The determination of the thiopurine S-methyltransferase activity (TPMT; EC 2.1.1.67) has become an important issue during thiopurine therapy due to its known genetic polymorphism resulting in a wide range of TPMT activity. Therefore, the standard thiopurine drug regimen is associated with increased hematopoetic toxicity in patients with low or absent TPMT activity, whereas patients with high activity may be insufficiently treated. However, presently available methods are labour intensive and time consuming and tend towards too high or too low enzyme activity due to their methodological approach. The use of instable substrate solutions (6-MP or 6-TG), organic solvents like dimethyl sulfoxide and too high substrate and co-substrate saturation concentrations contribute to this phenomenon. We therefore, established an optimized and fast isocratic HPLC linked TPMT assay based on the enzymatic methylation of mercaptopurine or thioguanine in RBC lysates with S-adenosyl-l-methionine as methyl donor. Unspecific non-enzymatic methylation was not detectable. The recovery of 6-methyl-mercaptopurine was 97-102%, the intra- and interday variation between 1.0 and 5.0%, respectively. The assay dispenses with a time consuming extraction procedure with organic solvents, a heating step, and a gradient elution and is therefore, favourable for clinical routine application. The TPMT activity was measured in 62 untreated children with acute lymphoblastic leucemia at the time of diagnosis (activity = 34.0+/-10.6 nmol/g Hb/h, range: 11.5-55.4 nmol/g Hb/h) and in 12 adult healthy volunteers (62.8+/-7.7 nmol/g Hb/h, range: 48-82 nmol/g Hb/h) reflecting the wide measurable TPMT activity found in erythrocytes.
