So, let’s pause, take a step back, and look at Barrett Esophagus with a lot less concern than we used to.
The take-home message is that we need to re-evaluate the evaluation of risk in BE, and guideline committees need to incorporate these data, recognizing that these data are all European, and the biopsy protocols are different in Europe.
Barrett Esophagus: Are We Screening Too Often?
Hello. I am Dr. David Johnson, Professor of Medicine and Chief of Gastroenterology at Eastern Virginia Medical School.
Surveillance and diagnosis are part of the mainstay of evaluation for Barrett’s esophagus (BE) in patients with reflux disease. This has been predicated on knowing that BE is a precancerous condition associated with progression to cancer at an estimated incidence of 0.4%-0.5% per year, and to high-grade dysplasia at 0.9% per year. The incidence that we quote to patients is 1.5% yearly progression to high-grade dysplasia cancer.
Those studies were done some time ago, and 2 new, very large epidemiologic studies are changing the waterfront of BE. One was published in the Journal of the National Cancer Institute
 in July; this study came from Ireland. It was a very large database, including all patients with BE, capturing data from more than 8000 patients. The mean duration of follow-up was 7 years (range: 1-20 years). The de novo progression to cancer from BE (excluding incident cancers) at 1 year was 0.13%. This incidence is very low and very different from 0.4%-0.5% per year, which had estimated the risk to be 30-40 times that of the general population. The risk is now down to 10-11 times the risk in the general population.
The most recent study, published in The New England Journal of Medicine
 in October, shows essentially the same information, this time coming from a study in Denmark that included all patients with BE, and excluded cancer found within the first year after diagnosis. The mean follow-up was 6 years, and the rate of progression to cancer was 0.12% per year. Progression to cancer was 3 times more common in men than women. The risk in women was inordinately low.
Let me put this in perspective for you. If you look at cost modeling studies that have been done in the past (one that comes to mind is from 2003 in the Annals of Internal Medicine
) they found that screening or surveillance of BE was effective from a cost standpoint, at 5-year intervals if the incident cancer risk was ≥ 1.9% per year. At 1.9%, we are talking about 4 times the rates that are being reported now. The incident rate is considerably lower than 0.1%-0.13%, the rate at which it was cost-effective to screen at 5-year intervals. The current guidelines recommend 3-year follow-up for BE. This raises the question: is it cost-effective at the present time, when, because of dollar restrictions we are tossing out things like Pap smears and prostate-specific antigen tests? Surveillance in BE is certainly now subject to cost evaluation.
Throw in the most recent study that comes from a Dr. Wani and colleagues, a 5-site multicenter study that looked at the progression from low-grade dysplasia to cancer. Data were analyzed for 210 patients, and specimens were reviewed by centralized pathologists. The progression rate from low-grade dysplasia to esophageal adenocarcinoma (with a follow-up of 6 years) was 0.44% per year. In progression to cancer risk, this group did not differ from patients who had no low-grade dysplasia at study entry. The study also reiterated that there is terrible concordance for agreement on low-grade dysplasia. The kappa value was 0.14, which indicates poor agreement. Intra-observer variability was very high. This study raises red flags about the paradigm of 6-month evaluations for patients with low-grade dysplasia.
No cancers were evident in the low-grade dysplasia patients within 2 years of the diagnosis of low-grade dysplasia, so it also raises the question, could we potentially lengthen out the surveillance of low-grade dysplasia patients to longer intervals? We return to the bottom line. In the present day, no data suggest that patients with BE die at a higher rate than patients without BE. No data say that patients with BE under surveillance die at a lower rate than patients without surveillance. Even for esophageal adenocarcinomas, the risk for death doesn’t seem to be higher in patients with BE under surveillance strategies.
The take-home message is that we need to re-evaluate the evaluation of risk in BE, and guideline committees need to incorporate these data, recognizing that these data are all European, and the biopsy protocols are different in Europe. The definition of intestinal dysplasia is not requisite for the diagnosis of BE in Europe as it is in the United States. These data raise a number of important questions about the relative risk for patients with BE. Do they need surveillance if they don’t have dysplasia? Do patients with dysplasia need surveillance at lengthened intervals?
These findings suggest that we should not be ablating patients with low-grade dysplasia or metaplasia outside of research protocols. Certainly in clinical practice, I don’t think that should be the standard of care, nor would it be justifiable in the present day based on the cost evaluation and the relative risks for these patients.
There is a lot on the plate for discussion. BE guidelines committees should convene in a more rapid fashion and reassess the relative risk, and provide guidance so we can achieve cost-effective medicine for our patients with BE.
So, let’s pause, take a step back, and look at BE with a lot less concern than we used to. There is no question that BE is a precancerous condition and that it is associated with adenocarcinoma of the esophagus. As the biologic behavior of BE has become better understood, the relative clinical significance has dwindled. In your next conversation with your patients with BE, put this into perspective and evaluate the strategies for surveillance. Hopefully, the guidelines will also provide us with some meaningful changes in the not-too-distant future.
I’m Dr. David Johnson. Thanks again for listening.