Progression of patients through the study. NB-UVB indicates narrowband UV-B; PUVA, oral psoralen followed by irradiation with UV-A. *The median number of treatments in each category is given in parentheses.
Improvement in body surface area affected by vitiligo (BSA-V). The data at each time point refer only to patients who received at least that number of sessions. NB-UVB indicates narrowband UV-B; PUVA, oral psoralen followed by irradiation with UV-A.
Improvement in vitiligo in patients treated with oral psoralen followed by irradiation with UV-A (PUVA) and narrowband UV-B (NB-UVB). A, A patient before therapy. B, Same patient as in A after 48 sessions of PUVA therapy. The repigmented skin is noticeably darker than the patient's unaffected skin. C, A patient before therapy. D, Same patient as in C after 48 sessions of NB-UVB therapy. The repigmented skin is the same color as the patient's unaffected skin.
A, Dermatology Life Quality Index (DLQI) scores before therapy, at the end of therapy, and at the end of follow-up. B, Visual Analog Scale (VAS) scores before therapy, at the end of therapy, and at the end of follow-up. The asterisks indicate extremes; circles, outliers, and the lines (from the bottom upward) minimum value, lower interquartile value, median, upper quartile value, and maximum value. NB-UVB indicates narrowband–UV-B therapy; PUVA, oral psoralen followed by irradiation with UV-A.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Yones SS, Palmer RA, Garibaldinos TM, Hawk JLM. Randomized Double-blind Trial of Treatment of Vitiligo: Efficacy of Psoralen–UV-A Therapy vs Narrowband–UV-B Therapy. Arch Dermatol. 2007;143(5):578–584. doi:10.1001/archderm.143.5.578
To compare the efficacy of oral psoralen–UV-A (PUVA) with that of narrowband–UV-B (NB-UVB) phototherapy in patients with nonsegmental vitiligo.
Double-blind randomized study.
Phototherapy unit in a university hospital.
Fifty-six patients with nonsegmental vitiligo.
Twice-weekly therapy with PUVA or NB-UVB.
Main Outcome Measures
The change in body surface area affected by vitiligo and the color match of repigmented skin compared with unaffected skin were assessed after 48 sessions of therapy, at the end of the therapy course, and 12 months after the end of therapy.
The results in the 25 patients each in the PUVA and NB-UVB groups who began therapy were analyzed. The median number of treatments was 47 in the PUVA-treated group and 97 in the NB-UVB–treated group (P = .03); we suspect this difference was because of the differences in efficacy and adverse effects between the 2 modalities, such that patients in the NB-UVB group wanted a longer course of treatment. At the end of therapy, 16 (64%) of 25 patients in the NB-UVB group showed greater than 50% improvement in body surface area affected compared with 9 (36%) of 25 patients in the PUVA group. The color match of the repigmented skin was excellent in all patients in the NB-UVB group but in only 11 (44%) of those in the PUVA group (P<.001). In patients who completed 48 sessions, the improvement in body surface area affected by vitiligo was greater with NB-UVB therapy than with PUVA therapy (P = .007). Twelve months after the cessation of therapy, the superiority of NB-UVB tended to be maintained.
In the treatment of nonsegmental vitiligo, NB-UVB therapy is superior to oral PUVA therapy.
The psychosocial effect of vitiligo should not be underestimated.1 Several treatments have been advocated, including topical therapy with potent topical corticosteroids, calcipotriol, or tacrolimus; pseudocatalase therapy; melanocyte transplantation; skin grafting; cosmetic camouflage or self-tanning preparations; and psychological therapy, but these are often unsatisfactory.2 The combination of treatment with psoralen followed by irradiation with UV-A (PUVA) is a well-established treatment for nonsegmental vitiligo, but it has many disadvantages. In the past decade, there have been reports of good efficacy using narrowband UV-B therapy (NB-UVB; 311-313 nm, TL-01 lamp, Koninklijke Philips Electronics NV, Amsterdam, the Netherlands) to treat the condition.3-18 To our knowledge, we report the first double-blind randomized trial of PUVA therapy using oral psoralen vs NB-UVB therapy for vitiligo.
The trial was approved by the local research ethics committee, and all participants gave informed consent. Sample size calculations suggested that for 80% power to detect a difference of at least 20 percentage points in the improvement in surface area affected (measured as percentage of the pretreatment surface area affected), using nonparametric analysis at the 5% significance level, 44 patients with vitiligo were needed. Therefore, to allow for dropouts, 56 patients with vitiligo were recruited through the phototherapy clinic of St John's Institute of Dermatology, London, United Kingdom, between April 1, 2002, and January 31, 2004; the study was terminated in September 30, 2005.
The inclusion criterion was nonsegmental vitiligo affecting 2% to 70% of the body surface area. Exclusion criteria included age younger than 18 years or older than 70 years, previous skin malignancy, previous failure of intolerance to photochemotherapy, more than 100 sessions of photochemotherapy in the patient's lifetime, treatment for vitiligo within the last 3 months (phototherapy, systemic therapy, or topical therapy with corticosteroid agents, vitamin D analogues, or tacrolimus), pregnancy, lactation, renal or hepatic disease, lupus erythematosus, a history of photosensitivity, or administration of a drug known frequently to cause photosensitization.
After enrollment in the study, patients were randomly allocated to receive either PUVA or NB-UVB therapy by means of a sequentially numbered list held in the pharmacy. The dermatologist conducting the assessments (S.S.Y.) was blinded to the treatment allocations; nursing and pharmacy staff conducting the treatment were necessarily aware of them. To ensure full patient blinding, the phototherapy cabinet labels were concealed, and 2 hours before treatment, all patients ingested identical-appearing tablets: placebo in the NB-UVB group and 8-methoxypsoralen (8-MOP; 10-mg Puvasoralen tablets; Crawford Pharmaceuticals Ltd, Milton Keynes, United Kingdom) in the PUVA group. The dose of 8-MOP was determined according to patient body surface area, namely, 25 mg/m2, and ranged from 30 to 60 mg. All patients were advised that nausea from the medication was a possibility. During the course of the study, patients intolerant of 8-MOP because of nausea instead were given identical-appearing 5-methoxypsoralen (5-MOP) tablets (20-mg Pentaderm tablets; Crawford Pharmaceuticals Ltd) at a dose of 50 mg/m2 (range, 60-80 mg) 3 hours before phototherapy.
Therapy with PUVA was administered in a Waldmann 6002 cabin (Herbert Waldmann GmbH, Villingen-Schwenningen, Germany) containing 40 Waldmann 100-W UV-A fluorescent tubes. The UV-A irradiance at the surface of the patient's skin was approximately 17.4 mW/cm2. The NB-UVB treatments were administered in a Waldmann UV5000 cabinet containing 24 Philips 100-W NB-UVB fluorescent tubes emitting predominantly in the wavelength range of 311 to 313 nm. The UV-B irradiance in the cabin at the surface of the patient's skin was typically between 7 and 8 mW/cm2. The exact irradiance of both sources was checked monthly.
Patients were treated twice weekly, initially with an irradiation dose of 0.1 J/cm2 (NB-UVB group) or 0.5 J/cm2 (PUVA group), followed by 20% increments (NB-UVB group) or 0.25 J/cm2 (PUVA group) at each visit, if tolerated. Doses were adjusted according to the maximum erythema occurring since the previous session, determined by patient self-report and by physical examination. Grade 1 erythema (just perceptible) resulted in repetition of the same dose. In patients with grade 2 erythema (easily perceptible but causing no symptoms or only mild discomfort), treatment was discontinued until resolution of symptoms and emollients were applied frequently; subsequently, the previous dose (not resulting in erythema) was used. After an episode of erythema, NB-UVB increments were reduced by moving 1 point down a scale, from 20% to 10% to 5% to 0%. Corresponding figures for PUVA were 0.25 J/cm2 to 0.1 J/cm2 to 0.05 J/cm2 to 0 J/cm2. After reduction to an increment of zero, if patients remained free of erythema for several sessions, a small increment was reintroduced. Episodes of erythema localized to small areas, such as the face, were not included and were managed by shielding with clothing or a broad-spectrum sunscreen during exposure until resolution of symptoms and then for part of subsequent exposures. There were no erythemal episodes of severity greater than grade 2. No further increments were applied once a dose of 2 J/cm2 (NB-UVB) or 5 J/cm2 (PUVA) was reached.
All patients were requested to apply aqueous cream (30% emulsifying ointment, 1% phenoxyethanol) twice daily and to use daily a moisturizing bath additive of an active emollient containing 13% isopropyl myristate and 37.8% light liquid paraffin (Hydromol; Ferndale Pharmaceuticals Ltd, Thorp Arc Estate, United Kingdom) throughout therapy and follow-up to standardize patient washing regimens and minimize any tendency for vitiligo progression as a result of the Koebner phenomenon. For 12 hours after treatment, all patients wore eye protection because of oral PUVA therapy. During treatment sessions, unaffected skin was covered insofar as practically possible with clothing and male patients wore genital protection.
All patients were assessed immediately before the commencement of treatment and subsequently after every 16 sessions, and photographs were taken on each occasion. All assessments were made by 1 blinded investigator (S.S.Y.) and consisted of estimation of the percentage of body surface area with vitiligo (BSA-V) in a darkened room using a Wood's lamp. The first assessment was made using the rule of nines or, for small areas, by tracing affected areas and comparing the tracings with diagrams of size corresponding to known percentages (planimetry). At subsequent assessments, comparison was made with the baseline photograph to determine the percentage improvement in BSA-V.
In addition, at each visit, patients completed the Dermatology Life Quality Index (DLQI)19 and a visual analog scale (VAS). The VAS was a visual scale of numbers from 0 to 10 prefaced with the question, “At the moment, how would you grade or rate your vitiligo on a scale of 0 to 10? Please circle as appropriate between 0 (no vitiligo) and 10 (the worst vitiligo you have ever had).”
At the end of the study, the nature of repigmented areas was examined using photographs, as follows. The color of repigmented areas was compared with that of the patient's unaffected skin and the matching was classified as either excellent or not excellent. The presence of follicular, peripheral, and/or diffuse repigmentation was recorded.
Treatment was terminated in the event of any of the following: complete or almost complete resolution of vitiligo, absence of improvement after 32 treatments or very slow progress or deterioration thereafter, intolerance of therapy necessitating termination, completion of 200 treatments in a patient's lifetime, or request by the patient for termination because of logistical reasons unrelated to the efficacy of treatment or adverse effects. After termination of treatment because of any of these reasons, patients were assessed every 3 months for 1 year.
Patients who enrolled in the study but withdrew before starting treatment were excluded from statistical analysis, whereas all those who commenced therapy were included regardless of withdrawal for any reason thereafter (on-treatment analysis). Analysis was performed using Statistical Package for Social Sciences software (version 11.5; SPSS Inc, Chicago, Ill).
Comparisons of the 2 treatment groups were made using the exact χ2 test for nominal data and the exact Mann-Whitney test for ordinal data. Pretreatment vs posttreatment BSA-V, DLQI, and VAS scores were compared using the exact Wilcoxon signed rank test. Correlations were assessed with the Spearman rank correlation (ρ) coefficient. All tests of significance were 2-sided, and statistical significance was assumed at P<.05.
Fifty-six patients were enrolled; their progression through the study is shown in Figure 1. Six patients withdrew before commencement of treatment and were not considered further. Fifty patients commenced therapy and all were included in the results; their characteristics are shown in Table 1. There were no statistically significant differences in baseline variables between the 2 groups. The median number of treatments in the PUVA group was 47 compared with 97 in the NB-UVB group (z = 2.1; P = .03). Only 1 patient in the PUVA group reached the maximum dose (5 J/cm2) compared with 8 patients in the NB-UVB group (2 J/cm2). Three patients had been previously treated with PUVA; in this study, 2 of these patients were treated with NB-UVB.
Both PUVA and NB-UVB therapy produced a significant improvement in BSA-V (PUVA, z = 3.5, P<.001; NB-UVB, z = 3.9, P<.001; Figure 2). Considering the BSA-V at the end of treatment in all 50 patients, the improvement in BSA-V with NB-UVB tended to be greater than with PUVA, but this was not quite statistically significant (P = .06; Table 2). We also considered the percentage improvement in BSA-V at the end of 48 treatments only in the patients who completed at least 48 sessions (13 in the PUVA group and 21 in the NB-UVB group); NB-UVB showed greater efficacy (z = 2.7; P = .007; Table 2). The improvement in BSA-V was not associated with sex or correlated with age, skin type, or disease duration. Four patients with disease duration longer than 30 years showed improvement greater than 70%.
In the NB-UVB group, 16 patients (64%) had peripheral repigmentation compared with 9 patients (36%) in the PUVA group (χ21 = 3.9, P = .09). Corresponding figures for follicular repigmentation were 22 patients (88%) in the NB-UVB group and 23 patients (92%) in the PUVA group. Diffuse repigmentation was not observed in any patients. All 25 patients in the NB-UVB group and 23 (92%) of 25 patients in the PUVA group had areas of repigmentation, and the assessment of the color match in comparison with the patient's unaffected skin at the end of treatment was excellent for all patients in the NB-UVB group but for only 11 patients (44%) in the PUVA group (χ21 = 19, P<.001). In the remaining 12 patients in the PUVA group, the repigmented areas were noticeably darker than the patients' unaffected skin (Figure 3).
At baseline, BSA-V scores were not correlated with the DLQI scores (rs = −0.01, P = .94). The DLQI scores were significantly lower after both forms of photochemotherapy than before therapy (Figure 4A; z = 4.3, P<.001), with a median of 9 before and 4 afterward, reflecting an improvement in quality of life. These improvements in DLQI were correlated with the percentage improvement in BSA-V (rs = 0.60, P<.001). Patients rated their vitiligo as less severe after therapy, with the median VAS score before therapy of 9 compared with 5 afterward (Figure 4B; z = 4.6, P<.001). These reductions in DLQI and VAS scores did not differ significantly between the NB-UVB and PUVA groups (reduction in DLQI, z = 0.2, P = .8; reduction in VAS, z = 0.8, P = .5).
Twenty-four patients (96%) in the PUVA group but only 17 patients (68%) in the NB-UVB group developed erythema at some stage during their treatment, a statistically significant difference (χ21 = 6.6, P = .02). The median number of erythemal episodes per patient was 9 in the PUVA group and 2 in the NB-UVB group (z = 3.7, P<.001), of which 8 were grade 2. Seven episodes were in patients being treated with PUVA (χ21 = 3.0, P = .2). For all 50 patients, the median number of sessions until the first episode of erythema was 5 in patients in the PUVA group and 21 in patients in the NB-UVB group (z = 5.1, P<.001). One patient in the PUVA group and no patients in the NB-UVB group terminated treatment because of adverse effects. There was no association between the number of erythemal episodes and sex, age, or skin type. Eight patients switched from 8-MOP to 5-MOP during the study because of nausea.
Thirteen patients in the PUVA group and 21 in the NB-UVB group were followed up at 3-month intervals for 12 months; an additional 6 patients in the PUVA group and 4 in the NB-UVB group were followed up for 3 to 9 months before declining further follow-up. Six patients in the PUVA group and none in the NB-UVB group declined all follow-up after the intervention period. A slight but gradual loss of repigmentation induced by the photochemotherapy was noted: 7 patients (28%) in the PUVA group and 3 patients (12%) in the NB-UVB group had more severe vitiligo than before they began treatment; nevertheless, 6 patients (24%) in the PUVA group and 9 patients (36%) in the NB-UVB group maintained greater than 75% improvement in BSA-V (Table 2). For improvement in BSA-V seen from the start of therapy until the last follow-up in the 44 patients, the difference between the 2 modalities was not significant (z = 0.8, P = .4).
At the end of follow-up, the color match in repigmented areas in all patients in the NB-UVB group remained excellent. In patients in the PUVA group, color matching improved slightly since the end of therapy, but was excellent in only 14 (61%) of 23 patients (NB-UVB vs PUVA: χ21 = 12, P<.001).
To our knowledge, this is the first double-blind randomized trial of oral PUVA therapy vs NB-UVB therapy in the treatment of vitiligo. Both methods produced a reduction in BSA-V. At the end of therapy, 16 patients (64%) in the NB-UVB group showed greater than 50% improvement compared with 9 patients (36%) in the PUVA group. The color match of the repigmented skin was excellent in all patients in the NB-UVB group but in only 11 patients (44%) in the PUVA group (P<.001). In patients who completed 48 sessions, the improvement in BSA-V was greater with NB-UVB therapy than with PUVA therapy (P = .007). It was previously found in a study of PUVA therapy for vitiligo that relapse usually occurs within a year of treatment.20 Therefore, we observed patients for up to 12 months after ceasing therapy; at this time, the previous differences between the 2 modalities tended still to be present but were not statistically significant. There were also fewer adverse effects with NB-UVB therapy. We were unable to detect a difference between the modalities insofar as the improvement seen in quality of life (DLQI scores) and patient-assessed disease severity (VAS scores). We did not detect any association between success of treatment and duration of vitiligo.
Perhaps the most dramatic finding was the difference in the nature of repigmentation, which was often markedly darker than the unaffected skin in patients treated with PUVA (Figure 3). This hyperpigmentation has been reported previously,18,20 but, to our knowledge, this is the first time this difference between PUVA and NB-UVB therapy has been confirmed in a double-blind study. At 12 months after treatment, the darker color tended to persist and was cosmetically unsatisfactory. The predominance of follicular repigmentation, with some patients also showing peripheral repigmentation, and an absence of diffuse repigmentation in patients exposed to both treatments is consistent with previous reports.21
There was a higher incidence of erythema in the PUVA group. This difference may have resulted because of a difference in protocols in the 2 groups, namely, an increment of 0.25 J/cm2 with PUVA vs a 20% increment with NB-UVB. The starting dose of PUVA was 0.5 J/cm2, so the initial increment with PUVA was quite high, 50%. However, only 1 patient in the PUVA group withdrew from treatment because of erythema, and it is sometimes claimed that erythema is a prerequisite for maximal PUVA efficacy.22 Eight (32%) of 25 patients switched from 8-MOP to 5-MOP because of severe nausea. It is possible that these patients may have become aware of their treatment allocation, but all patients were told beforehand that nausea was a possibility, and, overall, we believe most patients remained blinded to their treatment throughout the study.
Patients receiving PUVA had significantly fewer sessions than patients who received NB-UVB (median, 47 vs 97 sessions, respectively), although the usual stated reason for any early termination of therapy was related to commuting problems. However, we suspect that within the PUVA group a relative lack of repigmentation, a poor color match within repigmenting areas, and a high incidence of erythema may have combined to dissuade patients from continuing therapy.
The results of previous studies of NB-UVB therapy for vitiligo are shown in the eTable. These studies included patients with various characteristics and reported outcomes in various ways, making comparison between them difficult. However, overall they showed that approximately 50% of patients who continued with treatment for 6 to 12 months had 75% repigmentation or better. Our results suggest somewhat less efficacy than this. Of note, the starting doses of NB-UVB in the reported studies varied considerably, ranging from 75 mJ/cm2,3 to 740 mJ/cm,2,17 although these differences may reflect variations in calibration.23 Other studies used starting doses based on the minimum erythema dose.11,16
Three previous studies have compared NB-UVB therapy with PUVA therapy. The first, an observer-blinded study, suggested that NB-UVB might produce greater repigmentation than gel PUVA.3 Recently, a left-right comparison study of 15 patients treated with NB-UVB vs oral PUVA showed similar outcomes between the 2 modalities17 and a retrospective study suggested greater efficacy for NB-UVB compared with oral PUVA.18
It is unknown whether NB-UVB therapy is more effective if given 3 times vs twice per week. We administered PUVA twice a week rather than more frequently because PUVA-related erythema peaks at approximately 96 hours24; therefore, to enable effective masking of group allocation, we also administered NB-UVB twice per week. With PUVA, there is no convincing evidence to suggest a difference in efficacy between 8-MOP and 5-MOP, so we used 8-MOP because this is standard practice.
Before phototherapy, we could not detect any correlation between DLQI score and BSA-V. However, after the treatment course, the improvement in quality of life was correlated with the percentage improvement in BSA-V, suggesting that, at least in the context of photochemotherapy, quality of life in patients with vitiligo may be related more to recent disease changes than to overall severity of the disease.
In addition to its efficacy and favorable adverse effect profile, NB-UVB therapy offers other advantages over oral PUVA therapy that make it preferable for most patients. In particular, PUVA therapy requires the use of eye protection after treatment sessions, cannot be used in pregnancy, is contraindicated in patients with hepatic impairment or who are taking warfarin or phenytoin, and requires the somewhat inconvenient prior administration of psoralen. Of greatest concern, however, is the potential of PUVA to cause nonmelanoma and perhaps melanoma skin cancer, although in vitiligo this seems to be rare, perhaps because of the tendency to use much lower doses than for treatment of conditions such as psoriasis. Nevertheless, it is important that NB-UVB seems to be considerably safer than PUVA in comparable numbers of treatment sessions,25,26 and for this reason too, at least in part, NB-UVB is far preferable for use in children. Until further data become available, however, we prefer to limit the number of NB-UVB treatments to a lifetime cumulative total of 200 sessions.26
The mechanism of action of all phototherapy in the treatment of vitiligo may very probably involve diminution of the immunological process followed by the stimulation of residual melanocytes, particularly those residing in hair follicles. Our results firmly suggest that, in most patients, the NB-UVB form of this therapy is preferable to PUVA for the treatment of vitiligo.
Correspondence: Sami Sasi Yones, Dip Der, MsC, FCD, Photobiology Unit, Second Floor, St John's Institute of Dermatology, Division of Genetics and Molecular Medicine, Guy’s, King's and St Thomas' School of Medicine, King's College, London SE1 7EH, England (firstname.lastname@example.org).
Financial Disclosure: None reported.
Previous Presentations: This study was presented in part at the 14th Annual Meeting of the European Academy of Dermatology & Venerology; October 14, 2005; London, England; at the 15th Annual Meeting of the Photomedicine Society, March 2, 2006; San Francisco, Calif; and at the Second Congress of the Libyan Society of Dermatology & Venerology; September 16, 1006; Al-beida, Libya.
Accepted for Publication: November 29, 2006.
Author Contributions: Dr Yones had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Yones, Garibaldinos, and Hawk. Acquisition of data: Yones and Hawk. Analysis and interpretation of data: Yones, Palmer, and Hawk. Drafting of the manuscript: Yones and Palmer. Critical revision of the manuscript for important intellectual content: Yones, Palmer, Garibaldinos, and Hawk. Statistical analysis: Yones and Palmer. Administrative, technical, and material support: Yones, Garibaldinos, and Hawk. Study supervision: Hawk.
Funding/Support: Psoralen and placebo tablets were provided by Crawford Pharmaceuticals Ltd, which had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or preparation, review, or approval of the manuscript.
Additional Information: The eTable is available.
Acknowledgment: We thank the staff of the Dowling Day Treatment Unit (in particular, Maria Csazar), the staff of the Skin Therapy Research Unit, Bernard Higgins, PhD; Mohammed Ben-Gashir, PhD; Paul Seed, MSc; Manal Bugossa, Dip, and the Vitiligo Society. We also thank Crawford Pharmaceuticals Ltd for providing the psoralen and placebo tablets.
This article was corrected for error in data on 5/29/2007, prior to publication of the correction in print.