Direct Determination of Tetrahydrocortisol and Tetrahydrocortisone in Urine Samples from Patients under Steroid Drugs Treatment by LC-MS-MS

Direct determination of steroid metabolites in urine with minimal sample preparation The amount of tetrahydrocortisol and tetrahydrocortisone excreted in the urine during 24 hours and more particularly their ratio are directly correlable with the enzymatic activity of the 11-&beta-idroxysteroidodehydrogenase. Several methods have been proposed in the literature for their determination. Most of them require tedious and extensive extraction and derivatization steps and make use of gas-chromatographic techniques, including GC-MS. We present here a new approach for the direct determination in urine by LC-MS-MS requiring only a minimal sample treatment. The urine samples have been treated with a &beta-glucuronidase/arylsulfatase mixture from Helix pomatia for the hydrolysis of glucuronate and sulfate adducts, centrifugated and filtered on single use 0.2 &mu nylon membrane filters. The resulting clear solution was diluted 1:10 with water and directly analyzed by HPLC-MS-MS by Selected Reaction Monitoring (349-91 and 349-105 for tetrahydrocortisol; 347-91 and 347-149 for tetrahydrocortisone). HPLC separation was obtained on a C8 2.1x50 mm, 5 &mum particle size column using a gradient of water, methanol and acetonitrile containing 0.1% formic acid. Turbo-Ionspray (TIS), Atmospheric Pressure Chemical Ionization (APCI) and Atmospheric Pressure PhotoIonization (APPI) were tested as ionization techniques. Preliminary experiments pointed out the presence of partially interfering substances present in the matrix. This fact made the development of the chromatographic separation quite challenging. The best results were obtained using a gradient from solvent B (water/methanol 50:50 with formic acid 0.1%) to solvent A (methanol/acetonitrile with formic acid 0.1%) in 5 min. The method makes use of external standards calibration and have been tested on blank and spiked urine samples. TIS and APCI were tested on an Applied Biosystems API 4000 triple quadrupole mass spectrometer, whereas APPI was tested on an Applied Biosystems API 3000 instrument. APCI was tested on both machines to check the performance difference. TIS on the API 4000 revealed so far the best combination, mainly due to a lower noise level. We found out, however, a slightly higher signal intensity using APPI on the API 3000, suggesting that this ionization technique could represent the best choice. We are looking forward to testing APPI on the API 4000, whenever it will be available. A LOD of 0.5 ng/ml for pure standards in water and 2 ng/ml for spiked urine samples was observed. Owing to the 1:10 dilution this corresponds to a 20 ng/ml minimum level in the original urine sample. Matrix has some influence of the ionization efficiency, and this led to a slight underextimation of the spiked samples using TIS and to a slight overextimation using APCI and APPI. To override this problem we are at present considering the possibility to use deuterated internal standards and the preparation of the calibration curve standard in a blank matrix.