The determination of intracellular triphosphate metabolites of nucleoside analogs used in anti-HIV therapy is very challenging. Despite the well-known sensitivity and selectivity of LC-MS/MS, the measurement of the triphosphate metabolite of zidovudine (AZT-TP) remains difficult because of the interferences induced by endogenous nucleotides triphosphates. We describe a new approach that allows improved determination of AZT-TP simultaneously with AZT-monophosphate (MP). This was obtained, first, by monitoring a transition from the molecular ion of AZT-TP to a minor but very specific product ion. Then, the spiking of samples with a constant amount of AZT-TP allowed the signal to emerge from background, leading to increased sensitivity. Finally, the analytical run time was reduced to less than 10 min. The low limits of quantification were at 150 and 300 fmol per sample for AZT-TP and AZT-MP, respectively. Recoveries were higher than 85%. Inaccuracy and precision were lower than 10% and 15% (17% at the limit of quantification), respectively. The new method offers the possibility of determining simultaneously other nucleotide phosphates, as shown here for d4T-TP (the triphosphate metabolite of another nucleoside analog, stavudine or d4T) and 2'-deoxythymidine-5'-triphosphate or dTTP (the corresponding natural nucleotide triphosphate).

Besides liquid chromatographic (LC)/UV methods adapted to therapeutic drug monitoring, there is still a need for more powerful techniques that can be used for pharmacological research and clinical purposes. We developed an LC method coupled with tandem mass spectrometry (MS/MS) to separate, detect and quantify with high sensitivity the nucleoside analogues used in multitherapies (zidovudine, stavudine, zalcitabine, didanosine, lamivudine and abacavir) in plasma and in the intracellular medium. We worked on two essential issues: (i) the need to use two ionization modes in order to achieve the best sensitivity, which leads to the optimization of the chromatographic separation of drugs detected in the positive ionization mode and drugs detected in the negative ionization mode, and (ii) the need to optimize the extraction step in order to enhance sample recovery. The peripheral blood mononuclear cells were lysed in Tris buffer-MeOH. A clean-up procedure was performed by solid-phase extraction only for plasma samples. The LC separation was carried out on a Zorbax Stable Bond C(18) column followed by MS/MS analysis after electrospray ionization in either the negative or positive mode. The positive ionization mode was applied at the beginning of the run to detect zalcitabine and lamivudine, then the ionization mode was changed to negative for the detection of didanosine, stavudine, internal standard and zidovudine. The calibration range for all the analytes was 0.5-200 ng ml(-1). The recoveries were between 64 and 90%, with coefficients of variation (CVs) lower than 15%. The inaccuracy (bias) was +/-15% with CVs always lower than 12%. The analytes were stable at room temperature and in the extraction solvent for at least 24 h, after storage at -80 degrees C for 3 months, after three freeze-thaw cycles and in the injection solvent after 48 h at 4 degrees C. Together with the measurement of intracellular triphosphorylated metabolites thanks to the powerful plasma and intracellular assay method for intact drugs, it is possible to describe the behaviour of nucleoside analogues against HIV through plasma pharmacokinetics, cell membrane diffusion including drug transport involvement, and also the intracellular metabolism.