Advisors Split Over EPA Plan To Use Human Data To Regulate Insecticide

April 25, 2016

EPA science advisors are split over the agency's plan to tighten regulation of a commonly-used insecticide to protect against neurodevelopmental risks, with several advisors arguing the decision is based on limited epidemiological data, although others say the agency's novel approach is reasonable given the potential risk to children.

EPA for years has struggled with how to protect against a possible neurodevelopmental risk to children from exposure to chlorpyrifos, and late last year the agency proposed banning use of the substance on food to protect against risks from aggregate exposure, including neurodevelopmental risks to the developing fetus.

The agency is assessing the risk of chlorpyrifos and other organophosphate pesticides, and has shifted focus from protecting against inhibition of the enzyme acetylcholinesterase (AChE) in the nervous system to protecting against neurodevelopmental effects, which may occur at lower levels of exposure.

The change is based in part on epidemiological data which measures human health effects resulting from exposure rather than animal toxicology tests that have traditionally supported agency pesticide reviews. The approach has drawn criticism from the U.S. Department of Agriculture (USDA) and pesticide industry groups who say it is unnecessary and has not been properly vetted, and EPA convened an advisory panel to weigh in on the issue.

During the April 19-21 Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) SAP review in Arlington, VA, several panelists levied strong criticism against EPA's approach, arguing the plan is based on a single Columbia University epidemiological study that was not designed for regulatory purposes.

"EPA is put in a hard spot here because these [epidemiological] studies are good, but they were set up to be research studies and not to be used in regulation," said panelist Marion Ehrich, a professor of pharmacology and toxicology at the Virginia-Maryland College of Veterinary Medicine.

Ehrich and other panelists, including panel chairman James McManaman, a professor of obstetrics and gynecology at the University of Colorado, also raised concerns about limitations in the epidemiological study, including the analysis of the data.

"There has been a longstanding concern about the use of these data," McManaman said, noting that prior SAPs recommended against using the studies, and arguing the current SAP can not ignore the limitations.

But at least two other panelists generally backed EPA's approach, noting that the current regulation based on the AChE endpoint is inadequate to prevent potential neurodevelopmental risks in children.

Stella Koutros, of the National Cancer Institute, said EPA is taking reasonable steps to protect children. She also faulted other panelists for backing industry arguments that uncertainties surrounding the evidence are sufficient to hold off moving to a more protective endpoint.

Koutros also argued that the prior SAPs, in 2008 and 2012, had assessed the strength of the data and the role of the current SAP is to assist EPA in how best to use the study that agency officials believe is strong.

"That's not what we're here to judge," Koutros told panelists who criticized limitations in the epidemiological evidence. "The agency it appears to me has made the decision, they're using the data," and is asking for the panel's assistance in how to apply the information.

Industry groups and the USDA have argued that EPA's plan to strengthen oversight of chlorpyrifos and other organophosphate pesticides based on epidemiological data conflicts with long-standing agency practice. They have called for EPA to craft procedures for using epidemiological data in pesticide risk assessment and seek outside input on that process.

Additionally, industry has argued that EPA's plan to use the Columbia University study to ascribe a neurodevelopmental risk is inappropriate because the researchers have not provided the study's raw data to the agency for further analysis.

At the outset of the SAP, EPA officials said the agency recently wrote Columbia a letter requesting the raw data, and have been told they would receive it, though they have not yet.

During the meeting, EPA staff said the agency has followed the recommendation of a prior SAP and used modeling to help corroborate the Columbia study's findings. Agency officials also defended the study, calling it strong, and saying the research is especially informative given that the study came around the same time as EPA and industry in 2000 reached a voluntary agreement to end indoor uses of chlorpyrifos.

Anna Lowit, a senior scientist and toxicologist in EPA's pesticides office, said the study's analysis of chlorpyrifos levels in blood and neurodevelopmental effects in children were "strikingly different" before and after the regulatory change, resulting in a "pseudo dose response" between chlorpyrifos exposure and neurodevelopmental risks "that is unique to epidemiology."

Lowit also pushed back against panelists' assertions that EPA's decision is based on a single epidemiological study, saying the Columbia study is "one piece of a very large" body of research on chlorpyrifos that spans hundreds of studies and many years.

During the three-day SAP meeting, several panelists said they were troubled by the degree EPA is basing the decision on a single study that can not be reproduced.

"What I would like to see to reduce my uncertainty is replication of this study," said Isaac Pessah, a professor of pharmacology and toxicology at the University of California Davis' School of Veterinary Medicine.

Alvin Terry, a pharmacology and toxicology professor at Georgia Health Sciences University, said the study uses exposure estimates based on a single point in time and suggested a risk from such low levels of exposure as to not be biologically plausible. "It's very difficult for me to connect the dots and make a recommendation on a decision of such magnitude," he said.

But Koutros and another panelist, while acknowledging limitations in the data, said EPA's approach is reasonable and that agency risk assessors have advanced understanding and usefulness of the data through modeling of potential exposures.

"The fact that we do not understand the mechanism by which [a neurodevelopmental effect] occurs does not mean that it does not actually happen," Koutros said.

Sharon Sagiv, an epidemiology professor at the University of California at Berkeley, defended aspects of EPA's analysis but also agreed with other panelists' concern that the agency is putting significant stock in a single study.

"From a logical standpoint, how do we come to the middle here?" she asked. "As a panel we can't ignore the fact that we have an epi study suggesting an association with neurodevelopmental effects" even if there are limitations in the research.

The panel floated a variety of options for EPA to consider to strengthen the pool of data to better defend their approach. The ideas include contracting an independent lab to conduct additional analysis of the Columbia researchers' data, should it become available, or searching for other studies that may have blood samples containing chlorpyrifos that could be analyzed, or using a weight of evidence approach that looks at epidemiological data along with other factors. -- Dave Reynolds

Not a subscriber? Sign up for 30 days free access to exclusive environmental policy reporting.