Facing the challenge of data transfer from animal models to humans: the case of persistent organohalogens

Abstract

A well-documented fact for a group of persistent, bioaccumulating organohalogens contaminants, namely polychlorinated biphenyls (PCBs), is that appropriate regulation was delayed, on average, up to 50 years. Some of the delay may be attributed to the fact that the science of toxicology was in its infancy when PCBs were introduced in 1920's. Nevertheless, even following the development of modern toxicology this story repeats itself 45 years later with polybrominated diphenyl ethers (PBDEs) another compound of concern for public health. The question is why? One possible explanation may be the low coherence between experimental studies of toxic effects in animal models and human studies. To explore this further, we reviewed a total of 807 PubMed abstracts and full texts reporting studies of toxic effects of PCB and PBDE in animal models. Our analysis documents that human epidemiological studies of PBDE stand to gain little from animal studies due to the following: 1) the significant delay between the commercialisation of a substance and studies with animal models; 2) experimental exposure levels in animals are several orders of magnitude higher than exposures in the general human population; 3) the limited set of evidence-based endocrine endpoints; 4) the traditional testing sequence (adult animals--neonates--foetuses) postpones investigation of the critical developmental stages; 5) limited number of animal species with human-like toxicokinetics, physiology of development and pregnancy; 6) lack of suitable experimental outcomes for the purpose of epidemiological studies. Our comparison of published PCB and PBDE studies underscore an important shortcoming: history has, unfortunately, repeated itself. Broadening the crosstalk between the various branches of toxicology should therefore accelerate accumulation of data to enable timely and appropriate regulatory action.

Figure 1

Temporal distribution of studies addressing PCB (A) and PBDE (B) toxicity in experiments with animal models.

Figure 2

Temporal trends toward decrease of average daily dose in experiments with animal models. A – experiments addressing PCB toxicity; B – experiments addressing PBDE toxicity.

Figure 3

Distribution of endpoints of PCB and PBDE toxicity addressed in experiments with animal models.

Figure 4

Distribution of endocrine endpoints of PCB (A) and PBDE (B) toxicity addressed in experiments with animal models.

Figure 5

Trends towards increase in number of developmental experiments versus experiments with adult animals. A – developmental experiments addressing PCB toxicity; B – developmental experiments addressing PBDE toxicity; C – increase in number of pre- and perinatal versus neonatal exposures within developmental experiments with PCB.

Figure 6

Distribution of animal species in experiments addressing PCB (A, C) and PBDE (B, D) toxicity. A, B – big taxonomic groups, C, D – mammalian models.

Figure 7

Temporal trends in percent of studies using estimation of internal dose in animal models. A – studies addressing PCB toxicity; B – studies addressing PBDE toxicity.

Figure 8

Temporal trend towards increase in number of exposures via injections versus oral exposures in PCB experiments.