Feasibility of different mass spectrometric techniques and programs for automated metabolite profiling of tramadol in human urine.

The purpose of the study was to determine the advantages of different mass spectrometric instruments and commercially available metabolite identification programs for metabolite profiling. Metabolism of tramadol hydrochloride and the excretion of it and its metabolites into human urine were used as a test case because the metabolism of tramadol is extensive and well known. Accurate mass measurements were carried out with a quadrupole time-of-flight mass spectrometer (Q-TOF) equipped with a LockSpray dual-electrospray ionization source. A triple quadrupole mass spectrometer (QqQ) was applied for full scan, product ion scan, precursor ion scan and neutral loss scan measurements and an ion trap instrument for full scan and product ion measurements. The performance of two metabolite identification programs was tested. The results showed that metabolite programs are time-saving tools but not yet capable of fully automated metabolite profiling. Detection of non-expected metabolites, especially at low concentrations in a complex matrix, is still almost impossible. With low-resolution instruments urine samples proved to be challenging even in a search for expected metabolites. Many false-positive hits were obtained with the automated searching and manual evaluation of the resulting data was required. False positives were avoided by using the higher mass accuracy Q-TOF. Automated programs were useful for constructing product ion methods, but the time-consuming interpretation of mass spectra was done manually. Highquality MS/MS spectra acquired on the QqQ instrument were used for confirmation of the tramadol metabolites. Although the ion trap instrument is of undisputable benefit in MSn the low mass cutoff of the ion trap made the identification of tramadol metabolites difficult. Some previously unreported metabolites of tramadol were found in the tramadol urine sample, and their identification was based solely on LC/MS and LC/MS/MS measurements. Copyright (c) 2006 John Wiley & Sons, Ltd.