Barrett’s Esophagus Screening Best Evidence: In Whom?

We are all aware that the incidence of esophageal adenocarcinoma has been rising rapidly over the last 50 years. Since the cancer is associated with an endoscopically visible lesion in Barrett’s esophagus, screening for esophageal adenocarcinoma ought to be relatively easy compared to screening for other terrible cancers, such as pancreatic or ovarian. But there are many reasons why screening for esophageal adenocarcinoma may not be as effective as one would hope. For instance, surveillance biopsies may be insufficiently obtained in time or space, patients at greatest risk for the cancer typically are at greatest risk of competing causes of death, such as cardiovascular disease, and patients who are least likely to respond to endoscopic therapy appear to have the highest baseline risk of progressing to cancer to begin with. For the sake of argument, let’s assume that screening for esophageal adenocarcinoma is indeed worthwhile despite those shortcomings. In that case, who should we screen?

It is quite clear that we are performing a colossal number of scopes, but in the wrong patients.

Consider first how well we are doing currently in screening. Nationally, we perform roughly 2 million upper endoscopies a year in patients with symptoms of gastroesophageal reflux disease (GERD), costing society over $1 billion annually. And despite all of those endoscopies, fewer than 15 percent of patients with esophageal adenocarcinoma have undergone an upper endoscopy prior to their presentation with the cancer. It is quite clear that we are performing a colossal number of scopes, but in the wrong patients.

A major problem with our current approach is that a slight majority of patients with esophageal adenocarcinoma actually deny having had prior GERD symptoms of note.1 And the patients that we do scope with GERD symptoms are usually the patients whose symptoms have not responded adequately to proton pump inhibitors. Often these patients have atypical symptoms such as throat clearing or globus sensation, often with substantial comorbid anxiety, and they are unlikely to actually have pathological reflux as a cause of their symptoms. The patients who present with esophageal adenocarcinoma, if they have GERD symptoms at all, often tell me that they have had heartburn for years but just took over-the-counter antacids and never discussed their symptoms with a doctor since they were managing fine. If screening is to effectively reduce the burden of esophageal adenocarcinoma, it will need to rely on primary care providers or even the public health service to identify appropriate patients for screening. Relying on gastroenterologists to screen patients among the highly selected population of patients referred for consultation of refractory symptoms is a failed strategy.

So which individuals are the ones who should be screened? The highest risk patients are white men over the age of 50 or 60 who have experienced years of heartburn or regurgitation, have used tobacco and have abdominal obesity.2,3 But if we only screened that group of patients, we would have a limited impact on the burden of the cancer. Additional groups of patients could include men above 50 or 60 years of age with either GERD symptoms or a combination of abdominal obesity and tobacco use. Family history of esophageal adenocarcinoma in a first-degree relative might also be a reasonable indication for screening, irrespective of whether the patient has GERD symptoms. On the other hand, women of any age are at very low risk for esophageal adenocarcinoma, even if they have GERD symptoms.

None of the known risk factors for Barrett’s esophagus or esophageal adenocarcinoma have extremely high odds ratios. And the incidence of esophageal adenocarcinoma in the general population is still quite small (2.5 per 100,000 patients per year). So we can’t really identify patients who are at high risk. The best we can do with currently known risk factors is to identify the patients who are not at exceedingly low risk like the rest of the general population. It may not be a wise use of our society’s resources to screen for Barrett’s esophagus under such circumstances when the screening tool involves a highly trained (and expensive) endoscopist with a staff of nurses, technicians, monitoring equipment and possibly an anesthesiologist.

Alternative methods exist. Transnasal unsedated endoscopy removes much of the cost associated with sedated endoscopy, and non-endoscopic devices also hold promise. One such device, the cytosponge, is an abrasive sponge contained within a pill-sized capsule on a string that a nurse or technician can administer. After swallowing the capsule with water, the gelatin capsule dissolves and the sponge is pulled out, obtaining a semi- histologic quality cytology specimen that appears to be quite accurate for identifying Barrett’s esophagus. It may even be validated for identifying dysplasia. If such devices are very low cost (meaning something like $100 including analysis), then screening a wide enough spectrum of the population to capture most patients who are destined to develop the cancer might finally be an efficient strategy of screening for esophageal adenocarcinoma.

Dr. Rubenstein has no conflicts to disclose.


1. Rubenstein JH. Risk factors for Barrett’s esophagus. Current Opinion in Gastroenterology 2014;30:408-414.
2. Thrift AP, Kendall BJ, Pandeya N, et al. A model to determine absolute risk for esophageal adenocarcinoma. Clinical Gastroenterology & Hepatology 2013;11:138-44.e2.
3.Rubenstein JH, Morgenstern H, Appelman H, et al. Prediction of Barrett’s esophagus among men. American Journal of Gastroenterology 2013;108:353-62.

Join the discussion

Your email address will not be published. Required fields are marked *