An incidence chart of major cancers in the United States over the past three decades shows lines that are mostly flat or moderately sloped. But if the incidence of a relatively uncommon cancer, esophageal adenocarcinoma (EA), is added to the graph, the contrast is stark. After all, rarely do incidence rates of any disease, let alone a type of cancer, increase by 300 to 400 percent in 30 years.

Already infamous for being one of the most lethal types of cancer, these stunning gains in incidence have garnered EA a reputation as a looming threat.

"There are a lot of issues at play as to why those numbers have gone up," says Dr. Nicholas Shaheen, director of the Center for Esophageal Diseases and Swallowing at the University of North Carolina at Chapel Hill. "But I am concerned because we haven't been able to figure out what's causing the increase."

Epidemiologic studies have identified several factors that significantly increase EA risk, says Dr. Wong-Ho Chow, of the NCI Division of Cancer Epidemiology and Genetics. A 1999 case-control study conducted in Sweden, for instance, found that long-standing and severe reflux symptoms increased EA risk by more than 40-fold. When coupled with overweight or obesity, the risk increase jumped to more than 100-fold. Several studies have shown that smoking is also an important risk factor.

While these are risk factors for what is still a fairly rare cancer - approximately 7,000 new cases a year, mostly late-stage disease - it's as if something has happened to unleash what was once nothing more than a medical anomaly.

"The only change in Western society that really mirrored this increase in incidence is the epidemic of obesity," says Dr. Shaheen. "And that makes it the most logical candidate for why things are going the way they are. But we really don't know that. It's still just a theory at this point."

According to Dr. Chow, for a significant portion of patients, EA is thought to begin with reflux, which causes acid from the stomach to spout up into the lower portion of the esophagus. The acid displaces cells that line the lower part of the esophagus - just above where it meets the stomach - with a different, potentially cancerous cell type in a condition known as Barrett's esophagus.

People with Barrett's esophagus have at least a 30- to 40-fold increased risk of developing EA compared with the general population; however, their overall risk is very low: 90 to 95 percent of patients will never develop the cancer.

Getting Bigger Numbers Much of the uncertainty surrounding EA can be directly attributed to the fact that it is rare. To gather enough EA and Barrett's cases - and, more importantly, biological samples from those cases - to generate the statistical strength to provide a clearer picture of these diseases, NCI has formed a consortium of academic medical institutions and cancer centers that specialize in EA and Barrett's. Called BEACON and currently chaired by Dr. Thomas Vaughan of the Fred Hutchinson Cancer Research Center, the consortium includes investigators who have conducted case-control studies of EA and Barrett's in the United States, Canada, Great Britain, Sweden, and Australia.

"Why only a small percentage of Barrett's continues to become cancer, and what makes some people progress and others not, we just don't know yet," Dr. Chow says.

Unfortunately, most people who develop EA don't have symptoms until they already have late-stage disease, at which time the few available treatment options are not very successful. Even for patients who are amenable to treatment, 5-year survival ranges from just 5 to 30 percent.

A handful of investigators who have been involved in studies with Barrett's patients are searching for molecular indicators that might predict the likelihood of somebody with chronic reflux or Barrett's progressing to EA. At the University of Southern California/Norris Comprehensive Cancer Center, Dr. Peter W. Laird and colleagues are using a test called MethyLight to see if certain changes in the extent of DNA methylation - a way to activate or deactivate genes - can detect EA in its earliest stages or predict the likelihood of progression.

At Fred Hutchinson Cancer Research Center, Dr. Brian Reid, who heads the Center's Barrett's Esophagus Program, and colleagues are examining biopsy samples to determine whether mutations in two genes (p16 and p53), as well as abnormalities in the amount of DNA in a cell - that is, a condition called aneuploidy - are predictive of progression to EA.

Because EA is still so rare, much of this work is really about surveillance, Dr. Reid notes.

"If we can reassure those people who are at low risk, and get high-risk people into appropriate surveillance or treatment options, it would be a big help," he says.

The only current option for screening of people with chronic reflux or surveillance of Barrett's is endoscopy, which involves inserting a long, flexible tube - fitted with a light and video camera on one end - through the mouth and down into the throat to examine and take biopsies from the esophagus and stomach.

But using endoscopy as an everyday screening tool for EA isn't prudent or really feasible, Dr. Shaheen argues. An estimated 40 percent of people have chronic heartburn, "and most will never go on to have Barrett's or certainly will never have cancer, so you'd literally be searching for a needle in a haystack," he says. "We do that in other areas of medicine, but we generally don't do it with a test that costs more than $1,000."

Dr. William J. Blot, of the Vanderbilt-Ingram Cancer Center, is leading an effort that he hopes might help identify candidates for regular EA surveillance. Much like the Gail model, which assesses breast cancer risk, he and his colleagues are developing a model that takes into account all of a person's risk factors and stratifies them into a low- or high-risk category for developing EA over a subsequent 5-year period.

There is not nearly as much data on EA risk factors as there is for a far more common cancer such as breast cancer, Dr. Blot says, so the new model would be less precise than the Gail model. "But," he adds, "it's a start."

By Carmen Phillips