Background: Thiopurines, like azathioprine (AZA) and 6-mercaptopurine (6-MP), are effective in maintaining remission in Crohn's disease (CD). Due to potential adverse effects and the increased availability of biological therapies, their role might be questioned. We looked at clinical characteristics associated with an increased response to thiopurines. Moreover, ATG16L1 T300A genotyping has recently been suggested to identify patients who will benefit most from thiopurine treatment . Hence we evaluated this single nucleotide polymorphism (SNP) in our cohort.
Methods: Medical records of 230 British CD patients were retrospectively assessed. Response to thiopurines was defined as continued usage up to chart assessment or termination because of prolonged disease remission on thiopurine monotherapy; non-response as requirement for significant additional therapy (>1 course corticosteroid/year or addition of biologics) after 4 months on thiopurine. Patients who stopped thiopurines because of intolerance or immediately started combo therapy were excluded from genetic association analysis. Genotyping data (rs2241880, T300A) were available for 128 patients (UK IBDGC). Association analysis was performed via PLINK (chi-square test). A p-value <0.05 was considered significant.
Results: Most (87.0%) of the 230 included patients (111 men, median age at diagnosis 21 years) were administered AZA, with 23.9% ever receiving 6-MP. 24.8% of all patients had to stop thiopurines due to side effects (6 leukopenia, 5 abnormal LFT's, 6 pancreatitis, 1 lymphoma and 39 patients - intolerance unknown). A response rate of 57.8% was observed in patients who tolerated therapy. No difference in response rates was noticed depending on either disease location, disease behaviour, sex or smoking status. Conversely, absence of perianal disease was significantly associated with response to thiopurines (OR=2.8, p=0.003). The ATG16L1 minor allele, A, was not represented more often in responders compared to non-responders (minor allele frequency 41.8% vs 43.0% respectively, p=0.85). Additionally we could not identify a significantly higher proportion of AA homozygotes in thiopurine responders (14/71 vs 9/34, p=0.63), as identified in a Dutch cohort.
Conclusions: Although thiopurines have to be stopped due to intolerance in approximately one quarter of patients, they can still maintain clinical remission in an important subset. Apart from the absence of perianal disease, we could not identify any clinical parameters which might help to stratify patients towards thiopurine monotherapy. Similarly, we could not validate the previously reported predictive value of ATG16L1 genotyping.
 Wildenberg et al. (2016), The ATG16L1 risk allele associated with Crohn's disease results in a Rac1-dependent defect in dendritic cell migration that is corrected by thiopurines. Mucosal Immunol