Introduction The effectiveness of full dose azathioprine (FDA) for inflammatory bowel disease (IBD) is increasingly recognised as poor and often requires the use of anti-TNF therapies.

Current strategies to optimise azathioprine (AZA) includes the use low dose azathioprine with allopurinol (LDAA). However, many groups restrict LDAA to patients classed as shunters/hypermethylators (15% of non-responders) based on thiopurine metabolite (TM) assay profiles. This approach inadvertently denies i) 85% of non-responders ii) those without access to TM iii) those with poor understanding of these complex recommendations to benefit from LDAA.

Methods Records of IBD patients treated with LDAA were retrospectively analysed. AZA dose was reduced to 25–33% of the thiopurine methyl transferase (TPMT) adjusted dose. The response was determined after 6 months and assessed for steroid free clinical remission. A cohort of patients had thiopurine metabolite measurements: i) Non Response (NR) to FDA (n 30) then repeated on LDAA (n 20). ii) Further 70 on LDAA NR (n 20) and full response (FR) (n 50).

Results Four hundred patients received LDAA between 2009 and August 2015. Group 1 (G1)- 1^st line (n190); Group 2 (G2)-switched from FDA to LDAA (n210).

Overall 280 (70%) had a FR (G1 and G2), the rest non-responders (NR): In G1 FR was 79% and G2 65%.

Myelotoxicity occurred in 8 patients; 8 patients on LDAA had hepatotoxicity which resolved by increasing allopurinol to 200 mg (all FR).

The median length of therapy was 38 months.

Metabolites: i) For FDA: 20 complete responders (TGN mean 280) and 30 non responders (TGN mean 260): 6 non adherent (TGN < 150 and low MMP) and 4 shunters tendency (MMP > 5,000 and low TGN). In 20 switched to LDAA (TGN mean for Complete Response 420 and Non Response 440). ii) In Complete Response to LDAA (TGN mean 476), and FDA (TGNs mean 306). In LDAA Non Response (TGN mean 400).

Analysis: Mann-Whitney test. For commercial RBC TGNs: CR vs NR in FDA and LDAA p-value = 0.3132 and 0.9386 respectively. The TGNs increased significantly in both CR and NR switched from FDA to LDAA (p < 0.05).

Conclusion Appropriately dosed LDAA therapy is therapeutically effective, well tolerated and relatively safe compared to FDA. Giving LDAA to patients who are not shunters does not diminish the benefit of LDAA. Avoiding metabolite determined use of LDAA increases the number of patients who can benefit from AZA. The observed Increase in TGNs seen in patients switched to LDAA occurs in both NR and CR.

Disclosure of Interest None Declared