Toxic Cocktails

Consider these five real-life scenarios:

(i) The monitoring of pesticide residues in Europe has generated little concern about single substances but the simultaneous occurrence of multiple toxic residues could be a very different matter.
(ii) Lipophilic pollutants, such as PCBs, dioxins and brominated flame retardants often occur together in fatty foods, such as milk and dairy products, and in certain fish.
(iii) Mixtures of polychlorinated aromatic hydrocarbons are present in grilled or smoked meats.
(iv) A range of molds, which may produce different mycotoxins, are commonly found on pistachios, peanuts, corn and sorghum.
(v) Multiple essential minerals and vitamins, as well as enzymes, preserving and flavoring agents, can be found in feeds for intensive farming.

As for pesticides, 15-20 percent of samples contain residues of multiple monitored substances, albeit each within the maximum permitted level. The combinations of different substances may be additive in effect or may reciprocally modify the action or metabolism of other substances. How should such threats be assessed?

The most straightforward approach is the one adopted for dioxins, namely assessing the overall impact on a given mechanism, in this case interaction with the aryl hydrocarbon receptor. Since potency of each compound is known, the overall threat can be judged. But do we know the toxicity mechanisms for all relevant chemicals? Can we wait to gain knowledge of these mechanisms and in the meantime avoid making a decision? Do substances with dissimilar mechanisms always act independently and in isolation?  The answers are no, no and no.

In 2013, the European Food Safety Authority (EFSA) took a bold approach when dealing with the toxicological assessment of combined exposures to pesticide residues. Now, all substances that induce the same effect in a given target tissue or system are considered to act together in an additive way; for example, two pesticides that induce hypothyroidism are grouped together, irrespective of their different chemical structure. Beyond the mechanism at the molecular or biochemical level, it is the adverse outcome that matters, the EFSA states - provided that the outcome is clearly defined. Thus, a toxic cocktail in food may now be characterized by the components that produce the same effect. Available knowledge determines the criteria for grouping substances: the “effect” may be an alteration in laboratory animals or a pattern of metabolomic changes in a robust in vitro system.

One tricky issue for the approach is how to identify the effector components of any given cocktail. Although a cocktail may include a large number of, for example, lipophilic contaminants, only a few chemicals may drive the overall effect of the mixture. In straightforward cases, such as dioxins, a few potent and/or highly concentrated/persistent congeners are responsible for the lion’s share of the toxicity, and assessing just these substances provides a reasonable measure of the biological impact of dioxins as a whole for practical risk assessment. Clearly, the main drivers of toxicity are exposure and potency.

Potency has to be measured using assays and parameters that are comparable. Exposure is a topic that becomes exponentially more complex as the focus shifts from residues within a given commodity through to exposure across the whole diet; and since diet is impacted by factors as diverse as ethnicity, region, income and age, it is clear that exposure to relevant substances is also highly diverse. This is demonstrated in biomonitoring studies of the total burden of endocrine (for example, estrogenic) activity in human serum, which have shown that natural substances and xenobiotics from diet (and also from the environment) contribute to a given endocrine activity.

The downside of the approach is that the chances of drowning in complexity are just around the corner. The in-depth assessment of cumulative risks should, therefore, be better focused to answer risk management questions: for example, what pesticides should be regulated more strictly or what pollutants drive risk at a contaminated site? Finally, toxicologists should be most interested in producing new experimental data on whether shared adverse effects can be defined at the molecular level.