The present study evaluates for the first time the determination of 20 hydroxylated polychlorinated biphenyl (OH-PCB) congeners and their glucuronide and sulfate conjugates in urine as a biomarker of exposure to PCBs in humans. Thereby, a fast, sensitive and selective online solid phase extraction (SPE) method coupled to LCMS/ MS was validated for the determination of OH-PCBs in human urine, being previously successfully developed and applied for the separation and quantitation of OH-PCBs in human plasma. The lowest limit of quantification (LLOQ) ranged from0.01 to 0.19 ng mL(-1) and average extraction recoveries from 79 to 125% for all hydroxylated congeners. Within-run precision and between-run precision were between 2 and 17%. Extraction recovery tests were also performed in urine with different creatinine contents (0.52-3.92 g L-1) for an estimation of matrix influences and ranged between 69 and 125%. In order to evaluate the applicability of the method, the study was conducted in three different groups, which were distinctly separated as non-exposed to known sources of PCBs (N= 21), low-to-moderate PCB-exposed individuals (N= 25) and highly occupationally PCB-exposed individuals (N= 25), which included workers of a transformer recycling plant, their relatives and workers of surrounding companies from a German cohort. As part of the biomonitoring program HELPcB (Health Effects in High-Level Exposure to polychlorinated biphenyls), urine and blood samples were collected annually from 2010 to 2014. In this way, OH-PCB elimination profile in urine over time, correlations between OH-PCB levels in human plasma and urine, and associations with their parent compounds in plasma of the studied PCB cohort could be also assessed. Tri-chlorinated OH-PCBs were the predominant congeners in urine with concentrations up to 174 ng mL(-1). High chlorinated OH-PCBs (penta-through hepta-chlorinatedOH-PCBs) were also frequently detected in urine samples from non-exposed and occupationally exposed individuals, although levels were in general very low or lower than LLOQ. (C) 2016 Elsevier Ltd. All rights reserved.
