to reach clearance. However, the median number of days to clear was signicantly higher in the MPD-PUVA group than in the skin type-group (50 days (IQR = 4366) vs. 41 days (IQR = 36-50) respectively, P < 0.05). Nevertheless, the difference in the number of sessions per week between the two arms of the study makes these results difcult to compare.

In an intraindividual controlled, paired comparison study published by Collins et al.,24 there was no signicant difference in efcacy in 37 patients treated twice weekly with a PUVA initial dose and increment based on MPD on one body side while the other body side received a lower UVA dose regimen based on skin type with xed increments. The total number of sessions required to achieve clearance demonstrated a saving of 3 sessions with the MPD regimen (11 vs. 14 sessions) but with a higher cumulative UVA dose (median dose of 62.9 J cm2 on the MPD side vs. 30.5 J cm2 on the skin type side, P < 0.001). A meta-analysis of the three studies comparing initial dosing strategies in terms of psoriasis clearance, did not nd signicant differences between UVA dosing according to skin type and MPD (P = 0.22) (Fig. 5). One intra-individual study compared the effect of the number of weekly PUVA sessions on clearance rate.27 A total of 28 patients were randomized to receive PUVA either twice or three times weekly on each body side. There were no signicant differences in the median percentage of reduction in PASI at the end of the evaluation [92.9% (89.996.1) on the twice weekly side vs. 94.8 (91.896.8) on the thrice weekly side, P = 0.179]. A lower median

Study or subgroup Kirby 1999 26 Buckley 1995 25 Colins 1996 24

Phototype MPD Events Total Events Total Weight 38 37 30 41 41 37 27 35 31 44 42 37 32.6% 32.8% 34.6%

Odds ratio M-H, Random, 95% Cl 7.98 [2.12, 29.94] 1.85 [0.50, 6.87] 0.83 [0.25, 2.76] 2.26 [0.61, 8.36]

Odds ratio M-H, Random, 95% Cl

Total (95% CI) 119 123 100.0% Total events 105 93 Heterogeneity: Tau2 = 0.91; Chi2 = 6.30, df = 2 (P = 0.04); I2 = 68% Test for overall effect: Z = 1.22 (P = 0.22) MPD: Minimal Phototoxic Dose

0.01 0.1 1 10 100 Favours MPD Favours Phototype

Figure 5 Effect of initial UVA dose selection (MPD or skin phototype) on psoriasis clearance with PUVA.

JEADV 2012, 26 (Suppl. 3), 1121

2012 The Authors Journal of the European Academy of Dermatology and Venereology 2012 European Academy of Dermatology and Venereology

18

Archier et al.

Study (ref) Buckley 1995a 25 Yones 2006a 32 Gordon 1999a 34 Markham 2003a 33 Kirby 1999a 26 Collins 1996a 24 Sivanesan 2009b 29 Kirby 1999a 26 Collins 1996a 24 Buckley 1995a 25 Lahauranta 1981b 28

N 42 45 49 25 44 37 30 41 37 41

Dose according to MPD

Dose according to skin type

Fixed dose
20

10

20 30 40 50 60 Number of sessions needed for clearance*

70

* When available, median and interquartile interval or mean and 95% confidence interval. N: number of patients enrolled in each arm of the study MPD: Minimal Phototoxic Dose

Figure 6 Number of PUVA sessions needed for psoriasis clearance.

number of sessions was needed for clearance on the twice weekly side compared with the thrice weekly side with 15 sessions1125 vs. 22 1725 respectively, P < 0.0001). Similarly, the median cumulative UVA dose to obtain clearance was lower with the twice-weekly regimen [142.5 J cm2 (106.1316) vs. 241.4 (169.7366.3), P = 0.001]. Patients with thick psoriasis plaques (ostraceous subtype) showed different behaviour with signicant differences favouring the thrice-weekly regimen but not the number of sessions for clearance or the cumulative doses. Due to limited information about data distribution in original publications it was not possible to estimate with precision and condence the mean number of PUVA sessions required to achieve clearance. As shown in Fig. 6, the average number of PUVA sessions needed for clearance is about 17 with limited differences between UVA dosing strategy (Fig. 6). No reliable information could be extracted from publications concerning the cumulative UVA dose required for clearance. As shown in Fig. 4, PUVA regardless of the protocol used, was shown to produce a psoriasis clearance rate of about 80% (Fig. 4).
Comparative efcacy of PUVA and NB-UVB in chronic plaque psoriasis

Three studies3234 directly compared NB-UVB and PUVA therapies in psoriasis: Yones et al.32 treated 83 patients with twice-weekly NB-UVB or PUVA, starting at 70% of the minimum phototoxic or erythema dose, with increases of 20% in dose per session. Of the 71 patients with skin type I through IV, PUVA achieved clearance in 31 (84%) out of 37 patients compared with 22 (65%) out of 34 patients with NB-UVB (P = 0.02). The median number of treatments to clearance was 17.0 in the PUVA group and 28.5 in the NB-UVB group (P < 0.001). The clearance rate was

signicantly lower in patients with darker skin (24% for skin type V and VI vs. 74% to 75% for patients skin type I to IV). Six months after stopping treatment, 68% of PUVA treated patients were still in remission vs. 35% of NB-UVB treated patients (P = 0.02). In the study by Markham et al.33 54 patients received either 3-times weekly TL-01 or twice-weekly PUVA, based on minimal erythemal or phototoxic doses. All the 45 analyzed patients who completed the study cleared, no matter which phototherapy was used. Patients receiving NB-UVB required a higher number of sessions to achieve clearance with 25.5 sessions (18.032.5) vs. 19 sessions (14.625.0) for patients receiving PUVA (P = 0.03). There was no signicant difference in the number of days needed to clear for patients treated by PUVA or NB-UVB, with 66 days (52.092.6) vs. 67(47.981.7) respectively, P = 0.46) or in the duration of remission. At 6 months, 67% of the patients treated with NB-UVB were still cleared vs. 68% of those treated with PUVA. Gordon et al.34 randomized 100 patients to receive either twiceweekly TL-01 or PUVA. Psoriasis clearance occurred in a higher proportion of patients with PUVA than with NB-UVB (84% vs. 63%, respectively, P = 0.018). The median number of treatments for clearance with PUVA was also signicantly lower than with NB-UVB (16.7 vs. 25.3, respectively, P = 0.001). A total of 35% of PUVA patients remained clear 6 months after stopping treatment vs. 12% of patients treated with NB-UVB (P = 0.002). Figure 7 shows that the psoriasis clearance rate with PUVA based on three comparative studies is signicantly higher as compared to NB-UVB [OR = 2.79 95% CI: (1.405.55)] (Fig. 7). The meta-analysis investigating remission rate at 6 months shows that more psoriasis patients are still cleared at 6 months with PUVA than with NB-UVB [OR = 2.73, 95% CI: (1.186.27)] (Fig. 8).

JEADV 2012, 26 (Suppl. 3), 1121

2012 The Authors Journal of the European Academy of Dermatology and Venereology 2012 European Academy of Dermatology and Venereology

Efcacy of Psoralen UV-A therapy vs. Narrowband UV-B therapy

19

Study or subgroup Markham 2003 33 Yones 2006 32 Gordon 1999 34

PUVA NB-UVB Odds ratio Events Total Events Total Weight M-H, Random, 95% Cl 24 38 41 25 45 49 28 22 32 29 34 51 5.9% 41.3% 52.8% 0.86 [0.05, 14.45] 2.96 [1.02, 8.63] 3.04 [1.18, 7.84] 2.79 [1.40, 5.55]

Odds ratio M-H, Random, 95% Cl

T otal (95% CI) 119 114 100.0% T otal events 103 82 Heterogeneity: T au2 = 0.00; Chi2 = 0.71, df = 2 (P = 0.70); I2 = 0% T est for overall effect: Z = 2.93 (P = 0.003)

0.01 0.1 1 10 100 Favours NB-UVB Favours PUVA

Figure 7 Comparative clearance rates with PUVA vs. NB-UVB in psoriasis.

Study or subgroup Markham 2003 33 Gordon 1999 34 Yones 2006 32

PUVA NB-UVB Odds ratio Events Total Events Total Weight M-H, Random 95% Cl 8 17 31 25 49 45 9 6 15 29 51 43 29.3% 32.8% 37.9% 1.05 [0.33, 3.31] 3.98 [1.42, 11.22] 4.13 [1.70, 10.06] 2.73 [1.19, 6.27]

Odds ratio M-H, Random. 95% Cl

T otal (95% CI) 119 123 100.0% T otal events 56 30 Heterogeneity: T au2 = 0.27; Chi2 = 3.99, df = 2 (P = 0.14); I2 = 50% T est for overall effect: Z = 2.37 (P = 0.02)

0.01 0.1 1 10 100 Favours NB-UVB Favours PUVA

Figure 8 Comparative remission rate at 6 months of PUVA vs. NB-UVB in psoriasis.

Discussion
The aim of the present systematic review was to determine the respective efcacy of PUVA and NB-UVB for treating plaque psoriasis in preparation for evidence-based recommendations for the use of phototherapy in the treatment of plaque psoriasis. Overall, our data conrm that PUVA therapy is more effective than NB-UVB in treating chronic plaque psoriasis. We showed that clearance rate of psoriasis was about 80% with PUVA vs. 70% with NB-UVB. The meta-analysis of the three randomized controlled studies comparing PUVA and NB-UVB was also in favour of PUVA. The consideration of remission rate is also of clinical relevance. The probability of psoriasis remission at 6 months appears to be on average 2.7 times higher with PUVA than with NB-UVB. Our results are consistent with the French,35,36 British,37 and American38 guidelines. The French treatment guidelines on the use of PUVA in psoriasis, published in 2000 by Schmutz al, also mentioned psoriasis clearance in 80 to 90% of the patients.35 The French guidelines on the use of NB-UVB in psoriasis, published in 2009 by Beani et al., concluded that NB-UVB provides a good alternative to PUVA except for patients with large thick psoriasis plaques and patients with skin types from IV to VI. They estimated that 6090% of patients were cleared after 20 40 sessions.36 The British guidelines reviewed NB-UVB and broadband UVB (BB-UVB) efcacy in psoriasis but also in eczema and other skin

diseases. They suggest that NB-UVB is more effective than BBUVB for psoriasis and similar to PUVA in terms of efcacy in treating psoriasis in patients with skin type IIII, with approximately 6380% of patients cleared with a course of NB-UVB.37 The American guidelines, published by Menter et al. in 2010, also concerned PUVA, NB-UVB and BB-UVB. The guidelines did not make recommendation regarding treatment choice.38 Using a systematic literature review approach, we assessed different predictors of efcacy for both phototherapy procedures. Our results suggest that the modalities of determining the initial dose of UV, the increment regimen or the number of session per week do not dramatically inuence clearance rates with PUVA therapy or with NB-UVB. There was no difference in terms of efcacy between the regimen using MED MPD or those using skin type to determine the initial dose. But the paucity of RCT using the skin type to determine the phototherapy protocol contrasts with what is done in our daily practice.35 The American guidelines recommend determining the initial UV dose according to skin type or MED for NB-UVB and according to the patient skin type optionally including formal MED testing, plaque characteristics, and thickness for PUVA.38 The French guidelines on NB-UVB did not compare the effect of using DEM or skin type for determining the initial UVB dose.36 However, the French guidelines on PUVA suggest that protocols based on the skin type should be preferred for determining the initial dose as MPD depends on plasma and

JEADV 2012, 26 (Suppl. 3), 1121

2012 The Authors Journal of the European Academy of Dermatology and Venereology 2012 European Academy of Dermatology and Venereology

20

Archier et al.

cutaneous concentrations of psoralens, which are subject to change from day to day.35 In our review, the incremental regimen doesnt seem to affect clearance rates for either PUVA or NB-UVB phototherapy among psoriatic patients. Our results are in accordance with the French guidelines stating, however, that if the incrementation regimen is too fast, it causes more phototoxic accidents for NB-UVB therapy.36 The British guidelines specify that it also seems that near-erythemal treatment courses are not essential and that suberythemal treatment may be as effective, although this may take longer to achieve clearance. Moreover, if any of the selected studies in our analysis focused on the role of incrementation in PUVA efcacy, the French guidelines suggest that a suberythemal regimen clears psoriasis with lower cumulative doses of UVA.35 Session frequency did not inuence clearance rates. This is at variance with some guidelines strongly advocating for at least three sessions per week for PUVA or NB-UVB.35,38 The number of sessions needed to clear was lower for PUVA (approx 17 sessions) than with NB-UVB (approx 25 sessions). This is in accordance with the American guidelines suggesting that clearance within two weeks may be seen with an average of 1520 treatments with PUVA, and with 2025 treatments with NB-UVB.38 However, other potentially predictive factors such as the initial severity of psoriasis or the anteriority of treatments with phototherapy were not available in enough studies to be analyzed but may also play a role in clearance rate. Our work has several limitations. The review highlights the relative paucity of the literature on the subject, most studies being old and conducted in a small number of subjects. This indicates the results of the systematic review have to be interpreted with caution. Moreover, there was a high level of heterogeneity, particularly concerning the number of sessions per week and the dose selection process. The heterogeneity of studies, their small sample sizes, differences in primary end points, use of non validated end points to determine efcacy, variable treatment regimens and poor statistical analysis, often made it difcult to interpret and to combine the data. Despite theses limitations, both PUVA and NB-UVB appear to be effective treatments for chronic plaque psoriasis. Our results suggest that PUVA tends to clear psoriasis more reliably than NB-UVB, with fewer treatments, and for longer, and should therefore still be used in appropriately selected patients. The choice between the two phototherapy modalities should be inuenced by the respective long-term risk of skin cancer of PUVA and NB-UVB, which is addressed in an accompanying paper.39

References
1 Davidovici BB, Sattar N, Jorg PC et al. Psoriasis and systemic inammatory diseases: potential mechanistic links between skin disease and co-morbid conditions. J Invest Dermatol 2010; 130: 17851796. 2 Parrish JA, Fitzpatrick TB, Tanenbaum L et al. Photochemotherapy of psoriasis with oral methoxsalen and longwave ultraviolet light. N Engl J Med 1974; 291: 12071211.

3 van Weelden H, van der Leun JC. Improving the effectiveness of phototherapy for psoriasis [abstract]. Br J Dermatol 1984; 111: 484. 4 van der Leun JC, vanWeelden H. UV-B phototherapy: principles, radiation sources, regimens. Curr Probl Dermatol 1986; 15: 3951. 5 Paul C, Gallini A, Archier E et al. Evidence-based recommendations on topical treatment and phototherapy of psoriasis: systematic review and expert opinion of a panel of dermatologists. JEADV 2012; 26(Suppl. 3): 110. 6 Evers AW, Kleinpenning MM, Smits T et al. Itch and scratching as predictors of time to clearance of psoriasis with narrow-band ultraviolet B therapy. Br J Dermatol 2009; 161: 542546. 7 Kleinpenning MM, Smits T, Boezeman J et al. Narrowband ultraviolet B therapy in psoriasis: randomized double-blind comparison of high-dose and low-dose irradiation regimens. Br J Dermatol 2009; 161: 13511356. 8 Altiner DD, Ilknur T, Fetil E et al. Comparison of weekly and daily incremental protocols of narrowband ultraviolet B phototherapy for psoriasis. J Eur Acad Dermatol Venereol 2006; 20: 10761080. 9 Cameron H, Dawe RS, Yule S et al. A randomized, observer-blinded trial of twice vs. three times weekly narrowband ultraviolet B phototherapy for chronic plaque psoriasis. Br J Dermatol 2002; 147: 973978. 10 Dawe RS, Wainwright NJ, Cameron H et al. Narrow-band (TL-01) ultraviolet B phototherapy for chronic plaque psoriasis: three times or ve times weekly treatment? Br J Dermatol 1998; 138: 833839. 11 Dawe RS, Cameron HM, Yule S et al. A randomized comparison of methods of selecting narrowband uv-b starting dose to treat chronic psoriasis. Arch Dermatol 2011; 147: 168174. 12 Jacobe H, Wintereld L, Kim F et al. The role of narrowband UV-B plus alefacept combination therapy in the treatment of psoriasis. Arch Dermatol 2008; 144:10671068; author reply 1068-9. 13 Kirke SM, Lowder S, Lloyd JJ et al. A randomized comparison of selective broadband UVB and narrowband UVB in the treatment of psoriasis. J Invest Dermatol 2007; 127: 16411646. 14 Grundmann-Kollmann M, Ludwig R, Zollner TM et al. Narrowband UVB and cream psoralen-UVA combination therapy for plaque-type psoriasis. J Am Acad Dermatol 2004; 50: 734739. 15 Swinkels OQJ, Prins M, Veenhuis RT et al. Effectiveness and side effects of UVB-phototherapy, dithranol inpatient therapy and a care instruction programme of short contact dithranol in moderate to severe psoriasis. Eur J Dermatol 2004; 14: 159165. 16 Mahajan R, Kaur I, Kanwar AJ et al. Methotrexate narrowband UVB phototherapy combination vs. narrowband UVB phototherapy in the treatment of chronic plaque-type psoriasis-a randomized single-blinded placebo-controlled study. J Eur Acad Dermatol Venereol 2010; 24: 595 600. 17 Bagel J. LCD plus NB-UVB reduces time to improvement of psoriasis vs. NB-UVB alone. J Drugs Dermatol 2009; 8: 351357. 18 Selvaag E, Caspersen L, Bech-Thomsen N et al. Optimized UVB treatment of psoriasis: a controlled, left-right comparison trial. J Eur Acad Dermatol Venereol 2000; 14: 1921. 19 Wainwright NJ, Dawe RS, Ferguson J et al. Narrowband ultraviolet B (TL-01) phototherapy for psoriasis: which incremental regimen? Br J Dermatol 1998; 139: 410414. 20 Hofer A, Fink-Puches R, Kerl H et al. Comparison of phototherapy with near vs. far erythemogenic doses of narrow-band ultraviolet B in patients with psoriasis. Br J Dermatol 1998; 138: 96100. aute -Labre ` ze C, Saillour F, Che ne G et al. Saline spa water or com21 Le bined water and UV-B for psoriasis vs. conventional UV-B. Lessons arn randomized study. Arch Dermatol 2001; 137: from the salies de be 10351039. 22 Brands S, Brakman M, Bos JD et al. No additional effect of calcipotriol ointment on low-dose narrow-band UVB phototherapy in psoriasis. J Am Acad Dermatol 1999; 41: 991995. 23 Lowe NJ, Prystowsky JH, Bourget T et al. Acitretin plus UVB therapy for psoriasis. Comparisons with placebo plus UVB and acitretin alone. J Am Acad Dermatol 1991; 24: 591594.

JEADV 2012, 26 (Suppl. 3), 1121

2012 The Authors Journal of the European Academy of Dermatology and Venereology 2012 European Academy of Dermatology and Venereology

Efcacy of Psoralen UV-A therapy vs. Narrowband UV-B therapy

21

24 Collins P, Wainwright NJ, Amorim I. et al. 8-MOP PUVA for psoriasis: a comparison of a minimal phototoxic dose-based regimen with a skintype approach. Br J Dermatol. 1996; 135: 248254. 25 Buckley DA, Healy E, Rogers S et al. A comparison of twice-weekly MPD-PUVA and three times-weekly skin typing-PUVA regimens for the treatment of psoriasis. Br J Dermatol 1995; 133: 417422. 26 Kirby B, Buckley DA, Rogers S et al. Large increments in psoralenultraviolet A (PUVA) therapy are unsuitable for fair-skinned individuals with psoriasis. Br J Dermatol 1999; 140: 661666. 27 Valbuena MC, Hernandez O, Rey M et al. Twice- vs. thrice-weekly MPD PUVA in psoriasis: a randomized-controlled efcacy study. Photodermatol Photoimmunol Photomed 2007; 23: 126129. 28 Lauharanta J, Juvakoski T, Lassus A et al. A clinical evaluation of the effects of an aromatic retinoid (Tigason), combination of retinoid and PUVA, and PUVA alone in severe psoriasis. Br J Dermatol 1981; 104: 325332. 29 Sivanesan SP, Gattu S, Hong J et al. Randomized, double-blind, placebo-controlled evaluation of the efcacy of oral psoralen plus ultraviolet A for the treatment of plaque-type psoriasis using the Psoriasis Area Severity Index score (improvement of 75% or greater) at 12 weeks. J Am Acad Dermatol 2009; 61: 793798. 30 Legat FJ, Hofer A, Quehenberger F et al. Reduction of treatment frequency and UVA dose does not substantially compromise the antipsoriatic effect of oral psoralen-UVA. J Am Acad Dermatol 2004; 51: 746754. 31 Cooper EJ, Herd RM, Priestley GC, Hunter JA. A comparison of bathwater and oral delivery of 8-methoxypsoralen in PUVA therapy for plaque psoriasis. Clin Exp Dermatol 2000; 25: 111114. 32 Yones SS, Palmer RA, Garibaldinos TT et al. Randomized double-blind trial of the treatment of chronic plaque psoriasis: efcacy of psoralen-

33

34

35

36

37

38

39

UV-A therapy vs narrowband UV-B therapy. Arch Dermatol 2006; 142: 836842. Markham T, Rogers S, Collins P et al. Narrowband UV-B (TL-01) phototherapy vs oral 8-methoxypsoralen psoralen-UV-A for the treatment of chronic plaque psoriasis. Arch Dermatol 2003; 139: 325328. Gordon PM, Diffey BL, Matthews JN et al. A randomized comparison of narrow-band TL-01 phototherapy and PUVA photochemotherapy for psoriasis. J Am Acad Dermatol 1999; 41: 728732. Schmutz JL, Jeanmougin M, Martin S et al. Recommandations de la te franc rapie syste misocie aise de photodermatologie pour la PUVAthe que dans le psoriasis vulgaire. Ann Dermatol Venereol 2000; 127: 73 79. rapie UVB a ` spectre e troit dans Beani JC, Jeanmougin M. La photothe conisations de la socie te le psoriasis vulgaire: utilisation pratique et pre franc aise de photodermatologie. Ann Dermatol Venereol 2010; 137: 21 31. Ibbotson SH, Bilsland D, Cox NH et al. An update and guidance on narrowband ultraviolet B phototherapy: a British Photodermatology Group Workshop Report. Br J Dermatol 2004; 151: 283297. Menter A, Korman NJ, Elmets CA et al. Guidelines of care for the management of psoriasis and psoriatic arthritis. Section 5. Guidelines of care for the treatment of psoriasis with phototherapy and photochemotherapy. J Am Acad Dermatol 2010; 62: 114135. E Archier, S Devaux, E Castela et al. Carcinogenic risks of Psoralen UV-A Therapy and Narrowband UV-B Therapy in chronic plaque psoriasis: a systematic literature review. J Eur Acad Dermatol Venereol 2012; 26(Suppl. 3): 2231.

JEADV 2012, 26 (Suppl. 3), 1121

2012 The Authors Journal of the European Academy of Dermatology and Venereology 2012 European Academy of Dermatology and Venereology