Subscribe
Share
Share
Embed
This episode follows one fifty-eight-year-old man with long-standing heartburn to show that non-dysplastic Barrett esophagus is a field at risk, not a cancer, and that every decision aims to find dysplasia without over-labeling. It begins with who qualifies for a single screening endoscopy and why those criteria are limited. It then covers the American definition of Barrett, with a view and a biopsy, followed by Prague measurement and Seattle sampling. Next come how Barrett forms and progresses, how his risk sets a one-year then three-year surveillance interval, the role of PPI and aspirin, and when surveillance should stop.
Topics covered
Key decisions
This is an AI-generated podcast, and some pronunciations may be imperfect. Thank you for your understanding, and we hope you enjoyed this content.
For the full chapter with MCQs, tables, and primary-guideline references, visit www.boardpearls.com.
Questions or feedback: hello@boardpearls.com.
The most complete gastroenterology board-review podcast on the web, and it is entirely free. This is the GI series from Board Pearls: board review built around clinical reasoning, not recall. Every episode takes one high-yield topic and works it the way you would on rounds: a case to anchor it, the framework that sorts the differential, and the specific decisions the exam rewards. No trivia, no textbook dictation, just the thinking that earns the point.
Over 100 episodes span the full blueprint across nine modules (esophagus, stomach and duodenum, small bowel, colon, pelvic floor, liver, pancreas and biliary, endoscopy, and the cross-cutting topics), grouped by chapter and built from the guidelines and pivotal trials the boards are written from: ACG, AGA, AASLD, and ASGE.
Pair it with the written curriculum, question bank, and AI tutor at boardpearls.com. Questions or feedback: hello@boardpearls.com.
Welcome to Board Pearls. This is episode one of three of the Barrett Esophagus and Esophageal Cancer chapter, in the Esophageal Disorders module. This episode is Barrett esophagus without dysplasia, through one patient. A fifty-eight-year-old man with long-standing heartburn asks whether he needs a scope, and his case teaches who qualifies for screening and why the criteria are limited, what counts as Barrett at endoscopy and on biopsy, how the segment is measured and sampled so dysplasia isn't missed, how Barrett forms and progresses, how his risk sets the surveillance interval, what he takes to lower that risk, and when surveillance stops.
Here's the patient. A fifty-eight-year-old man has had heartburn several times a week for fifteen years, controlled with an over-the-counter proton pump inhibitor, PPI, that he takes before breakfast. He smoked for twenty years and quit five years ago. His waist is forty-three inches. He's white, and he's never had an endoscopy. He feels fine. His brother-in-law was just diagnosed with esophageal cancer, and he asks whether he needs a scope.
His question is whether he should be screened for Barrett esophagus, a metaplastic change of the distal esophagus in response to chronic acid and bile injury, and the precursor of esophageal adenocarcinoma. There are rules for screening because Barrett is too uncommon in the general population to justify scoping everyone with heartburn, so screening is limited to people whose risk is high enough.
The twenty twenty-two American College of Gastroenterology guideline supports a single screening endoscopy for a patient with chronic reflux, meaning weekly heartburn or regurgitation for five years or more, plus at least three other risk factors. Those are male sex, age fifty or older, white race, central obesity, a history of smoking, and a first-degree relative with Barrett esophagus or esophageal adenocarcinoma. Central obesity means a waist over forty inches in men or thirty-five in women, or a waist-to-hip ratio over zero point nine in men or zero point eight in women. It's measured at the waist rather than by body mass index, because abdominal fat drives reflux and Barrett independently of overall weight. American Gastroenterological Association guidance treats reflux as one risk factor among the others rather than a requirement, and would screen anyone with three or more, whatever their reflux history.
Counted against those criteria, he qualifies. He has chronic reflux, plus male sex, age over fifty, white race, central obesity, and a smoking history. That's five. One thing doesn't count: his brother-in-law isn't a blood relative, so the family history criterion doesn't apply, though it's the reason he came in.
Screening has real limits, and they explain where it's heading. Roughly forty percent of patients who develop esophageal adenocarcinoma never had heartburn. In cohort data, about fifty-five percent of patients with new adenocarcinoma wouldn't have qualified under the guideline criteria. Screening hasn't been proven to lower mortality in randomized data, and lead-time and length-time bias inflate the apparent benefit of surveillance in observational cohorts. So the field is moving toward broader, cheaper screening without endoscopy.
Two devices collect esophageal cells without a scope. One is a gelatin capsule containing a compressed mesh sponge on a string. The capsule dissolves in the stomach, the sponge is pulled back up through the esophagus, and the cells it collects are tested for trefoil factor three, a marker of intestinal metaplasia, with optional p53 staining for dysplasia risk. The other is a capsule-enclosed balloon with a textured surface that's dragged through the distal esophagus, with the cells tested by next-generation sequencing for methylation of the vimentin and cyclin A one genes. Both detect about eighty to eighty-five percent of Barrett, with specificity around ninety percent, and they're the emerging alternative to endoscopy for population-scale or primary care screening.
He qualifies, so he has a screening endoscopy, and reading it requires the definition of Barrett, which has two parts, a view and a biopsy. The view starts with the landmark. The gastroesophageal junction is identified by the top of the gastric folds and the distal end of the esophageal palisade vessels, and both are reported, because the diagnosis depends on where the columnar lining sits relative to that point. The squamocolumnar junction, or Z line, marks the upper edge of the columnar lining. Barrett requires salmon-colored columnar mucosa extending at least one centimeter above the gastroesophageal junction. And the biopsy must show intestinal metaplasia: columnar cells containing goblet cells.
That's the American definition, and it differs from the British one. The British Society of Gastroenterology accepts columnar lining without goblet cells, called cardiac mucosa, as Barrett, arguing that it's itself a metaplastic response to reflux that can express the intestinal markers CDX2 and villin and accumulate the same DNA, deoxyribonucleic acid, content abnormalities as goblet cell metaplasia. American guidelines require goblet cells, because the published cancer risk estimates come from cohorts defined that way, and because most studies suggest cardiac mucosa alone carries minimal cancer risk. So in the United States, columnar mucosa without goblet cells isn't Barrett.
The one-centimeter rule matters in practice, and it's a trap. An irregular Z line, with small tongues of salmon mucosa less than a centimeter above the junction, should not be biopsied at all. Goblet cells there are intestinal metaplasia of the cardia, not Barrett. They're found in fifteen to twenty percent of people without symptoms and carry no measurable cancer risk, so a biopsy only creates a label that commits the patient to years of needless surveillance.
His endoscopy shows salmon mucosa extending fully around the esophagus for two centimeters above the junction, with one tongue reaching five centimeters. Before the biopsies come back, the segment is measured and sampled, and both decide his future intervals.
The extent is recorded with the Prague classification. C is the circumferential extent: the height above the junction at which the salmon lining still goes all the way around. M is the maximal extent, including the longest tongue. He's C two M five. Agreement between observers is good for segments over one centimeter and poor below that, one reason ultra-short Barrett, under one centimeter, has historically been treated separately. Barrett three centimeters or longer is long-segment, and under three is short-segment. His maximal extent of five makes him long-segment, and that will shorten his surveillance interval.
Sampling follows the Seattle protocol, and it starts with looking, not biopsying. Dysplasia is patchy, and random biopsies sample only a small fraction of the lining, so the whole segment is inspected slowly with high-definition white light and dye-based or virtual chromoendoscopy. A careful exam takes more than a minute per centimeter of Barrett, because faster exams measurably miss more dysplasia and cancer. Anything visible, a nodule or an irregular area, is sampled separately, and preferably removed by endoscopic mucosal resection rather than biopsied, because a resection gives the pathologist the whole lesion and its depth, which forceps biopsies can't.
The rest of the segment gets four-quadrant biopsies, one from each quarter of the circumference, every two centimeters along the segment. Once dysplasia has been found, that tightens to every centimeter. For a segment under four centimeters, the practical version is at least eight random biopsies plus one per centimeter of any tongue.
A brush can add to that yield. Wide-area transepithelial sampling with computer-assisted three-dimensional analysis uses an abrasive brush across the whole segment, and a neural network reassembles the loose cells into three-dimensional images that flag possible metaplasia and dysplasia for the pathologist. In the largest prospective series, adding it to forceps biopsies roughly doubled both Barrett detection and the dysplasia yield. The American Society for Gastrointestinal Endoscopy supports it as an adjunct, and the twenty twenty-two American College of Gastroenterology guideline made no recommendation. It's a supplement to biopsies, not a replacement.
His biopsies show intestinal metaplasia with goblet cells and no dysplasia. He has non-dysplastic, long-segment Barrett esophagus. Long-segment Barrett is itself definitive objective evidence of reflux disease under the Lyon consensus, so he'd never need reflux monitoring to prove it.
What he has is a field at risk, not a cancer, and how it formed explains the rest of his care. Refluxed acid and pepsin destroy the squamous lining at the junction, exposing progenitor cells that repopulate the wound. The candidates include basal cells of the squamous lining, glands of the cardiac mucosa, submucosal gland duct cells, and a population of transitional basal cells at the junction identified in mice. Ongoing reflux reprograms the transcription factors in those cells, and the wound heals not as squamous lining but as intestinal metaplasia with goblet cells. Cardiac mucosa is itself often acquired metaplasia that expresses CDX2 and villin, and it's increasingly seen as the precursor of intestinal metaplasia rather than a normal variant.
From there, the sequence runs from intestinal metaplasia to low-grade dysplasia, high-grade dysplasia, intramucosal carcinoma, and invasive adenocarcinoma. Bile acids and acid together drive it, by causing oxidative DNA damage, double-strand breaks, p53 mutations, and a type two helper T cell-leaning cytokine environment in the lining. A slow traditional pathway loses tumor suppressors in turn, CDKN2A, SMAD4, and p53, followed by oncogene amplification.
A faster pathway explains why some cancers appear between surveillance exams. About sixty percent of Barrett cancers arise along it: an early clone with a p53 mutation doubles its whole genome to a tetraploid state and then tolerates further mutations quickly. That can move from non-dysplastic mucosa to invasive cancer faster than a three- to five-year interval can catch. So surveillance lowers risk, but it can't eliminate it.
His rate of progression is low but real. Non-dysplastic Barrett progresses to adenocarcinoma at roughly two to five tenths of a percent a year. The risk rises with male sex, smoking, a longer segment, and confirmed low-grade dysplasia. A published progression model weights them: nine points for male sex, five for smoking, one per centimeter of Barrett, and eleven for confirmed low-grade dysplasia, with annual progression reaching about two percent in the highest-risk group. He has the first three.
Two tissue tests can refine that risk. Aberrant p53 immunostaining on biopsy is the most validated single marker of progression to high-grade dysplasia or cancer. A tissue systems pathology test integrates nine biomarkers, including p53, with morphology into a quantitative score that risk-stratifies low-grade dysplasia better than expert pathologists. A low-risk score is reassuring, with a negative predictive value around ninety-six percent, and the twenty twenty-two American Gastroenterological Association practice update endorses it for risk stratification in non-dysplastic Barrett.
Visible Barrett is only part of the cancer picture. About half of patients diagnosed with esophageal adenocarcinoma have no detectable Barrett at diagnosis, and cohort data suggest the cancer arises from a Barrett field in about forty-six percent of cases. Cancers with intestinal metaplasia present survive longer in matched cohorts. That fits a field-cancerization model: visible Barrett marks the size of the field and the chance to survey it, while cancers without it have likely overgrown the precursor lining by the time they're found.
His surveillance intervals follow that risk, and his first repeat endoscopy is at one year. The highest yield for finding dysplasia or cancer that was missed is the first follow-up after diagnosis. A meta-analysis of fifteen non-dysplastic Barrett series found that about a quarter of cancers were diagnosed within the first year after the index exam, and most cancers and high-grade dysplasia were found at or within that first year. Those lesions were most likely already present and missed, which is why there's an early second look even when the first exam was clean.
If that one-year exam is clean, the twenty twenty-two guideline surveys non-dysplastic Barrett every three years for a long segment, three centimeters or more, and every five years for a short segment. The older three-to-five-year window combined the two, and the guideline separated them because long segments progress more often. So he returns in one year, and if that's clean, every three years.
Alongside surveillance, he takes medication to lower the risk. Every patient with Barrett takes a PPI at least once daily. A large randomized trial of about twenty-five hundred patients followed for a median of almost nine years compared esomeprazole forty milligrams twice daily with twenty milligrams once daily, with or without aspirin three hundred milligrams daily. High-dose PPI lowered a combined outcome of death from any cause, high-grade dysplasia, and adenocarcinoma. Aspirin lowered the same outcome once patients taking other nonsteroidal anti-inflammatory drugs were censored, and the two effects appeared additive. The results weren't strong enough to recommend both for everyone, so high-dose PPI and aspirin are individualized decisions, not routine. Antireflux surgery isn't done to prevent cancer, because the data don't show it protects better than a PPI. He continues his once-daily PPI before breakfast.
Surveillance also has an end point, because it helps only if the patient lives long enough for a detected cancer to be found and treated. Modeling suggests stopping non-dysplastic surveillance at about eighty-one in a healthy man and seventy-three in a man with severe illness, and earlier in women: about seventy-five in a healthy woman and sixty-nine in a woman with severe illness. A large randomized trial of more than three thousand patients, comparing endoscopy every two years with endoscopy only for symptoms over at least ten years, found no difference in overall or cancer-specific survival and no difference in stage at cancer diagnosis. So surveillance is a judgment about benefit rather than a fixed rule, and his will be revisited as he ages.
Two related rules come up in patients like him. A patient with Los Angeles grade C or D erosive esophagitis needs a repeat endoscopy after PPI healing, because severe inflammation can hide Barrett underneath it, and a first exam read as negative can miss it. And distal esophageal biopsies are also part of the workup for eosinophilic esophagitis, where an eosinophil infiltrate can be confused with the reactive change seen in Barrett.
Here are the traps the boards set in this patient.
In screening, scoping a patient with reflux but only one or two other risk factors goes beyond the guideline. And counting a relative by marriage as family history misreads the first-degree criterion.
In diagnosis, biopsying an irregular Z line under a centimeter creates a Barrett label for intestinal metaplasia of the cardia that carries no measurable risk. Calling columnar mucosa without goblet cells Barrett fails the American definition. Taking random biopsies without a careful, unhurried inspection first misses visible lesions. And biopsying a nodule instead of resecting it gives up the depth the pathologist needs.
In surveillance, skipping the one-year follow-up after diagnosis misses the lesions most likely to have been overlooked. Using a five-year interval for a long segment undersurveys a higher-risk patient. Offering a fundoplication for cancer prevention isn't supported. And continuing surveillance in a patient whose life expectancy is too short to benefit adds risk without gain.
So, start to finish.
Screen once with chronic reflux plus at least three risk factors: male, fifty or older, white, central obesity, smoking, or a first-degree relative. Non-endoscopic cell collection is the emerging screening option.
Barrett is salmon mucosa at least a centimeter above the top of the gastric folds, with goblet cells on biopsy, and an irregular Z line isn't biopsied. Record it with Prague C and M, inspect slowly at more than a minute per centimeter, resect anything visible, and take four-quadrant biopsies every two centimeters. Long-segment Barrett is itself proof of reflux disease.
Without dysplasia, progression runs about two to five tenths of a percent a year. Repeat at one year, then every three years for three centimeters or more and every five years for less. Keep every patient on at least a once-daily PPI, individualize high-dose PPI and aspirin, and stop surveillance when life expectancy no longer supports it.
Episode two is the patient whose biopsies come back with dysplasia: why a second pathologist reads them first, and how each grade, from indefinite to high-grade, leads to a different next step.