To the Editor:

We report a case of a novel thiopurine methyltransferase variant resulting in a misdiagnosis of TPMT deficiency. The patient had been diagnosed with distal ulcerative colitis (UC) on the basis of standard clinical, endoscopic, and histological criteria. Having experienced two flares of his UC, both of which required treatment with corticosteroids, the patient was initiated on azathioprine. TPMT levels were requested, but to avoid delay azathioprine was commenced at a low dose equivalent to 0.5 mg/kg. Blood monitoring was performed fortnightly and with no evidence of leukopenia throughout, the dose of azathioprine was increased to a maintenance dose of 2 mg/kg.

A TPMT level was finally available after the patient had been taking azathioprine for a total of 3 months and a full treatment dose of 200 mg for 6 weeks. TPMT activity was 5 pmol/h/mg Hb (normal activity 26–50, intermediate activity 10–25, very low activity <10), indicating TPMT deficiency and a high risk of severe myelotoxicity.

A further sample was sent for repeat analysis of TPMT phenotype and the dose of azathioprine was reduced to 100 mg once daily (1 mg/kg). A sample was also sent for 6-thioguanine nucleotides (6-TGNs) and methyl 6-mercaptopurine (MMP) levels. The results were available after a further 4 weeks during, which time the full blood count was monitored weekly and the white cell count remained normal. 6-TGNs levels were significantly above the therapeutic range at 1181 pmol/8×108 red blood count (RBC; normal range 200–400 pmol/8×108 RBC) and MMP levels were 0 pmol/8×108 RBC, in keeping with reduced TPMT activity. A second TPMT level was again 5 pmol/h/mg.

Despite the normal white cell counts and the patient remaining in remission, after discussion, and in view of the perceived risk of potential myelotoxicity, azathioprine was discontinued. One month after cessation of azathioprine the patient developed a further flare of his colitis, which failed to respond to oral steroids, intravenous steroids, and an infliximab infusion. Subtotal colectomy was performed and histology from the resection specimen showed fulminant colitis compatible with UC.

In view of the discrepancy between the result of the TPMT phenotype assay and the absence of leukopenia, TPMT genotype analysis was performed. The patient was found to be compound heterozygous for the sequence variant TPMT*3A and a novel mutation, TPMT c.349>A, which results in the amino acid substitution p.117G>R. This variant has been designated TPMT*28 following reports of TPMT*26 and a novel splice site variant recently reported.1, 2

It is generally recommended that individuals intermediate for TPMT activity should be treated with half-dose therapy.3 Although there are limited reports of the safe use of thiopurines at very low doses in individuals with very low TPMT activity,4 in general thiopurine treatment is not recommended in these patients, as severe myelosuppression occurs quickly and has been described as inevitable.3

The correlation between genotypic and phenotypic analysis is, however, not perfect. One analysis found that although all patients genotyped as heterozygous had phenotypic activity in the intermediate range, ≈33% of patients with intermediate activity on phenotype testing were wildtype on genetic analysis.5

Why this patient was able to tolerate azathioprine in total for 4 months and at full dose for 6 weeks requires explanation. This patient was compliant with medication, as evidenced by supratherapeutic 6-TGN levels. A more likely explanation is that TPMT*28 is an unstable enzyme variant. One of the patient's TPMT alleles, TPMT*3A, is known to be associated with very low TPMT activity and severe toxicity to azathioprine therapy. We suggest that the other allele, the novel TPMT*28, is an unstable enzyme variant associated with significant residual TPMT activity in nucleated cells able to continually synthesize TPMT but with unrecordable levels of activity when measured in erythrocytes with the standard TPMT activity assay.

The case described would suggest that in patients already established on therapy subsequently found to have very low TPMT phenotypic activity, TMPT genotyping should be performed to assess for novel or rare variants which might allow thiopurine therapy to be continued with close monitoring. Furthermore, in thiopurine-naive patients with very low TPMT phenotypic activity, genotyping may allow identification of a small but significant cohort of patients where therapy can be considered.


  1. Top of page
  • 1
    Kham SK, Soh CK, Aw DC, et al. TPMT*26 (208F[RIGHTWARDS ARROW]L), a novel mutation detected in a Chinese. Br J Clin Pharmacol. 2009; 68: 120123.
  • 2
    de Beaumais TA, Fakhoury M, Pigneur B, et al. Characterization of a novel TPMT mutation associated with azathioprine-induced myelosuppression. Br J Clin Pharmacol. 2009; 68: 770772.
  • 3
    Sandborn WJ. Rational dosing of azathioprine and 6-mercaptopurine. Gut. 2001; 48: 591592.
  • 4
    Kaskas BA, Louis E, Hindorf U, et al. Safe treatment of thiopurine S-methyltransferase deficient Crohn's disease patients with azathioprine. Gut. 2003; 52: 140142.
  • 5
    Winter J, Gaffney D, Shapiro D, et al. Assessment of thiopurine methyltransferase enzyme activity is superior to genotype in predicting myelosuppression following azathioprine therapy in patients with inflammatory bowel disease. Aliment Pharmacol Ther. 2007; 25: 10691077.

J. Landy MRCP*, N. Bhuva MRCP*, A. Marinaki PhD†, J. Mawdsley MRCP, PhD*, * Gastroenterology Department West Middlesex University Hospital Isleworth, UK, † Purine Research Laboratory, GSTS Pathology Guy's St Thomas' NHS Foundation Trust, London, UK.