Response to the NRCD Petition 

2,4-D is Not an Endocrine Disruptor

NRDC claims that 2,4-D has “extensive hormone disrupting activity” and that these effects “may occur at low doses”. In fact, the scientific record indicates 2,4-D has a very low endocrine disrupting potential. Endocrine effects in animal toxicity studies attributed to 2,4-D are restricted to unrealistically high doses that exceed saturation of renal clearance of 2,4-D in rodents and thus are of limited value to human or animal risk assessment. In addition, the doses to which animals are exposed in studies are all substantially greater than possible human exposures as demonstrated by extensive occupational and general population biomonitoring studies (see Exposure Section below).

Although NRDC claims that EPA “illogically ignores the existing data,” EPA in March, 2007 issued a data call-in for a two-generation reproduction study (Guideline No. 870.3800) that specifically requires examination of thyroid, gonadal, reproductive and other endocrine-sensitive endpoints. In addition, EPA has also requested an assessment of both developmental neurotoxicity and immunotoxicity potential. The Agency on several occasions has rejected the position that “data gathered under the Endocrine Disruptor Screening Program (“EDSP”) is a prerequisite to a safety determination under FFDCA section 408.” (US Federal Register 2007a at 68662, 68676) (denying NRDC petition to cancel DDVP, Dec. 5, 2007); (US Federal Register 2006 at 43906, 43919-43921) (denying states petition to cancel certain pesticides).

While the Task Force is conducting the required reproduction and developmental neurotoxicity studies, EPA will keep in place an additional 10X uncertainty factor (for a total 1000X) to account for the data gap. EPA has stated the additional 10X factor is reasonable because the doses to be used in the requested toxicity studies likely will be similar to animal NOAELs and LOAELs previously established from existing studies (EPA response to public comments OPP-2004-0167-0090 16-Dec-04). Thus, the results of the data call-in studies are unlikely to alter the RED 2,4-D risk assessments by a factor greater than those already imposed by the 10X data gap uncertainty factor and may provide state-of-art toxicological evidence to either eliminate or reduce application of the full currently applied 10X data gap factor.

The 2,4-D Task Force has completed the in-life portion of a state-of-art GLP study examining potential life-stage impacts of dietary 2,4-D administration to rats on reproductive, endocrine, developmental neurotoxicity and immunotoxicity responses. The study addresses the concerns raised by the EPA data call-in (USEPA 2007) requesting an evaluation of these endpoints. The Task Force will submit to EPA the final study report in 2009.

NRDC asserts that EPA “relies on the hollow excuse that a formal screening program does not yet exist to avoid examining potential endocrine effects”. This is incorrect. EPA has imposed an extra 10X data gap uncertainty factor that reasonably accounts for the lack of a state-of-art reproductive assessment which includes a more detailed assessment of endocrine endpoints. EPA issued a data call-in for studies to support the final re-registration of 2,4-D. It important to note that the studies required by the data call-in will provide an apical assessment of the potential endocrine properties of 2,4-D in rats, and as such, for purposes of 2,4-D risk assessment, will supplant any mammalian toxicity information elements gleaned from the proposed Endocrine Disrupting Screening Program.

 Contrary to NRDC’s assertions, the estrogenic properties of 2,4-D have been extensively examined in both mammalian and non-mammalian in vitro and in vivo studies. These studies discussed below, do not support NRDC’s claim that 2,4-D is a “potent” endocrine disruptor in any species. It is very unlikely that 2,4-D is an endocrine disruptor or stressor given the extremely low human exposures to the herbicide.

NRDC cites Xie et al. (2005) as reporting that 2,4-D has a “relatively potent estrogenic effects in fish.” However, the authors of the study refer to their findings as showing elevated estrogenic activity, as measured by elevation of plasma vitellogenin, under conditions of a “worst-case scenario” exposure (1.64 mg/L 2,4-D). The NOEC and LOEC values for this response were reported as 0.0164 mg/L and 0.164 mg/L, respectively. As noted in the comments above in response to the NRDC claim that 2,4-D was detected in 50% of sampled waters, a U.S. National Surface Water Monitoring program reported finding 2,4-D in only 1.6% of samples, and then only at a maximum concentration of 1.9 ppb (0.0019 mg/L). Thus repeated field monitoring indicates a large separation between the responses in this laboratory study and 2,4-D concentrations reported in surface water.

Interpretation of Xie et al. is further confounded by several other observations. The reported vitellogenin responses appear to be variable and inconsistent across experimental parameters such as dose and presence of surfactants. For example, control vitellogenin responses ranged from near zero to as high as approximately 10 ng/mg total plasma protein; such variation confounds clear identification of NOEC and LOEC values reported in this study. Xie et al. also noted that a combined treatment with 1.64 mg/L 2,4-D and 1.46 mg/L of the surfactant R-11 resulted in approximately 16-18 ng/mg vitellogenin expression. A subsequent experiment using a similar treatment period and 2,4-D concentration (1.64 mg/L) but a lower concentration of R-11 (0.89 mg/L) resulted in much higher vitellogenin expression (approximately 50 ng/mg). In addition, the mechanism of the high-concentration induced vitellogenin response remains unclear. Numerous studies have demonstrated that 2,4-D does not bind to or activate mammalian or fish estrogen receptors (Table 1). Further, a major photolysis/hydrolysis environmental metabolite of 2,4-D, 2,4-dichorophenol, also exhibits no or extremely weak in vitro estrogen receptor activity (Korner et.al., 1998; Nishihara et.al. 2000). Thus, Xie et.al. (2008) does not provide evidence that 2,4-D has “potent” estrogenic activity, and the variable nature of the reported responses requires further investigation and replication before inferring either fish or mammalian environmental health risks.

Evidence of 2,4-D estrogenic activity has not been demonstrated in several whole organism toxicity tests. In ovo exposure of alligator eggs to a range of 2,4-D concentrations had no effect in the hatchlings to the estrogen-sensitive endpoints of 1) production of females at male-determinant egg incubation temperatures; 2) gonadal and reproductive histology (females: epithelial cell in Mullerian duct and medullary regression of ovary; males: sex-cord diameter); and 3) hepatic or gonad-adrenal-mesonephros (GAM) aromatase activity (Spiteri et.al, 1999).

In a series of GLP subchronic 90-day dietary toxicity studies in rats, 2,4-D acid and its commercial salt and ester derivatives produced only minimal or no responses in estrogen and/or androgen sensitive tissues, and when observed, were restricted to high doses which exceeded either or both the maximum tolerated dose (“MTD”) and renal clearance capacity (Charles et. al., 1996). In a GLP developmental toxicity study of 2,4-D and its salt and ester derivatives in rats and rabbits, offspring effects were again limited to high maternal doses exceeding renal clearance saturation and were not characteristic of endocrine-directed targets (Charles et.al., 2001). Finally, reproductive studies conducted with either a mixture of 2,4-D and picloram (Oakes et.al., 2002) or with 2,4-D alone (study submitted to support re-registration of 2,4,D, Rodwell and Brown, 1986) resulted in high-dose only effects that were not characteristic of endocrine-mediated responses (i.e., no effects on reproductive tract malformations, retained nipples in males, pre-coital length, fertility, histological alterations in accessory sex glands, epididymides, testes, ovaries and uteri). Although the Rodwell and Brown study reflected guideline protocol design when it was conducted in 1985, the Task Force’s ongoing extended one-generation reproduction (“one-gen reproduction”) study to satisfy EPA’s data call-in will further supplement and update reproductive performance using state-of-art assessments including endocrine-sensitive endpoints (e.g., estrus cycle activity, ano-genital distance, and sperm parameters. etc.).

NRDC concerns regarding potential human health implications of 2,4-D effects on thyroid hormone alterations also are not supported by the data. In the subchronic rat studies cited by NRDC and described above (Charles et.al., 1996), decreased serum levels of T4 and/or T3 were noted only at high doses that exceeded renal clearance, and minor histological effects in the thyroid were also seen only at a dose which exceeded the MTD (300 mg/kg/day). The ewe study cited by NRDC (Rawlings et.al 1998) is too limited in its design to provide useful information regarding potential thyroid effects of 2,4-D. The study used only a single 2,4-D dose and a limited number of animals per treatment group. While the study noted a small but significant decrease in serum T4, this observation was not accompanied by any change in thyroid histopathology and TSH was not measured.

NRDC also describes a series of other studies it contends provides evidence of 2,4-D endocrine activity. None, however, provide compelling evidence of potential environmental or human endocrine health concerns. Specific comments on these studies are as follows:

1. Liu et al. (1996) found no significant effects of 2,4-D in hCG-stimulated release of testosterone in rat Leydig cells cultures. Although NRDC states 2,4-D produced a significant increase in estrogen release from these cells, it neglects to mention the minimal effect concentration for this effect was 500 µM. This concentration translates to approximately 110 µg/ml, and is substantially greater than the maximum 2,4-D plasma concentration reported in rats treated with 100 mg/kg/day in the diet (63 µg/ml; Saghir et.al., 2006). Given that 100 mg/kg 2,4-D is well above saturation of renal clearance, in vitro findings in this concentration range are not indicative of an endocrine risk.
2. The in vitro mechanistic study of Kim and co-workers (2005) is based on their earlier report that 2,4-D increased prostate weight in a rats treated with 50 mg/kg/day by oral gavage (Kim et al. 2002). However, the effect on prostate weight occurred at a dose and route of administration (oral gavage) known to be above the threshold for saturation of renal clearance, Thus, the observation underpinning the mechanistic in vitro study not only is yet to be examined in standardized guideline toxicity studies, the use of the high dose renders the usefulness and interpretation of this in vitro study of questionable value to potential risk. The effects of dietary 2,4-D treatment on prostate weight and pathology is currently being examined in the data call-in studies requested by EPA and being conducted by the 2,4-D Task Force.
3. The effects of 2,4-D on prolactin, progesterone and estrus cycle alterations described in Duffard et al. (1995) were at a dose level well above saturation of renal clearance and thus are of limited value for risk assessment.
4. Lerda and Rizzi (1991) does not support the contention that 2,4-D causes endocrine effects in farmworkers because of several reporting and methodological weaknesses. Many study-specific details were not reported including background of controls, number of participants excluded due to spermatogenesis-affecting health conditions, method used in “consideration” of external factors, detection limit for 2,4-D in urine, time period of urine collection, and ranges of sperm parameters and 2,4-D urine levels evaluated (only means provided in the report). In addition, the selection of controls was inappropriate as comparison was to workers in the field exposed to 2,4-D, but the controls were not agricultural workers doing similar field work. Field work involves exposure to other confounding factors (e.g., increased temperature, dusts, allergens) that likely alter sperm parameters.
5. The study of Garry and co-workers (1996) assessing potential associations of herbicide use and birth defect outcomes in an agricultural region of Minnesota was later described by this same research team as “a first effort” that suffered from a “limited reporting timeframe” (Garry et.al., 2002). In the latter report, the authors concluded that a more detailed cross-sectional analysis of this area showed no statistically significant correlations between 2,4-D use and excess adverse birth or neurodevelopmental effects.

In summary, the Task Force respectfully submits that the weight of evidence does not support NRDC’s claim that 2,4-D “has been shown to have extensive hormone-disrupting activity.” The Task Force’s ongoing one-gen reproduction study, conducted to satisfy the requirements of Data Call-In Notice, which specifically assesses endocrine-related endpoints, will supplement existing in vivo and in vitro studies that demonstrate that 2,4-D does not effect the endocrine system. The non-GLP studies cited by NRDC suffer from significant methodological or reporting limitations, or they administered doses far in excess of those to which humans would ever be exposed. Finally, EPA has repeatedly stated that data gathered under the FQPA EDSP is not a prerequisite to a determination that a pesticide chemical residue meets the safety standards of FFDCA § 408.

Next: Neurodevelopmental Toxic Effects