Screening for possible interferences from steroidal compounds other than the target analytes (endogenous or exogenous) is well established in LC-MS/MS assay development for steroid quantification in a routine clinical setting. However, interferences from non-steroidal substances have, hitherto, not been explored. After screening more than 150 pharmaceuticals and their metabolites by analyzing commercial quality control samples from TDM analysis kits (Recipe, Chromsystems) with a multisteroid LC-MS/MS assay (protein precipitation followed by HybridSPE filtration, biphenyl column, methanol-water gradient with NH4F additive), we can report the finding of two newly discovered potential interferences from non-steroidal drugs. Antidepressant paroxetine (PX) was identified as an interference to 17-hydroxyprogesterone (17P), and α-hydroxytriazolam (α-OH-TZM)-a major metabolite of benzodiazepine triazolam (TZM)-was identified as an interference to aldosterone (ALDO). Despite different elemental and structural compositions and nominal masses, the M+1 isotopologues of PX and α-OH-TZM produced overlapping signals in ion traces monitored for the respective analytes (m/z 331 → 109/97 and 361→315/343, respectively). PX and TZM are frequently prescribed drugs, and their therapeutic ranges are far exceeding the reference ranges of 17P or ALDO (µmol vs nmol); therefore, these interferences should be considered clinically relevant. Striving for faster multi-analyte methods with high sample turnover, especially in the field of steroid quantification, can limit assay selectivity and specificity. Therefore, supported by the findings of this study, screening for potential interferences in multi-analyte LC-MS/MS method development should not cover only substances of the same class but also include a set of common drugs.
Read full abstract