Background and aims: Not all of the adverse effects to thiopurine therapy can be explained by thiopurine methyltransferase (TPMT) polymorphisms. This study was intended to evaluate the value of TPMT genotype and phenotype measurement during the first year of thiopurine therapy.
Methods: Consecutive patients with inflammatory bowel disease (IBD) who were receiving azathioprine or 6-mercaptopurine were followed up for 12 months. TPMT genotypes and phenotypes were examined in patients with IBD before thiopurine therapy and in unrelated healthy volunteers by polymerase chain reaction and high-performance liquid chromatography.
Results: A total of 199 patients and 300 healthy volunteers were included at 2 centers. Forty-seven of the 199 patients (23.62%) exhibited adverse effects during the entire course of thiopurine therapy. Two (1%) patients carrying TPMT*3C developed leucopenia at week 4 of azathioprine treatment. The TPMT*3C had a specificity of 100% (163/163) but a sensitivity of 5.56% (2/36) for predicting leucopenia. The calculated optimal cutoff activity for high TPMT activity and decreased TPMT activity was 4.75 U/mL red blood cells. The risk of leucopenia increased in the decreased TPMT group (odds ratio: 20.25; 95% confidence interval: 2.19-187.17; P = 0.004) and increased more during the initial 3 months of thiopurine therapy (odds ratio: 34.80; 95% confidence interval: 3.71-326.77; P = 0.001). Leucopenia occurred more frequently in the patients cotreated with 5-aminosalicylates than in those not cotreated (32.81% versus 11.11%, respectively, P < 0.001).
Conclusions: The results of this study suggest that the value of TPMT genotyping before thiopurine therapy is limited in Chinese patients with IBD, considering the low sensitivity of predicting leucopenia, and that phenotyping is a more cost-effective tool that can be successfully used in patients. The coadministration of 5-aminosalicylates results in a high frequency of leucopenia in patients receiving azathioprine or 6-mercaptopurine.