An automated multi-analyte screening method for the identification and quantification of 92 drugs and metabolites based on on-line solid-phase extraction-high-performance liquid chromatography-diode array detection technique was developed and successfully validated. In addition, a database with 870 entries including UV-spectra, relative/retention times and response factors of toxicologically relevant compounds was created. Plasma samples (0.2 mL) were treated with methanol, diluted with buffer and on-line extracted (Strata X, 20 ×2 mm, 25 µm) at pH 9. Analytical separation was carried out on a Gemini NX column (150 ×4.6 mm, 3 µm) using gradient elution with acetonitrile-water (90:10,v/v) and 0.05 m potassium dihydrogen phosphate buffer (pH 2.3). Linear calibration curves with correlation coefficients ≥0.9950 were obtained for 78 analytes. As an additional benefit, the newly developed method allows the quantification of 42 analytes (e.g. antidepressants, neuroleptics and anticonvulsants) in a concentration range suitable for therapeutic drug monitoring. Limits of quantitation ranged from 0.02 mg/L (chlordiazepoxide) to 3.4 mg/L (mexiletine). Inter- and intra-day precisions of quality control samples (low/high) were better than 15% (zolpidem) and accuracy (bias) ranged from -11% (opipramol, venlafaxine) to 11% (venlafaxine, trazodone). Tests for carry-over and sample stability under different storage conditions were also performed and stability was adequate. Four cases of poisoning analysis are presented.

A multi-analyte screening method for the quantification of 50 acidic/neutral drugs in human plasma based on on-line solid-phase extraction (SPE)-HPLC with photodiode array detection (DAD) was developed, validated and applied for clinical investigation. Acetone and methanol for protein precipitation, three different SPE materials (two electro-neutral, one strong anion-exchange, one weak cation-exchange) for on-line extraction, five HPLC-columns [one C18 (GeminiNX), two phenyl-hexyl (Gemini C6 -Phenyl, Kinetex Phenyl-Hexyl) and two pentafluorophenyl (LunaPFP(2), KinetexPFP)] for analytical separation were tested. For sample pre-treatment, acetone in the ratio 1:2 (plasma:acetone) showed a better baseline and fewer matrix peaks in the chromatogram than methanol. Only the strong anion-exchanger SPE cartridge (StrataX-A, pH 6) allowed the extraction of salicylic acid. Analytical separation was carried out on a Gemini C6 -Phenyl column (150 × 4.6 mm, 3 µm) using gradient elution with acetonitrile-water 90:10 (v/v) and phosphate buffer (pH 2.3). Linear calibration curves with correlation coefficients r ≥ 0.9950/0.9910 were obtained for 46/four analytes. Additionally, this method allows the quantification of 23 analytes for therapeutic drug monitoring. Limits of quantitation ranged from 0.1 (amobarbital) to 23 mg/L (salicylic acid). Inter-/intra-day precisions of quality control samples (low/high) were better than 13% and accuracy (bias) ranged from -14 to 10%. A computer-assisted database was created for automated detection of 223 analytes of toxicological interests. Four cases of multi-drug intoxications are presented.

BackgroundAccurate assessment of kidney function is needed for a variety of clinical indications and for research. The measurement of the serum clearance of iohexol has emerged as a feasible method to reach this objective. We report the analytical validation and clinical application of a new high-performance liquid chromatography (HPLC) – tandem mass spectrometry (MS/MS) assay to quantify iohexol in human serum. Specificity was enhanced due to the use of method specific acceptance limits for relative ion (RI) intensities.MethodsThe internal standard ioversol was added to 50 μL serum prior to protein precipitation with methanol. Linear gradient elution was performed on a Waters Oasis® HLB column. Three transitions for both iohexol and ioversol were monitored allowing calculation of RIs. Measurements acquired during method validation were used as a training set to establish stricter acceptance criteria for RIs which were then tested retrospectively on clinical routine measurements (86 measurements) and on mathematically simulated interferences.ResultsThe method was linear between 5.0 μg/mL (lower limit of quantification [LLOQ]) and 100.3 μg/mL iohexol. Intraday and interday imprecision were ≤2.6% and ≤3.2%, respectively. Bias was −1.6% to 1.5%. All validation criteria were met, including selectivity, recovery, extraction efficiency and matrix effects. Retrospectively acceptance limits for RIs could be narrowed to ±4 relative standard deviations of the corresponding RIs in the training set. The new limits resulted in an enhanced sensitivity for the simulated interferences.ConclusionsCriteria for validation were met and the assay is now used in our clinical routine diagnostics and in research.