Familial Association in Barrett Esophagus

G&H What are the risk factors for and prevalence of Barrett esophagus in the United States, in terms of age, gender, and ethnicity?

YR Risk factors for Barrett esophagus include advanced age, male sex, frequent gastroesophageal reflux disease symptoms of prolonged duration, white ethnicity, and to a lesser extent, tobacco and obesity. The average age at which a patient is diagnosed with Barrett esophagus in the United States is 63. There is, however, a clinical bias at play in this statistic because many young, healthy people do not seek medical care or undergo endoscopy. It is quite possible that a person diagnosed with Barrett esophagus at age 63 has actually had it for 10–20 years but was not previously aware of its presence.

Population-based studies suggest that the ratio of men to women with biopsy-proven Barrett segments of greater than 2 cm in length is about 2 men for every 1 woman. The ratio of men to women who have Barrett and progress to cancer is much higher, at a ratio of 9 men for every 1 woman, for unknown reasons. Although many potential explanations for the gender difference have been espoused, a definitive reason has not yet been found.

The role of ethnicity in the presence or absence of Barrett esophagus is a fascinating topic. The prevalence of long-segment Barrett esophagus among whites of European descent in developed countries, ranges between 0.34% and 0.5%, based on two population-based studies. The prevalence of intestinal metaplasia, of any length, is estimated at 1.6%. Although it can be argued that the rate of obesity in people of Hispanic or AAfrican-American descent is higher than in their white counterparts, they have only one-twentieth the risk of developing Barrett esophagus. It is not yet clear if this difference in prevalence represents a genetic propensity among whites to develop Barrett esophagus, a genetic protective factor associated with those of nonwhite descent, or an environmental factor that is not yet equally shared among the groups.

G&H Has the incidence of esophageal adenocarcinoma in developing countries increased in the past 30 years because of an increasing incidence of Barrett esophagus?

YR This is a common misconception. The short answer is no. This was best demonstrated in a population-based study by Conio and colleagues in 2001 that found that over 30 years, as more patients were endoscoped, more cases of short- and long-segment Barrett esophagus were diagnosed. In other words, two phenomena were key and greatly impacted the estimate of disease prevalence: high utilization of diagnostic endoscopy and physician recognition. The first case of short-segment Barrett esophagus was diagnosed at the Mayo Clinic in Rochester, Minnesota, in 1985. It was not that a patient with short-segment Barrett esophagus had not crossed our doors before 1985; short-segment Barrett esophagus was just not recognized as a disease until that time. The common clinical practice then had been to only biopsy salmon-colored mucosa that was 3 cm or longer in length. Physician recognition and behavior played an important role in disease identification and, hence, affected disease prevalence estimates. Without knowing the number of people endoscoped, the prevalence of Barrett esophagus can erroneously be interpreted as rising.

With regard to the incidence of Barrett esophagus, no one actually knows what it is. No one knows how many people without Barrett at their first endoscopy have progressed to it over a given time frame, such as 5 years. Conversely, because esophageal cancer is highly and rapidly lethal, one can estimate that the incidence and prevalence of esophageal cancer are similar. Because one can harbor Barrett for a long period of time, it is not appropriate to use prevalence to estimate incidence for this disease process.

G&H Have studies supported the idea of a familial association in Barrett esophagus?

YR There are numerous case reports of families in which multiple members have Barrett esophagus, hiatal hernia, and, in some cases, esophageal adenocarcinoma. Case series can be biased, however, because GERD is rather common. One of the Mayo Clinic's first studies on this topic, published in 1997, showed that GERD symptoms aggregate in families, a finding that has since been confirmed by two independent groups. Specifically, the study showed that the parents and siblings of patients with Barrett esophagus or esophageal adenocarcinoma met criteria for GERD symptoms twice as often as the parents and siblings of spouse controls.

These results naturally led to the design of a study examining whether Barrett esophagus and reflux esophagitis also aggregated in families. Because our results have not yet been published, specific data cannot be revealed, but I can share that the independent predictors of Barrett esophagus included advanced age, male sex, and frequent GERD symptoms of prolonged duration. Even after the independent risk factors were taken into account, however, there was still a greater than 2-fold increased risk of Barrett esophagus in first-degree relatives compared to the controls who lacked a family history of Barrett esophagus or esophageal cancer.

A descriptive cohort study published by Chak and associates in 2004 also confirmed family history to be a risk factor for Barrett esophagus.

G&H What recent advances have been made in the identification of genes associated with Barrett esophagus?

YR Our group has learned a great deal by observing the experience of investigators in hereditary colon cancer and inflammatory bowel disease. For example, only 2–4% of patients with colon cancer have familial adenomatous polyposis or hereditary nonpolyposis colon cancer as the etiology. Nonetheless, most of what we know about the genetics of colon cancer stem from these rare families with a genetic predisposition to colon polyps/cancer. Henry Lynch began collecting high-penetrance families five or more decades ago.

In the same vein, the Barrett's Esophagus Genomic Study (BEGS) Group was formed in 1998. We are a group of private practice and academic physicians and scientists from Europe, South America, Canada, and the United States, who collaborate by identifying families in which at least two members have biopsy-proven longsegment Barrett esophagus, with or without esophageal adenocarcinoma. Because short segments of Barrett epithelium can be challenging to confirm or to distinguish from routine intestinal metaplasia of the cardia, we aim for families with long segments of Barrett. Focusing on a specific phenotype is especially important when hunting for markers of genetic predisposition.

To date, of the 466 families that have come to the Mayo Clinic or have been referred to the BEGS Group, 101 have been confirmed to be high-penetrance kindreds. The first linkage analysis study has been completed on 31 families seen at the Mayo Clinic, identifying susceptibility loci for familial Barrett esophagus and esophageal adenocarcinoma and, on separate chromosomes, loci for familial hiatal hernia and reflux esophagitis. We are very excited by the results and are working to fine map and validate the regions of interest. Preliminary results have been presented in abstract form, and a manuscript is in progress. We are on the path to finding the genes for these familial disorders.

G&H What are the next steps for research in familial Barrett esophagus?

YR Once we find the genes/loci for familial Barrett esophagus, our next step will be to investigate whether the same chromosomal regions are associated with sporadic Barrett esophagus. The Mayo Clinic Esophageal Adenocarcinoma and Barrett's Esophagus (EABE) Registry Consortium was constructed to address this aim. Since 2001, every patient with long-segment Barrett esophagus, esophageal adenocarcinoma, and, as our control group, squamous cell carcinoma of the esophagus, has been invited to participate in the EABE Registry. Blood, tissue, demographic information, reflux symptoms, risk factors, and quality of life data are collected at baseline and annually. As of April 18, 2007, 1,674 (60%) of 2,787 unique patients have consented to participate in the EABE Registry, which is supported by partners in industry and the National Institutes of Health, and has been expanded to Mayo Clinic Jacksonville and will soon include Mayo Clinic Arizona.

Once chromosomal regions of interest are found for familial and sporadic Barrett esophagus, we look forward to further collaboration with basic science colleagues to identify the pathways and mechanisms, functional polymorphisms, and other components that are important in predisposing to the phenotypes of interest.

G&H How do you envision the ultimate role of genetic testing in the screening of Barrett esophagus?

YR Once the results of our study showing that first-degree relatives of patients with Barrett esophagus have a higher risk of having reflux esophagitis and/or Barrett esophagus, regardless of GERD symptoms, are confirmed by another group, it is possible that the guidelines for upper endoscopy screening may change. Once the genes and loci that predispose to familial Barrett esophagus and, perhaps someday, sporadic Barrett esophagus, are identified and confirmed, the ultimate goal will be to develop a blood or buccal swab DNA test to screen the general population, regardless of their perceived GERD symptoms. Right now, this is just a dream; we are very early on in this line of investigation.

Were this dream to come to fruition, I would envision that people who screened positive would then undergo an imaging study, be it upper endoscopy or capsule endoscopy, for confirmation. Patients found to have Barrett would still require thorough educational sessions regarding environmental risk factors for reflux (eg, caffeine, fat, alcohol, tobacco) because progression to neoplasia would still likely be due to a combination of genetic predisposition and environmental exposure.

G&H Could genetic testing also be utilized in the diagnosis of esophageal cancer?

YR At present, most people with a precancerous esophageal lining (Barrett) are not diagnosed in advance and, hence, miss the opportunity to potentially benefit from chemoprevention, ablation, and/or surveillance. This is a tragic situation. We, as a scientific society, cannot hope to diminish the mortality rate of esophageal adenocarcinoma until everyone who is at risk can be identified at an early, curable stage in their disease progression. Hopefully, this will eventually come to pass, perhaps facilitated by genomic testing.

In 1990, Cameron and associates conducted an autopsy study designed to estimate the prevalence of Barrett esophagus in Olmsted County, Minnesota. One of their conclusions was that for every 17 people with longsegment Barrett esophagus, only 1 person was aware of their diagnosis. In other words, only 1 of 17 people with long-segment Barrett had been clinically diagnosed. Even were we to allocate 100% of our resources to the 1 person known to have Barrett, we would not diminish the mortality rate, as 16 others would still be at risk for neoplastic transformation. In the 2001 follow-up study, performed in the same population by Conio and colleagues, 1 of 7 people with long-segment Barrett esophagus had been clinically diagnosed. Although these figures represent significant progress over merely a decade, the odds remain against us. The mortality rate of esophageal adenocarcinoma will not significantly decline until the majority of persons at risk for the disease have been identified.

Perhaps 3–10% of people with long-segment Barrett esophagus will progress to cancer in their lifetime. The annual cancer risk for each individual with Barrett esophagus is quite low, about 0.4–0.5% per year. Although this is clinically favorable, from a scientific point of view, these low incidence rates make esophageal adenocarcinoma difficult to study. Power calculations from Shaheen and other researchers have suggested that large prospective cohort studies following at least 10,000 Barrett patients for 10 years will be necessary to find key factors in the progression from Barrett to cancer. The EABE Registry may be a helpful resource in this process.

For more information on the Barrett's Esophagus Genomic Study Group and how to join it, please e-mail Dr. Romero at romero.yvonne@mayo.edu.