Background: Several retrospective studies have found an increased risk of non-melanoma skin cancer (NMSC) in patients with inflammatory bowel disease (IBD) on thiopurines (TP). Other demographic and behavioural risk factors are well established for skin cancer development in the general population, however, these have not been controlled for in the IBD studies. Our study compared the influence of IBD patients' demographic and behavioural characteristics with that of thiopurine therapy on their NMSC prevalence.
Methods: IBD patients were recruited from 4 specialist centres in Australia (Brisbane latitude 27°25'S, Perth latitude 31°57'S and Adelaide latitude 34°52'S) and New Zealand (Christchurch latitude 43°33'S). Patients completed a comprehensive validated questionnaire. Data pertaining to age, ethnicity, skin colour, freckling, methods of sun-protection and number of sunburns in childhood, dose and duration of TP therapy, as well as corresponding metabolite levels were collated. Skin cancer prevalence was documented via self-reporting. Multiple logistic regression was performed.
Results: 691 patients with IBD were included in this cross-sectional study with 62 (9%) patients reporting NMSC development. The median age was 49 years old and 56% patients were women. The majority of patients (81%) stated exposure to a thiopurine medication. The rate of thiopurine exposure was similar in patients who developed skin cancer and those without skin cancer (92 vs. 89%, p=0.3). There was no significant association between NMSC and TP dose or 6-thioguanine nucleotides (6-TGN) levels. Subtropical Brisbane has the highest UV exposure index compared to the other three sites. In multivariate models, 4 factors were independently and significantly associated with NMSC; age per 1 year increase (OR 1.05; 95% CI 1.03–1.07), residing in Brisbane vs. Christchurch (OR 3.3; 95% CI 1.6–6.8), never staying in the shade vs. staying in the shade ≥50% of the time (OR 3.8; 95% CI 1.4–10.5) and having a skin type that never tanned vs. other skin types (OR 6.9; 95% CI 2.9–16.0). Thiopurine duration >3 years vs thiopurine duration <3 months (OR 1.5, p=0.26) and never using sunscreen (OR 1.03, p=0.95) did not reach statistical significance.
Conclusions: Our study demonstrated the significant impact of traditional risk factors on skin cancer development in an IBD population, but failed to find an association between TP dose, or metabolite levels, and NMSC. This suggests that traditional risk factors play a substantial role in NMSC development irrespective of TP exposure. Future studies investigating the role of TPs in skin cancer development need to control for traditional risk factors.