WASHINGTON, Dec. 16 – Environmental Protection Agency officials received a mixed message from scientists assembled to review evidence of whether glyphosate is a human carcinogen.

The members of a Scientific Advisory Panel concluded their four-day meeting in Arlington, Virginia, by offering opinions on EPA’s conclusion that the active ingredient in Roundup, the world’s most widely used herbicide, is “not likely to be carcinogenic to humans.”

The result: The panel was split on the issue. Some members backed EPA’s finding and others said that the evidence was “suggestive” of carcinogenic potential for the chemical.

Some on the panel were not expecting that. “I’m a little surprised there’s this controversy. I thought EPA did a pretty good job,” said Marion Ehrich, co-director of the Laboratory for Neurotoxicity Studies at the Virginia-Maryland Regional College of Veterinary Medicine. She supported the “not likely” conclusion, saying “there just isn’t enough (evidence) there” to support any other finding.

“Rarely does a panel disagree as much as this one,” said SAP veteran Kenneth Portier, a statistician with the American Cancer Society. He urged EPA to figure out how it plans to use epidemiology to inform its risk assessments, an issue raised recently by CropLife America, which petitioned the agency Nov. 29 to postpone regulatory decisions informed by epidemiological studies until it has adopted concrete data quality standards.

Panel members who supported EPA said the evidence from the epidemiological and animal studies was simply not strong enough to support anything else.

But there was uncertainty all around the table as members wrestled with EPA’s review of the data in its September 2016 “white paper.”

“I am in agreement with EPA with a little reservation,” said Aramandla Ramesh, a professor in the Department of Biochemistry & Cancer Biology at Meharry Medical College in Nashville, Tennessee. He said that glyphosate’s role in “promotion or progression of spontaneously induced lesions cannot be ruled out.”

Lianne Shepard, assistant chair in the Department of Environmental and Occupational Health Sciences at the University of Washington, said that epidemiological studies that looked at whether glyphosate causes non-Hodgkin lymphoma (NHL) were enough to tilt the scale in favor of a “suggestive” finding. Individually, the studies may not contain strong evidence of a connection between glyphosate exposure and NHL, but taken together, they do, she said.

“Clearly, it’s suggestive to me, and it’s the most appropriate public health conclusion to reach,” she said.

Kenny Crump, a private consultant and statistical modeler, repeatedly warned of the danger of “recall bias” in the epidemiological studies on glyphosate and non-Hodgkin lymphoma. Recall bias refers to the problem of study participants accurately remembering what they did in the past. In the matter of the non-Hodgkin lymphoma studies, the issue would be farmers’ recollection of pesticide use.

“It’s very difficult to remember what pesticides you’re exposed to,” Crump said.

But Shepard said that while she agreed that recall bias is a feature of case-control studies, “it doesn’t necessarily mean that they should be disregarded.”

The meeting featured more than an entire day of public comments, including a presentation by Peter Infante, a consultant and former OSHA epidemiologist who was originally included on the panel but was left off the roster after CropLife America questioned his objectivity. Speaking with reporters, Infante said he did not leave the panel voluntarily but declined to go into more detail. EPA has repeatedly declined to comment on the matter. 

Infante said his review of the epidemiological data, which focused on non-Hodgkin lymphoma, showed that glyphosate should be considered a likely human carcinogen.

There also was a lengthy presentation by representatives of Monsanto, glyphosate’s original registrant. The scientists supported EPA’s evaluation and noted that every regulatory body that has looked at glyphosate has concluded it is not carcinogenic.

The controversy over the chemical primarily stems not from any regulatory finding, but from a conclusion by the World Health Organization’s International Agency for Research on Cancer in March 2015 that glyphosate probably causes cancer in humans.

At the SAP meeting, representatives of both the European Food Safety Authority and the German Federal Institute for Risk Assessment offered testimony disagreeing with the IARC conclusion.

“Glyphosate is unlikely to be genotoxic, neurotoxic or toxic for the reproduction or development and is unlikely to pose a carcinogenic hazard to humans,” the EFSA presentation said.

Dow AgroSciences, Syngenta, BASF and DuPont Crop Protection also made presentations backing the “not likely” conclusion. On the other side of the question were groups such as Center for Food Safety, the Natural Resources Defense Council and Consumers Union.

NRDC said EPA improperly dismissed evidence of elevated cancer in rodent studies. CFS said that even a finding of carcinogenicity for glyphosate would not lead to its banning. “Many pesticides that EPA has designated ‘likely to be carcinogenic to humans’ or ‘probable human carcinogens’ are in widespread use today,” the group said.

More meeting materials are here.