Abstract No. 3 Is there a skin cancer risk with narrowband ultraviolet B phototherapy? Preliminary data from the second phase of the Dundee follow-up study

Abstract

We do not know whether narrowband ultraviolet B (NB‐UVB, TL‐01) phototherapy carries any increased risk of skin cancers. Follow‐up of 126 patients treated with NB‐UVB in Germany (Weischer M, Blum A, Eberhard F et al. No evidence for increased skin cancer risk in psoriasis patients treated with broadband or narrowband UVB phototherapy: a first retrospective study. Acta Derm Venereol (Stockh) 2004; 84: 370–4) did not detect any increased skin cancer risk. Our follow‐up study (with linkage to Scottish Cancer Registry data complete to 1998) showed greater than expected (compared with age‐ and sex‐matched Scottish population) numbers of basal cell carcinomas (BCCs) in those treated with NB‐UVB (Man I, Crombie IK, Dawe RS et al. The photocarcinogenic risk of narrowband UVB (TL‐01) phototherapy: early follow‐up data. Br J Dermatol 2005; 152: 755–7). Whether the modestly increased standardized incidence rate (SIR) of 213 (95% confidence interval, CI 102–391) was due to treatment or factors such as ascertainment bias was not clear. Analysis of data from the next phase of this follow‐up study, with linkage to the Scottish Cancer Registry complete to December 2002, is ongoing. The summary phototherapy records of 4050 patients treated with NB‐UVB between 1985 and 2002 were examined. The median duration of follow‐up was 5·8 years (up to 16·7) from end of first NB‐UVB course. We now have 26 633 person‐years of follow‐up data. The most frequent primary diagnoses were: psoriasis (55%), atopic dermatitis (15·4%), polymorphic light eruption (7·7%) and chronic urticaria (3·9%). Indirect standardization, adjusting for age and sex, was used to compare numbers of skin tumours identified with numbers expected in the general Scottish population. We did not find significantly more melanomas [SIR 164 (95% CI 60–357), P = 0·16] nor squamous cell carcinomas [SIR 173 (95% CI 79–329), P = 0·08] than expected. More BCCs than expected were registered in patients who had received NB‐UVB and psoralen plus ultraviolet A [SIR 184 (95% CI 128–255), P = 0·0007] and who had received only NB‐UVB [SIR 211 (95% CI 135–314), P = 0·0007, n = 3082]. So far, these findings are reassuring. A possible explanation for the excess BCCs is a causative association between NB‐UVB exposure and BCC development. Alternatively, more BCCs may have been registered in our treated patients because of careful surveillance leading to more BCCs being detected earlier and treated with modalities leading to a histopathological diagnosis (and, therefore, registration). However, only with more prolonged follow‐up of patients who have received high cumulative numbers of treatments might a definite skin cancer risk become detectable. Therefore, we should remain cautious as we cannot dismiss the possibility that NB‐UVB is carcinogenic in humans.