Direct stereoselective assay of fluoxetine and norfluoxetine enantiomers in human plasma or serum by two-dimensional gas–liquid chromatography with nitrogen–phosphorus selective detection

A method was developed and validated for the direct enantioselective assay of fluoxetine and norfluoxetine in human plasma or serum by two-dimensional capillary gas–liquid chromatography (GC). A Rtx-1 fused-silica capillary (15 m×0.25 mm I.D., 1.0 μm film thickness) and a hydrodex-β-6-TBDM fused-silica capillary (25 m×0.25 mm I.D., 0.25 μm film thickness) were used. A three-step liquid–liquid extraction was used for sample preparation with fluvoxamine and nisoxetine as internal standards. The method provided linear calibration between about 5 and 250 ng/ml for (R)- and (S)-fluoxetine as well as 15 and 250 ng/ml for (R)- and (S)-norfluoxetine. The limits of detection were about 1.5 and 6 ng/ml, respectively. Intra-day precision (coefficient of variation) was estimated as being between 5.4 and 12.7% at plasma levels of 25, 100 and 200 ng/ml for the four enantiomers. Inter-day precision was between 5.3 and 9.1% at 100 ng/ml. The enantioselective separation of some racemic psychopharmaceuticals was tested with various cyclodextrin GC-capillaries. Advantages and disadvantages of direct enantioselective GC are discussed for the assay of racemic psychopharmaceuticals. Samples from a patient who was treated with racemic fluoxetine were measured. In agreement with literature, plasma levels of the (R)-enantiomers of fluoxetine and norfluoxetine were considerably decreased in comparison to the (S)-enantiomers.