Randomized Controlled Trial of Topical Corticosteroid and Home-based Narrowband Ultraviolet B for Active and Limited Vitiligo

Results of the Hi-Light Vitiligo Trial

K.S. Thomas; J.M. Batchelor; P. Akram; J.R. Chalmers; R.H. Haines; G.D. Meakin; L. Duley; J.C. Ravenscroft; A. Rogers; T.H. Sach; M. Santer; W. Tan; J. White; M.E. Whitton; H.C. Williams; S.T. Cheung; H. Hamad; A. Wright; J.R. Ingram; N.J. Levell; J.M.R. Goulding; A. Makrygeorgou; A. Bewley; M. Ogboli; J. Stainforth; A. Ferguson; B. Laguda; S. Wahie; R. Ellis; J. Azad; A. Rajasekaran; V. Eleftheriadou; A.A. Montgomery


The British Journal of Dermatology. 2021;184(5):828-839. 

In This Article


Recruitment and Participant Characteristics

Recruitment was from 3 July 2015 to 1 September 2017, with 517 participants randomized (398 adults and 119 children). Primary outcome data were available for 370 (72%) participants (Figure 1). Baseline characteristics were well balanced across treatment groups (Table 1).

Figure 1.

CONSORT flow diagram. Reasons for noncollection of primary outcome data at 9 months were: not assessed in clinic (n = 4), withdrew consent (n = 60), discontinued due to adverse effects (n = 3), lost to follow-up (n = 75) and other (n = 5). These reasons were similarly distributed within each treatment arm. Of those who withdrew consent, 11 stated that this was due to lack of treatment response and 33 due to time burden. Of those lost to follow-up, one stated that this was due to lack of treatment response and two due to time burden. TCS, topical corticosteroid; UVB, ultraviolet B. aPatients could have more than one reason for exclusion


The median percentage of NB-UVB treatment days (actual/allocated) was 81% for TCS, 77% for NB-UVB and 74% for the combination, and for ointment 79% for TCS, 83% for NB-UVB and 77% for combination. Just under half of participants used the treatments for > 75% of the expected duration (Table S2; see Supporting Information). Assuming 100% adherence, and a participant with a skin type requiring dose escalation to the maximum dose in the treatment schedule, we estimate the maximum possible total dose of NB-UVB received over the 9-month treatment period to be 4 mW cm−2 × 822 s × 135 treatment sessions = 443·9 mJ cm−2.

In addition to written and online video training,[32] participants received face-to-face training (mean 70 min) prior to using the treatments at home. For participants using active light devices, the median time taken to administer the treatment was approximately 20 min, including time for set-up, administering the light, and documenting timings and side-effects in the treatment diary.

Difficulties in using the treatments are summarized in Table S2 (see Supporting Information). Burden of treatment was identified as an issue by 42 of 142 (30%) in the TCS group, 38 of 140 (27%) in the NB-UVB group and 36 of 149 (24%) in the combination group, although interpretation is difficult as all three groups used both treatments throughout (either active or dummy/placebo). Overall, NB-UVB treatment was reported to be more burdensome than treatment with TCS. Burden of treatment and side-effects were the most commonly cited difficulties for both groups and were common reasons for discontinuation of treatment, along with lack of treatment response.


At the 9-month visit, investigators reported possible unblinding for 21%, 28% and 27% of participants in the TCS, NB-UVB and combination groups, respectively. More participants reported possible unblinding (39%, 55% and 44% in the TCS, NB-UVB and combination groups, respectively), supporting the need for confirmation of the primary outcome using blinded outcome assessment.

Primary Outcome

The proportions of participants who reported treatment success (a lot less noticeable or no longer noticeable) at 9 months were 20 of 119 (17%) for TCS, 27 of 123 (22%) for NB-UVB and 34 of 128 (27%) for combination treatment (Table 2). Combination treatment was superior to TCS: adjusted between-group difference 10·9% (95% CI 1·0–20·9%; P = 0·032; number needed to treat = 10). NB-UVB alone was not superior to TCS: adjusted between-group difference 5·2% (95% CI − 4·4% to 14·9%; P = 0·29; number needed to treat = 19) (Table 3). The proportions of participants achieving treatment success at each timepoint are shown in Figure S1 (see Supporting Information).

All sensitivity analyses were consistent with the primary analysis. Treatment effects were largest among participants who adhered to the interventions ≥ 75% of the time (Figure S2; see Supporting Information).

There was no evidence that any of the treatments were more effective than others for any of the predefined subgroups (Table S3; see Supporting Information). Post hoc exploration of treatment response by skin type (types I–III vs. types IV–VI) also found no differences between the groups (Table S3).

Secondary Outcomes

Treatment success using digital images, assessed by people with vitiligo who did not participate in the trial, was consistent with the primary analysis but suggested greater treatment effects than trial participants' VNS assessments (Table 4).

Participant-reported treatment success at 9 months (all assessed patches) was lower for patches on the hands and feet than on other body regions. However, the relative effectiveness of the three treatment groups remained similar in different body regions (Figure S3 and Table S4; see Supporting Information).

Most participants had onset of treatment response by 3 months, defined as the target patch having improved or stayed the same (Figure S4; see Supporting Information), with 40% in TCS, 61% in NB-UVB and 60% in the combination group showing improvement in their vitiligo (that is, more than stopped spreading).

Treatment success, defined as ≥ 75% repigmentation, supported the finding that combination treatment was superior to TCS alone, but NB-UVB alone was not superior to TCS: this occurred in four patients (3%) for TCS, nine (8%) for NB-UVB and 18 (15%) for combination. This gives an adjusted odds ratio of 4·62 (95% CI 1·50–14·2) for combination compared with TCS, and 2·22 (95% CI 0·66–7·51) for NB-UVB compared with TCS (Table 5).

Long-term follow-up

The percentages of participants followed up at 12, 15, 18 and 21 months after randomization were 56%, 52%, 47% and 43%, respectively. VNS scores throughout the 21-month study period are shown in Figure S1 (see Supporting Information). During the follow-up phase, > 40% of participants reported loss of treatment response by 21 months, across all groups (Table S5; see Supporting Information). Both generic and vitiligo-specific quality-of-life scores were similar at follow-up across the treatment groups (Table S6; see Supporting Information).


In total 124 (25%) participants reported 206 treatment-related adverse events: 33 events from 24 participants (14%) in the TCS group, 69 events from 48 participants (28%) in the NB-UVB group and 104 from 52 participants (30%) in the combination group (Table 6). There were five serious adverse events reported from five participants, but none was related to a trial intervention.

Details of adverse events of particular interest (grade 3 or 4 erythema and skin thinning) are shown in Table 6. Grade 3 and 4 erythemas constituted the majority of adverse events in the NB-UVB and combination groups, and these erythemas accounted for the higher overall adverse event rates in these groups. Fewer adverse events were reported in children than in adults.