Comparative assessment of screening workflows for characterizing occupational exposure to chemicals in humans: an EU-PARC interlaboratory study
摘要
Biomonitoring of occupational exposure to mixtures of chemicals remains a major challenge, as conventional targeted approaches can only cover a limited number of compounds. Suspect and non-target screening (SS/NTS) using high-resolution mass spectrometry offers a more comprehensive approach, however wider implementation requires robust and comparable analytical strategies with broad chemical coverage. An interlaboratory study was conducted within the European Partnership for the Assessment of Risks from Chemicals (PARC) to evaluate the comparability and complementarity of SS/NTS approaches and to support the development of a standard operating procedure (SOP) for occupational biomonitoring. Urine and plasma samples fortified with 81 occupationally-relevant exposure markers at three concentration levels were distributed to eleven European laboratories from eight Member States and measured via their respective in-house Liquid Chromatography- and Gas Chromatography-High Resolution Mass Spectrometry (LC- and GC-HRMS) workflows. Thirteen different analytical methods were applied: nine to both matrices, three to urine only and one exclusively to plasma samples. Eleven methods enabled detection of more than half of the fortified exposure markers (> 40 compounds) in the standard mixture, whilst eight and seven methods achieved the same for urine and plasma samples, respectively. Despite substantial methodological differences, numerous compounds were shared across Reversed-Phase (RP) and Hydrophilic Interaction Liquid Chromatography (HILIC) methods (57 in urine and 34 in plasma), with RP providing higher overall detection rates (eleven and 29 compounds detected exclusively by RP in urine and plasma, respectively). In addition, four compounds were detected exclusively by GC-HRMS in the biological matrices, demonstrating that combining SS/NTS workflows broadens chemical space coverage. The solvent selected for extraction appears to have an influence on the detection performance, with solvents providing broader polarity coverage, such as mixtures of methanol or acetonitrile with water, generally associated with optimal results, across the different sample preparation techniques employed. Solid phase extraction and deconjugation did not show an improvement in performance in these large-scale SS applications. Based on this interlaboratory study, a minimal SOP is proposed to support SS/NTS implementation in future large-scale human biomonitoring studies, including characterization of occupational cohorts within PARC.
Graphical Abstract