Potential and limitation of retrospective HRMS based data analysis: ?Have meat-producing animals been exposed to illegal growth promotors such as SARMs??

Targeted strategies using gas or liquid chromatography coupled to triple quadrupole mass analyzers are mostly used in food safety control laboratories. However, when a new chemical hazard emerges, the target list needs to be updated and thus the should be reanalyzed. Under this scenario, the constant emergence of compounds that can illegally be used as growth promotors in cattle, such as selective androgen receptor modulators (SARM), and the phasing out of older SARMs, would require routine control laboratories (RCL) to constantly re-develop and re-validate their once developed targeted methods. Nevertheless, high-resolution mass spectrometry (HRMS) is not limited to analyzing samples using only a predefined targeted list and therefore the implementation of untargeted HRMS methods in RCLs to detect SARMs would enhance the laboratories throughput. In this study, two HRMS profile databases of cattle urine samples gathered over the past years and collected by two independent laboratories, were retrospectively analyzed to find out if a new SARM has been used. This study assesses different retrospective screening approaches based on the Schymanski identification confidence levels. A screening purely based on identification confidence levels 4 and 5 does not permit reliable detection of any exogenous trace compound that does not belong to a certain food or environmental matrix sample due to the high number of false detects. Therefore, it is shown that the availability of a physical reference substance to increase the identification confidence up to level 1 is essential to discriminate a suspected detection between a false positive detection or a confirmed finding.