View Report Details

Biomarkers, toxicokinetics and default safety factors

Project Code: T01017

31/03/2004

University of Southampton
Tullberg, S ; Keene, W; Rakkar, P; Toor, M; Renwick, A

A series of studies has been performed in human volunteers and experimental animals to assess the adequacy of the uncertainty factors used in the risk assessment of non-genotoxic chemicals present in human food. A 100-fold uncertainty factor has been used internationally for over 40-years to convert the no observed adverse effect level (NOAEL) in animal studies into a safe intake for humans (the acceptable daily intake or ADI). The studies in this project have used a series of approved food additives dosed at the ADI and at 10-times the ADI in humans and at the NOAEL and at one tenth of the NOAEL in animals. The toxicokinetic data were compared with the proposed default toxicokinetic uncertainty factors of 4.0 for species differences and 3.16 for human variability. In addition, experiments have been performed to investigate the potential for the concentration in saliva to be used as a biomarker of the circulating concentration and therefore the internal dose, and also for in vitro studies to be used to predict in vivo species differences in kinetics.

The data for BHT showed that the kinetic measurements were essentially independent of the dose given, and that the toxicokinetic default uncertainty factors for species differences and human variability were appropriate for this additive.

The plasma concentrations of curcumin were close to or below the limit of quantification in animals and humans, despite the development of an analytical method capable of measuring curcumin in the low pg/ml range. The limited available data did not indicate that the default uncertainty factors for species differences and human variability were inadequate for this food additive.

The plasma concentrations of thiabendazole (TBZ) in humans dosed at the ADI were almost 10,000 times lower than those in rats dosed at the NOAEL, so that the TBZ-specific interspecies adjustment factor would be 0.01 rather than the usual default of 4.0. The difference was highly dependent on the dose administered, because a 10-fold reduction in dose in rats reduced the area under the plasma concentration-time curve (AUC) by 100 times, while a 10-fold increase in dosage in humans increased the AUC by almost 100 times. In consequence, the species difference at dose equivalence (in mg/kg body weight) was only 1-2 fold rather than 0.01 fold. There was higher human variability at 10 times the ADI, which is consistent with variability in the extent of saturation of first-pass metabolism. The default uncertainty factor for human variability would be barely adequate for TBZ at doses higher than the ADI and/or leading to saturation of first pass metabolism.

The plasma concentration-time curves for propyl gallate in rats showed rapid absorption and elimination, with significantly higher concentrations in males than in females. The plasma concentrations in humans dosed at the ADI were close to the limit of quantification, whereas more reliable data were obtained at a dose of 10 times the ADI. The kinetics were essentially linear in both rats and humans at the doses studied. The data indicated that the default uncertainty factors interspecies differences and human variability would be adequate for propyl gallate. Studies with octyl gallate and dodecyl gallate showed the presence of extremely low plasma concentrations, which combined with the low ADI values for these food additives, meant that further in vivo studies were impracticable.

The concentrations of foreign chemicals in saliva are known the represent the non-protein bound concentrations in plasma, and therefore saliva has the potential to act as a non-invasive biomarker of the internal dose. All of the compounds analysed showed extensive plasma protein binding, such that the concentrations in human saliva were close to the limit of detection, even when the additive was dosed at 10-times the ADI. Salivary concentrations of these food additives would not be practicable biomarkers of exposure without a very large, possibly 100-fold, increase in analytical sensitivity. In consequence, it was not possible to investigate the population sub-groups (such as children and pregnant women) as was originally intended, and the studies were replaced by investigations in healthy volunteers at 10-times the ADI which gave significant insights into dose-selection for toxicokinetic comparisons and the analytical requirements for the use of saliva as a biomarker. The results underline the importance of analytical sensitivity and dose selection for the determination of the most appropriate biomarker of internal exposure.

In vitro studies on BHT, propyl gallate and TBZ, using rat and human liver preparations, showed that there were minor species differences in the rates of metabolism and in the estimated Vmax and Km values. The estimated inter-species adjustment factors based on Vmax/Km were 1.17 for BHT, 2.4 for propyl gallate and 0.68 for TBZ, compared with in vivo values of 3.97 for BHT, 2.0 for propyl gallate and 0.85 for TBZ (at dose-equivalence). The in vitro values do not take into account differences in organ blood flow, and such analyses would require the development of a full PBPK model. Interestingly the Km values for BHT and propyl gallate were considerably higher than the in vivo plasma levels, indicating little likelihood of saturation of metabolism. In contrast the plasma levels of TBZ were close to the Km value, and this is consistent with marked non-linearity found with this substrate.

Main File

Some of the files on this site may be in a format that your computer can't read. However, you can download Readers and Viewers for the following document types below: