Non-tricyclic antidepressants are often encountered in overdose cases, and for some of them therapeutic monitoring might be warranted [1Hendrix L., e al. Deliberate self-poisoning: characteristics of patients and impact on the emergency department of a large university hospital Emerg Med J 2013 ;  30 : e9

Cliccare qui per andare alla sezione Riferimenti]. Therefore we developed and validated an ultra-performance liquid chromatography-tandem mass spectrometry method for the quantification of 15 common non-tricyclic antidepressants in serum: citalopram, desmethylcitalopram, desmethylfluoxetine, desmethylvenlafaxine, duloxetine, fluoxetine, fluvoxamine, melitracen, mianserin, mirtazapine, paroxetine, reboxetine, sertraline, trazodone and venlafaxine. The selected pharmaceuticals represent (nor) tramadol (added because of the desmethylvenlafaxine interference) and all the non-tricyclic antidepressants available in Belgium.

Serum samples (250μL) were precipitated with methanol and acetonitrile containing 5ng of the internal standards: mirtazapine-D4, paroxetine-D5, reboxetine-D6 and venlafaxine-D6. The supernatant was evaporated at 56°C under a flow of nitrogen and reconstituted in 100μL of 50:50 methanol:water with 2mM ammonium acetate. Five microliters was injected and the target analytes were separated on a 100×2.1mm ACQUITY BEH C18 column (Waters, Zellik, Belgium) using a Waters ACQUITY Ultra-Performance Liquid Chromatography system. The analytes were eluted within 7.5min, using a gradient of 2mM ammonium acetate and 0.1% formic acid in water and in methanol at a flow rate of 0.4mL/min. Quantification was performed on a Waters triple quadrupole ACQUITY TQD using multiple reaction monitoring (MRM) in positive mode (2 MRMs per analyte). A six-point calibration curve was used to cover a large concentration range from 10ng/mL to 1000ng/mL. The method was validated based on the Scientific Working Group for Forensic Toxicology (SWGTOX) guideline.

No interfering signals from matrix, internal standard or tricyclic antidepressant drugs were found. Limits of quantification and limits of detection ranged from 10 to 85ng/mL and 2.0 to 28ng/mL, respectively. Ion suppression from matrix effects varied from 36 to 108% for the compounds and 47 to 102% for the internal standards. Process efficiency varied from 22 to 106%. Intra- and interassay imprecision varied from 3.5 to 12.3% and 3.1 to 12.3%, respectively. The bias of the assay was lower than 15% for all the compounds, except the lowest concentrations of desmethylcitalopram, tramadol and trazodone (bias<20%). No carryover was observed.

Good validation performance was obtained thanks to the use of selected deuterated internal standards that adjusted for all of the compounds, to compensate for the global method variability. This method is therefore suitable for both therapeutic drug monitoring and quantification in suspected overdose cases.