Improved HPLC methodology for monitoring thiopurine metabolites in patients on thiopurine therapy

Standardization of thiopurine metabolite testing is currently lacking. This paper presents in-depth methodological analysis and optimization of two currently available HPLC procedures (Lennard–Singleton [J. Chromatogr. 583 (1992) 83] and Dervieux–Boulieu [Clin. Chem. 44 (1998) 551]) to improve precision, turn-around time, ruggedness, and cost effectiveness. Design and methods: Reversed-phase chromatography with UV detection was performed on a Waters HPLC system. The two protocols were improved with regards to internal standardization (IS), chromatographic conditions, as well as reagent preparation, storage, and use. 6-Thioguanine nucleotides (6-TGNs) were analyzed by our optimized techniques in samples from patients on thiopurine therapy (n = 24) and the results were compared. Results: 6-Mercaptopurine (6-MP) was an ideal internal standard in either procedure. Isocratic elution with 5% acetonitrile (ACN) in 20 mmol/l phosphate buffer pH 2.5 allowed for minimal background interference in both protocols. 6-Thioguanine, 6-mercaptopurine, and 6-methylmercaptopurine (6-MMP) eluted at around 4, 5, and 6 min, respectively. Dithiothreitol (DTT) was critical only during the acid hydrolysis step. Less mercury-containing waste was generated in the Lennard–Singleton procedure. With our optimized protocols recovery of 6-TGNs was on average 1.38-fold higher in the Dervieux–Boulieu method over a range of 10–678 pmol/8 × 108 RBC and no interfering peaks hindered analysis. Specific extraction of thiopurines before their analysis as per Lennard–Singleton procedure may be redundant. Conclusions: We improved the quality and cost effectiveness of two known procedures for thiopurine metabolite assay. Through common chromatographic conditions and internal standardization, future comparison studies are now facilitated a great deal. The less tedious Dervieux–Boulieu procedure for routine thiopurine metabolite testing is warranted.