Updated systematic assessment of human, animal and mechanistic evidence demonstrates lack of human carcinogenicity with consumption of aspartame.

Aspartame has been studied extensively and evaluated for its safety in foods and beverages yet concerns for its potential carcinogenicity have persisted, driven primarily by animal studies conducted at the Ramazzini Institute (RI). To address this controversy, an updated systematic review of available human, animal, and mechanistic data was conducted leveraging critical assessment tools to consider the quality and reliability of data. The evidence base includes 12 animal studies and >40 epidemiological studies reviewed by the World Health Organization which collectively demonstrate a lack of carcinogenic effect. Assessment of >1360 mechanistic endpoints, including many guideline-based genotoxicity studies, demonstrate a lack of activity associated with endpoints grouped to key characteristics of carcinogens. Other non-specific mechanistic data (e.g., mixed findings of oxidative stress across study models, tissues, and species) do not provide evidence of a biologically plausible carcinogenic pathway associated with aspartame. Taken together, available evidence supports that aspartame consumption is not carcinogenic in humans and that the inconsistent findings of the RI studies may be explained by flaws in study design and conduct (despite additional analyses to address study limitations), as acknowledged by authoritative bodies.

Development of a neurotoxic equivalence scheme of relative potency for assessing the risk of PCB mixtures.

PCBs produce adverse effects in humans and animals by several modes of action. The first mode of action is binding of coplanar or mono-ortho-PCBs to the aryl hydrocarbon (Ah) receptor leading to effects associated with the activation of this receptor. The remaining PCB congeners do not activate this receptor and have different modes of action underlying their toxic effects. One mode of action that has been shown for di-ortho-substituted non-coplanar PCBs (PCB congeners with two or more chlorines in the ortho-positions) is the interference with intracellular signaling pathways dependent on Ca(2+) homeostasis and the resulting cellular, organ-level and organismal effects. The ortho-substituted non-coplanar congeners produce other cellular or organ-level effects including changes in protein kinase C translocation, changes in cellular dopamine (DA) uptake, formation of reactive oxygen species, and thyroid effects. Here, we propose a scheme for developing relative potency estimates (REP) for the PCB congeners not considered in the TEF scheme used to assess the toxicity of coplanar and mono-ortho-PCBs and chlorinated dioxins and furans. Because a number of the modes of action listed here for the ortho-substituted non-coplanar PCB congeners have been implicated in the neurotoxic effects of these PCBs congeners, this relative potency scheme is referred to here as the Neurotoxic Equivalent (NEQ) scheme for estimating toxicity of PCB mixtures. The Neurotoxic Equivalent (NEQ) values are developed in a way similar in concept to the derivation of the well-known TEF congener values. Although this scheme is in its infancy and the set of NEQ values are limited by the current data, there are several compelling reasons for proposing such a scheme now. First, it should open discussions as to how different modes of action can be utilized to predict congener potency differences for the effects they produce. Second, consideration and evaluation of the ability of the proposed NEQ scheme to predict the toxicity of PCB mixtures will assist in the identification of the specific modes of action relevant to the effects produced by non-coplanar PCBs. If other modes of action are suggested and subsequently identified, then other schemes of relative potency could be developed specifically for those modes of action, distinct from either the TEF scheme or the NEQ scheme. Knowing these other modes of action and the relative toxicity of the various congeners would advance our understanding of PCB toxicology and thereby ultimately improve our ability to estimate the toxic potency of PCB mixtures for each identified mode of action. Third, a quantitative scheme for assessing the toxicity of the non-coplanar PCB congeners present in a mixture has the potential to improve significantly future risk assessments of PCB mixtures.

Development of an oral cancer slope factor for Aroclor 1268.

Rodent cancer bioassays indicate that substantial differences exist among PCB mixtures in terms of tumorigenic response, although no bioassay has been conducted with Aroclor 1268. The USEPA has used data from these studies to develop three sets of PCB cancer slope factors (CSFs) ranging from 0.07 to 2.0(mg/kg-day)(-1). Selection of the appropriate CSF for risk assessment purposes is largely a function of the exposure circumstances rather than the PCB mixture involved. Since the congener composition of Aroclor 1268 differs substantially from that of the predominant PCB mixture (Aroclor 1254) used to derive the CSFs, the validity of applying existing CSFs to Aroclor 1268 is questionable. We have therefore undertaken the task of developing cancer potency estimates specifically for Aroclor 1268. Potency estimation approaches for Aroclor 1268 were based in part on existing potency estimates for other PCB mixtures, coupled with the relative 2,3,7,8-tetrachlorodibenzo-p-dioxin toxic equivalents (TEQ) content and bioaccumulation potential of PCB mixtures. As such, both Ah-dependent and independent mechanisms of tumorigenesis were considered relevant. Both empirical evidence and mechanistic considerations indicate Aroclor 1268 is substantially less toxic and carcinogenic than the PCB mixtures that have been used by the USEPA to develop CSFs. The present analysis indicates that Aroclor 1268 is likely to be 1-2 orders of magnitude less potent than Aroclor 1254 in terms of potential tumorigenicity. Therefore, we suggest an upper-bound cancer potency factor of 0.27(mg/kg-day)(-1) for Aroclor 1268, a value that is 7- to 8-fold lower than the USEPA's current default, but nonetheless adequately conservative.

Evaluation of the carcinogenicity of 1,1-dichloroethylene (vinylidene chloride).

The U.S. Environmental Protection Agency has classified 1,1-dichloroethylene (vinylidene chloride; VDC) as a "C" carcinogen and has developed an inhalation unit risk value and an oral cancer slope factor for this chemical. The development and use of these cancer potency estimates for risk assessment purposes are questionable. The inhalation unit risk value is based on increased kidney adenocarcinomas in Swiss mice from one study. This type of cancer was not increased in female mice or in rats or hamsters in the same study nor in male mice of a similar strain in another study with higher VDC exposures. The VDC oral cancer slope factor is based on a non-statistically significant increase in adrenal pheochromocytomas in male rats following oral exposure in a standard National Toxicology Program chronic bioassay. Both human and animal literature relevant to VDC carcinogenicity was reviewed according to the USEPA draft Guidelines for Carcinogen Risk Assessment with the objective of determining the weight-of-evidence for VDC carcinogenicity. We conclude that information currently available for VDC is most appropriately characterized in a weight-of-evidence narrative by the descriptor "inadequate for an assessment of human carcinogenic potential." For chemicals with this descriptor, dose-response assessment is not indicated. Under this guidance, quantitative estimates of cancer risks associated with VDC exposure are inappropriate until additional, more definitive evidence for human carcinogenicity becomes available.