Some weed killers may need to be treated with more respect.

By Janet Raloff

Web edition: May 28, 2009

A new study links two weed killers with pancreatic cancer in pesticide applicators and their spouses. The authors, most of whom work for the National Cancer Institute, note that they are the first to link this particular malignancy with the farm chemicals — pendimethalin and EPTC — and really don’t know how either would trigger cancer (although they do have a theory). But for now they are recommending that “these findings should be considered hypothesis generating and in need of confirmation.”

How’s that for not hyping your data?

Pancreatic cancer is a rare disease and nearly always fatal. Producing few early symptoms, the disease is usually diagnosed once the cancer is advanced. But even where the cancer is found fairly early and removed, only one in five patients survive five years. And there is basically no cure. Which explains why this malignancy is the fourth leading cause of cancer deaths in the United States.

With such a poor prognosis, there’s been a push to understand what might predispose someone to develop the disease. Cigarette smoking seems to double risk, and a host of other factors also up the odds that someone will get this cancer — such as obesity, chronic pancreatitis, diabetes, cirrhosis and chewing tobacco. Taken together, however, these risks still only explain about a quarter of cases.

A couple epidemiological studies had suggested that certain agricultural chemicals, principally weed killers, might spike someone’s risk. So Gabriella Andreotti of NCI’s division of cancer epidemiology and genetics, in Bethesda, Md., and her colleagues decided to probe risks of the disease among the 89,000 participants of a long-running federally funded Agricultural Health Study. Its participants include 57,000 private, licensed pesticide applicators (mostly white men) and some 32,000 wives of applicators. All lived in Iowa or North Carolina when they were recruited in the mid-1990s.

Throughout the first seven years of followup, 93 cases of pancreatic cancer developed — 64 in applicators, the rest in spouses. As in other studies, smoking, having diabetes, or tipping the scales at an undesirable weight all increased an individual’s odds of developing cancer. But after adjusting for these risk factors, two of some 50 pesticides showed a statistically significant correlation with the cancer.

Farmers treat their fields with pendimethalin (sold under such trade names as Accotab, Go-Go-San, Herbadox, Penoxalin, Prowl, Sipaxol, Stomp and Way-Up) to get rid of certain annual grasses and broadleaf weeds that threaten their corn, potatoes, rice, cotton, soybeans, tobacco, peanuts and sunflowers. The other herbicide, EPTC (also known as Eptam, Eradicane, Shortstop and Genep), targets grassy and broadleaf weeds in fields planted with beans, forage legumes, potatoes, corn and sweet potatoes.

These were the only two pesticides that seemed to increase pancreatic cancer risk in an exposure-dependent manner, Andreotti's group reports in the May 15 International Journal of Cancer. That is, cancer risk increased 40 percent among people who had incurred moderate exposure to pendimethalin or had employed lots of protective gear when applying. Among mores highly exposed individuals, the pancreatic cancer risk was triple the rate seen among those applicators never exposed to this herbicide. Similarly, there was an 80 percent increased risk for relatively low-level exposures to EPTC. That risk jumped among the more heavily exposed individuals to two-and-a-half times the cancer risk in pesticide applicators who had never used this chemical.

Although Andreotti’s group acknowledges that the mechanisms by which either chemical might trigger pancreatic cancer remain unknown, they suspect it might have something to do with the fact that the one can harbor the carcinogen nitrosamine as a trace impurity and both weed killers are able to form related N-nitroso compounds. Nitrosamines and related compounds are suspected human carcinogens affecting tissues including the pancreas.

I guess the lesson is that anything we broadcast into the environment at concentrations expected to kill something should be treated with respect.