Background Azathioprine (AZA) is a common treatment for connective tissue diseases (CTD). AZA is a pro-drug which is metabolised to active moieties of 6-thioguanine (6-TGN) and 6-methylmercaptopurine (6-MMP). These metabolites, rather than the absolute dose of AZA, are associated with clinical efficacy and toxicity. An idiosyncratic skewed metabolism towards 6-MMP in some patients (“shunters”) increases the risk of hepatotoxicity and treatment failure. Allopurinol can correct such shunts1. Therapeutic protocols using metabolite concentrations have been shown to be cost effective, improve efficacy, and decrease treatment morbidity in IBD1. Only one small study estimated a therapeutic 6-TGN level for SLE patients2. There are scant data on AZA shunters in connective tissue disease3.
Objectives Explore the proportion of AZA therapeutic failure, toxicity, and shunters in CTD patients.
Methods Retrospective, multicentre audit of AZA metabolite levels in CTD patients from 2012–2016. Patient demographics, treatments, disease activity and drug toxicity were also extracted.
Results 61 testing episodes occurred in 34 patients whose mean age was 55 (32–79) years; predominantly female (N=26, 77%), with SLE (N=19, 56%). Active disease was present in 15/61 (25%) episodes. 20/34 (60%) patients were on HCQ + AZA. 25/34 (76%) of patients were either on no or <10mg/day of prednisone. 22/61 (36%) episodes had bone marrow suppression, and 8/61 (13%) had moderate liver function derangement. 12/35 (34%) patients were AZA shunters.
Based on AZA metabolite levels, patients were classified into four categories:
SLE patients were over-represented in the underdosed/non-compliant category (7/9). Patients in underdosed and 6-MMP shunter categories had more active disease (44% & 31%), compared with appropriate and overdosed (17% and 19% p=0.24). Surprisingly, patients who were overdosed by 6TGN levels had the lowest mean AZA dose. 6-MMP shunters had significantly lower median 6TGN levels (147pmol/8x108 RBC), yet toxic MMP levels (9630pmol/8x108 RBC) that was associated with worst LFT derangement prevalence (4/13, 31% vs <13%, p=0.16).
Conclusions AZA dose is poorly predictive of 6-TGN and 6-MMP levels. Overall, one third of CTD patients were shunters, with lower 6-TGN levels and more hepatotoxicity. Only 40% of test episodes were appropriately dosed. Patients with low disease activity are more likely to have adequate 6TGN levels, however, 40% of patients with CTD won't have adequate disease control despite therapeutic 6-TGN levels. AZA metabolites can identify the multiple causes for therapeutic failure (under dosing, shunter, or true non-response), and enable early detection of shunters to avoid liver toxicity.
Croyle, L., & Morand, E. F. (2015). Optimizing the use of existing therapies in lupus. Int J Rheum Dis 18(2), 129–37.
Askanase, AD et al (2009). Use of pharmacogenetics, enzymatic phenotyping, and metabolite monitoring to guide treatment with azathioprine in patients with systemic lupus erythematosus. J Rheum, 36(1), 89–95.
Haines, ML et al (2011). Clinical usefulness of therapeutic drug monitoring of thiopurines in patients with inadequately controlled inflammatory bowel disease. Inflam Bowel Dis, 17(6), 1301–7.
Disclosure of Interest None declared