Comparison of analyte identification criteria and other aspects in triple quadrupole tandem mass spectrometry: Case study using UHPLC-MS/MS for regulatory analysis of veterinary drug residues in liquid and powdered eggs.

Ultrahigh-performance liquid chromatography (UHPLC) coupled with triple quadrupole tandem mass spectrometry (MS/MS) is one of the most powerful tools for the multiclass, multiresidue analysis of veterinary drugs, pesticides, mycotoxins, and other chemical contaminants in foods and other sample types. Until approximately 2010, commercial MS/MS instruments using multiple reaction monitoring (MRM) were generally limited to minimum dwell (and inter-dwell) times of 10 ms per ion transition. To achieve the needed accuracy and detection limits for hundreds of targeted analytes, older UHPLC-MS/MS methods typically acquired only two ion transitions per analyte (yielding only one ion ratio for qualitative identification purposes), which is still the norm despite technological advancements. Newer instruments permit as little as 1 ms (inter-)dwell times to afford monitoring of more MRMs/analyte with minimal sacrifices in accuracy and sensitivity. In this study, quantification and identification were assessed in the validation of 169 veterinary drugs in liquid and powdered eggs. Quantitatively, an "extract-and-inject" sample preparation method yielded acceptable 70-120% recoveries and < 25% RSD for 139-141 (82-83%) of the 169 diverse drug analytes spiked into powdered and liquid eggs, respectively, at three levels of regulatory interest. Qualitatively, rates of false positives and negatives were compared when applying three different regulatory identification criteria in which two or three MRMs/drug were used in each case. Independent of the identification criteria, rates of false positives remained <10% for 95-99% of the drugs whether 2 or 3 ions were monitored, but the percent of drugs with >10% false negatives decreased from 25-45 to 10-12% when using 2 vs. 3 MRMs/analyte, respectively. Use of a concentration threshold at 10% of the regulatory level as an identification criterion was also very useful to reduce rates of false positives independent of ion ratios. Based on these results, monitoring >2 ion transitions per analyte is advised when using MS/MS for analysis, independent of SANTE/12682/2019, FDA/USDA, or 2002/657/EC identification criteria. (Quant)identification results using all three criteria were similar, but the SANTE criteria were advantageous in their greater simplicity and practical ease of use.