Estimating plasma half-lives of dioxin like and non-dioxin like polychlorinated biphenyls after occupational exposure in the German HELPcB cohort.

Abstract Polychlorinated biphenyls (PCBs) are carcinogenic persistent organic pollutants that have been used as mixtures in transformers, electrical capacitors or hydraulic oils in underground mining until their ban in the late 1980s in Germany. The widespread use of PCBs has led to an age-dependent body burden in the general population. In order to determine the human half-lives of different PCB-congeners, we have used data collected between 2010 and 2017 from the prospective investigation of the German HELPcB-cohort with high initial occupational exposure to PCBs from a transformer recycling company. PCBs were quantified in plasma of the study participants in six cross-sectional investigations using gas-chromatography coupled to mass-spectrometry. Applying logistic regression on the individual plasma levels after statistical elimination of outliers, the elimination half-lives of 18 PCB-congeners were determined. Further stratifications were performed using adjustment for blood lipids, calculation of the total amount of PCB in the body and by using a statistical model taking into account the individual age-dependent background burden. The calculated plasma half-lives ranged from 0.8 years for PCB 52 until a maximum of 28.9 years for PCB 189, depending on the model applied. E.g. the total body amount related half-live for PCB 28 is 2.41 years, for PCB 74 it is 12.81 years, for PCB 118 it is 6.65 years and for PCB 153 = 10.75 years. The model with adjustment for age dependent background burden led to shorter half-lives. The analysis revealed structure-related differences in the kinetics for the PCB-congeners examined. Congeners with a chlorine substitution in 2,4,(5)- and 4′-position showed longer half-lives than other isomers with one (or two) free 4-positions. Due to the high number of included cases and repeated measurements, our results provide valid half-live data for a large number of PCB congeners. Using these data, a rough estimation of former occupational exposures from current PCB-levels seems feasible.