Response to the NRCD Petition 

Neurodevelopmental Toxic Effects

NRDC claims that “2,4-D is neurotoxic.” It supports this contention by citing findings in several studies in the published literature. NRDC ignores, however, significant problems with the design, route of administration, and in some cases even the test material in the studies cited, making the findings in these studies an inappropriate basis for risk assessment or regulatory decision-making on 2,4-D. For example, NRDC cited a zebrafish embryo study by Ton et al (2006) as evidence of developmental neurotoxicity. This study did not assess developmental neurotoxicity but rather was a method development exercise designed to create a screening system for potential developmental neurotoxicity. There were no negative controls in the Ton et al study. In addition, the lowest concentrations affecting some of the endpoints cited by NRDC, such as disrupted motor neuron growth, occurred at concentrations that exceeded the LC50 for zebrafish (despite NRDC’s claim that the exposure was “sub-lethal”); therefore, the reported endpoints do support the contention that they are indicators of developmental neurotoxicity. As an attempt to create a screening assay, its protocol clearly lacked the sensitivity and specificity required to generate data that would enable a regulator to assess whether 2,4-D affects the developmental neurotoxicity of fish (and even less to humans).

NRDC next cite the same Ton et al 2006 zebrafish paper to support their claim that 2,4-D is associated with developmental neurotoxicity in mammals, “including decreased motor activity and Parkinson’s-like tremors.” (NRDC Pet. at 6.) The problem is Ton et al. never tested mammals. In the introduction to their fish study, Ton et al. do claim of an association between 2,4-D and decreased motor activity and tremors in mammals but Ton’s support for their claim, in turn, is a news and view commentary by Giasson and Lee, 2000 who discuss using the pyrethroid insecticide and pesicide rotenone as a model for Parkinson’s disease. Giasson and Lee never even mentioned 2,4-D. NRDC then cites a collection of rat studies (NRDC footnotes 34 - 38) all of which administered high dose levels that have been shown to overwhelm metabolic clearance. Specifically, these studies all tested doses of 2,4-D at 70-100 mg/kg ip or sc, which exceeds the threshold for renal clearance of 2,4-D (established in oral [gavage and dietary] studies: van Ravenzwaay et al., 2003; Saghir et al., 2006). Further, a recent unpublished range-finding study of 2,4-D in CD® rats showed overt maternal and pup toxicity in the same dose range, reflected in severely decreased maternal feed consumption during lactation and marked pup weight gain decreases and increased mortality (at a 100 mg/kg bw/day dietary dose). Among the mammalian studies cited by NRDC was a study relating to neurotoxicity or developmental neurotoxicity that did not appear to test overwhelmingly high doses (Bortolozzi et al. 2001). This study, however, was an experimental test where 2,4-D was directly administered into the rat’s brain. This route is not relevant to human or mammalian risk assessment and it is unfortunate that NRDC relies on such a study.

The mammalian studies cited by NRDC are summarized below:

1.  NRDC cites Evangelista de Duffard et al (1995) in support of the claim that 2,4-D causes “delays in brain development and abnormal behavior patterns, including apathy, decreased social interactions, repetitive movements, tremor and immobility.” (NRDC Petition at 7). First, the study does not assess brain development. Second, the findings in this study cannot be attributed to 2,4-D alone. Indeed, the study is entitled “Altered behavioral responses in [2,4-D] treated and amphetamine challenged rats” designed to be a pharmacological challenge study in adult animals (thus not assessing brain development). The effects were a result of acute toxicity of intraperitoneal amphetamine, spiperone and/or haloperidol, possibly enhanced by a prior high (50 or 100 mg/kg) acute intraperitoneal dose of 2,4-D. The authors clearly state in the abstract that “all behaviors were not seen in the 2,4-D treated rats.” The observations NRDC reports of “decreased social interaction and apathy” were made in animals given combined 2,4-D and amphetamine doses that caused lethality, which indicates that these behaviors may be non-specific effects rather than any selective neurological effect. The data supporting effects on these particular “social” behaviors is especially weak. There is no description in the methods section of how “social interaction” or “social apathy” was evaluated, and the reporting of these results is not sufficiently rigorous to allow any conclusions to be made.
2. NRDC cites Bortolozzi et al (2001) to support its claim that “[r]odent studies have revealed a region-specific neurotoxic effect on the basal ganglia of the brain, resulting in an array of effects on critical neurotransmitters and adverse effects on behavior.” (NRDC Pet. at 7). The route of administration in Bortolozzi et al. was direct injection of 2,4-D in DMSO into specific brain areas of the rat. Bortolozzi et al. indicated that the estimated 2,4-D concentrations in each brain area were 2-4 mM, which the author defined as 40 to 100-fold the concentration in brain after systemic treatment. This study is not a developmental neurotoxicity study but a very high dose adult toxicity study using a route of exposure that has no relevancy to human exposures.
3. NRDC concludes by citing three papers from the same research group (Rosso et al. 2000; Duffard et al. 1996 and Konjuh et al. 2008) to allege that “this herbicide specifically appears to impair normal deposition of myelin in the developing brain.” (NRDC Pet. at 7). These were the only evaluations of actual developmental neurotoxicity in mammals among the studies cited by NRDC. These studies, however, were conducted at excessively high doses of 70 and/or 100 mg/kg/day 2,4-D either through subcutaneous or intraperitoneal injection, which are dose levels that would be expected to exceed the threshold level for saturating renal clearance of 2,4-D. The two earlier studies (Duffard et al 1996; Rosso et al 2000) were methodologically flawed, because the litter was not used as the unit for statistical analyses, which is critical to a developmental neurotoxicity evaluation to avoid bias due to litter effects. The reporting of the histopathological evaluation of these two studies is limited, with no sample size reported, and insufficient reporting of methods and results to allow a conclusion that unbiased objective measurements were made using a sufficient sample size. The authors themselves admit that their findings could be due in part to undernutrition and resulting weakness caused by maternal toxicity instead of a specific effect on the developing nervous study.

The hypothesis that the effects reported by Duffard et al. and by Rosso et al. may have been due to undernutrition was tested in a more recent study by this group (Konjuh et al, 2008), in which the experimental design was improved compared to the earlier studies noted above. In the Konjuh et al study “undernutrition” was achieved by retaining 14 pups per litter, while “normal” nutritional status was achieved by culling to the more standard litter size of 8 pups litter. Although Konjuh et al (2008) report some differences in myelin parameters between undernourished pups and those exposed intraperitoneally to 2,4-D, the general pattern of 9 of the 12 myelination parameters evaluated actually appeared similar between undernourished and 2,4-D well-nourished pups. This evaluation is considered too limited to be more than a hypothesis generating study because there was only one exposed dose group (70 mg/kg/day), the sample size was too small (single pups from each of 4-5 litters), and no corrections were made for multiple statistical comparisons.

In summary, the papers cited by NRDC do not provide credible or substantive evidence that 2,4-D causes developmental neurotoxicity at exposure levels or routes of administration relevant to humans. As noted, in response to EPA Data-Call-In requirements, the 2,4-D Task Force is currently sponsoring a one-gen reproduction study in rats of 2,4-D in the diet. The protocol for this study includes assessment of developmental neurotoxicity endpoints. To date, there have been no dose-related statistically significant indications of developmental neurotoxicity, even at dose levels demonstrated to be well above the renal clearance threshold in rat dams and pups.

Next: Mutagenic and Genotoxic Effects