Article Text

Download PDFPDF
PTU-042 Ratio of methylated metabolites to thioguanine nucleotides predicts hepatotoxicity and non-response to thiopurine therapy
  1. M A Smith1,
  2. A M Marinaki2,
  3. M Arenas2,
  4. E Escuredo2,
  5. P Irving1,
  6. J D Sanderson1
  1. 1Department of Gastroenterology, Guy's and St Thomas’ NHS Foundation Trust, London, UK
  2. 2Purine Research Laboratory, Guy's and St Thomas’ NHS Foundation Trust, London, UK


Introduction Azathioprine (AZA) and 6-mercaptopurine (6MP) are the immunomodulators of choice for inflammatory bowel disease (IBD). There are two main problems with these therapies: lack of response and toxicity. These problems conflict, making dosage decisions difficult, especially as the complex metabolism of thiopurines is subject to inter-individual variation such that patients on the recommended dose are at risk of either non-response or toxicity. Therapeutic monitoring of thioguanine nucleotides (TGNs), the active metabolic end-product of AZA and 6MP, is thought to predict both response to treatment and toxicity, and is the subject of ongoing research. TGN levels can be used to detect non-adherence, under- and over-dosing, as well as treatment resistance. TGNs are measured alongside methylated metabolite levels (MeMP) and there are previous reports in small cohorts that the ratio of these two metabolites predicts non-response to treatment and hepatotoxicity with a cut off of 11.

Methods MeMP/TGN measurements taken on a series of 190 patients attending IBD clinics at Guy's and St Thomas Hospitals on AZA or 6MP were reviewed. In particular we noted incidences of hepatotoxicity and looked for any association between MeMP/TGN ratio and non-response to treatment. Associations were tested for significance using the 2-sided Fisher's exact test. The relationship between MeMP/TGN ratio and TPMT was analysed using linear regression analysis.

Results Twelve patients were found to be non-compliant with both metabolite levels <10 pmol/8×108 RBC, and were excluded from further analysis. Eight patients developed abnormal LFTs that were thought to relate to thiopurine treatment. One was diagnosed with genetic haemochromatosis and was excluded from further analysis. Of the remaining seven patients, six had MeMP/TGN ratios ≥11, (compared to 15 of the 147 without hepatotoxicity) p<0.0001. A MeMP/TGN ratio of ≥11 was also associated with non-response (104/134 responders with ratio <11 vs 8/19 with ratio ≥11, p=0.001). The MeMP/TGN ratio did not correlate with TPMT activity (r2=0.02).

Conclusion Our results confirm in a large cohort, that the ratio of MeMP/TGN is a useful predictor of non-response to thiopurines and the development of thiopurine-related hepatotoxicity. This is particularly important, since by adding allopurinol to thiopurines at 25–50% of usual target dose, both these problems can be easily circumvented without losing thiopurines as a treatment option. The lack of relationship to TPMT is interesting and could relate to dosing differences as reported in other cohorts.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.