Subscribe
Share
Share
Embed
Board Pearls is a gastroenterology board review built around clinical reasoning, not recall. Each 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.
The gastroenterology series covers 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. Episodes are grouped by chapter and built from the primary guidelines and pivotal trials the boards draw from (ACG, AGA, AASLD, ASGE), not from textbook summaries.
Use it as an audio companion to the written curriculum, MCQs, 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. In this episode we cover Barrett diagnosis, pathogenesis, and surveillance: Barrett lives in the conjunction of a visible columnar segment above the gastroesophageal junction and intestinal metaplasia on biopsy, the dysplasia ladder drives surveillance interval, and low-grade dysplasia is the shared-decision branch between continued surveillance and endoscopic eradication.
Start with the definition, because every other piece of the Barrett vignette is downstream of it. Barrett esophagus is a metaplastic response of the distal esophagus to chronic acid and bile injury. The American definition has two parts and the boards reward candidates who hold both. The first part is endoscopic. Salmon-colored columnar mucosa must extend at least one centimeter above the gastroesophageal junction. The second part is histologic. The biopsy must show intestinal metaplasia, meaning columnar epithelium containing goblet cells. Either part alone is not Barrett.
The gastroesophageal junction itself is where the diagnosis lives or dies, because the disease is defined relative to that landmark. Two endoscopic anchors mark it. The proximal extent of the gastric folds is the first. The distal end of the esophageal palisade vessels is the second. The candidate should expect to see both reported in any Barrett vignette, because the location of the columnar lining relative to that landmark is what makes the disease the disease. Above the gastroesophageal junction is the squamocolumnar junction, the Z line. The Z line is the proximal extent of the columnar lining and is what the Prague system measures.
The one-centimeter floor is not arbitrary. Salmon mucosa with goblet cells less than one centimeter above the gastroesophageal junction is intestinal metaplasia of the cardia. It is found in roughly fifteen to twenty percent of asymptomatic patients on endoscopy, and it does not carry measurable cancer risk. The corollary is procedural. An irregular Z line with tongues under one centimeter is not biopsied at all. A candidate who biopsies an irregular Z line at the gastroesophageal junction and labels the resulting goblet cells as Barrett has made the wrong move. There is no Barrett below the one-centimeter threshold to find.
The requirement for goblet cells is the point of contention between the American and British frameworks and the favored test of definitional precision on the boards. The British Society of Gastroenterology accepts cardiac, non-goblet columnar mucosa above the gastroesophageal junction as Barrett. The argument is that cardiac mucosa is itself a metaplastic response to reflux, can express the intestinal markers CDX2 and villin, and accumulates similar DNA-content abnormalities. The American framework requires goblet cells, because the published cancer-risk estimates the surveillance ladder relies on came from cohorts in which goblet-cell metaplasia was the entry criterion. The boards expect the American definition. Salmon mucosa over one centimeter above the gastroesophageal junction plus intestinal metaplasia with goblet cells on biopsy.
Once the diagnosis is on the table, the Prague C and M criteria quantify extent. C is the maximal circumferential extent of the columnar segment in centimeters, measured from the gastroesophageal junction to the most proximal point at which the salmon lining still extends fully around the lumen. M is the maximal extent of the columnar segment, including any tongue that extends above the circumferential portion. A Barrett segment described as C three M six has a circumferential portion three centimeters tall, with the longest tongue reaching six centimeters above the gastroesophageal junction. Interobserver agreement for the Prague system is good for segments over one centimeter and poor for segments under one centimeter. That measurement problem is part of the reason ultra-short Barrett has historically been treated as a separate clinical entity. Long-segment Barrett is conventionally three centimeters or longer. Short-segment Barrett is under three centimeters. Ultra-short Barrett is under one centimeter.
The Seattle biopsy protocol is the sampling strategy that compensates for the patchy distribution of dysplasia inside a Barrett segment. The logic is that dysplasia is focal and the visible appearance of the segment underestimates its histologic complexity. The exam starts with careful inspection of the entire columnar lining under high-resolution white-light endoscopy with either dye-based or virtual chromoendoscopy. Every visible nodule or irregularity is sampled separately, and ideally resected by endoscopic mucosal resection rather than forceps-biopsied. We will pick up the resect-first principle in episode two. The remainder of the segment gets four-quadrant random biopsies at two-centimeter intervals in non-dysplastic Barrett, and at one-centimeter intervals once dysplasia has been documented. For short Barrett segments under four centimeters, the practical adaptation is at least eight random biopsies plus one biopsy per centimeter of any tongue.
The Seattle protocol is imperfect, because even four-quadrant biopsies sample only a small fraction of the mucosal surface. That is why visible-lesion targeting and chromoendoscopy add yield, and why Barrett surveillance demands patience. A careful exam takes more than one minute per centimeter of segment, and shorter examinations measurably miss more dysplasia and cancer. A nine-minute look at a six-centimeter segment is not adequate.
Two adjuncts deserve recognition because the vignette will name them. Wide-area transepithelial sampling with three-dimensional computer-assisted analysis, WATS-3D, uses an abrasive cytology brush to sample the entire columnar segment. A neural network reassembles the disaggregated cells into three-dimensional images that flag possible foci of intestinal metaplasia or dysplasia. In the largest prospective series, WATS added to forceps biopsies increased Barrett detection from roughly sixteen to thirty-three percent and roughly doubled the dysplasia yield. The ASGE supports WATS as an adjunct to the Seattle protocol. The 2022 ACG guideline declined to make a recommendation. WATS is a supplement, not a replacement.
Non-endoscopic cell-collection devices are the modern alternative to indiscriminate screening EGD. The Cytosponge is a gelatin capsule containing a compressed mesh sponge attached to a string. The capsule dissolves in the stomach, the sponge is withdrawn through the mouth, and the cytology specimen is analyzed for trefoil factor three as a marker of intestinal metaplasia. Optional p53 staining adds dysplasia risk information. EsoCheck is an encapsulated balloon device with a textured surface that is dragged through the distal esophagus, with the EsoGuard assay reading methylation of vimentin and cyclin A one by next-generation sequencing. Sensitivity is roughly eighty to eighty-five percent and specificity around ninety percent. On the boards, the role is recognition. A vignette describing population-scale Barrett screening or a primary-care setting points to a non-endoscopic device rather than universal endoscopy.
Risk factors are the gating criteria for screening, because the general-population prevalence is too low to justify universal endoscopy. The 2022 ACG guideline supports a single screening EGD for patients with chronic GERD symptoms plus at least three additional risk factors. The factors are male sex, age fifty or older, white race, central obesity, tobacco smoking, and a first-degree relative with Barrett or esophageal adenocarcinoma. Chronic GERD is operationalized as weekly heartburn or regurgitation for five or more years. The AGA Clinical Practice Update considers GERD a risk factor rather than a strict requirement and would screen patients with three or more risk factors regardless of reflux history. Central obesity is defined by waist circumference over forty inches in men and thirty-five in women, or waist-to-hip ratio over zero point nine in men and zero point eight in women. Central adiposity drives reflux and Barrett independent of body mass index, which is why the criterion uses waist circumference rather than BMI.
The limitations of screening are real and the boards have started to reward candidates who can articulate them. Roughly forty percent of patients who present with esophageal adenocarcinoma have no heartburn history. Cohort data suggest fifty-five percent of incident adenocarcinoma patients would not have qualified for screening under the ACG criteria. Screening has not proven a mortality benefit in randomized data, and observational surveillance cohorts are inflated by lead-time and length-time bias. The candidate should still answer the criteria as written, while recognizing the field is shifting toward broader, cheaper, non-endoscopic tools.
One adjacent point belongs here, because it sits at the intersection of the prior chapter and this one. LA grade C and LA grade D erosive esophagitis demand a follow-up EGD after proton-pump-inhibitor healing, precisely because acute inflammation can mask underlying Barrett. Long-segment Barrett is itself definitive objective evidence of GERD that obviates reflux monitoring under Lyon Consensus version two. Those are pointers from the GERD chapter that fold into the Barrett evaluation.
Move now to pathogenesis, because the cancer risk that comes next is graded by where the tissue sits on a histologic sequence. Barrett is the price the distal esophagus pays for chronic acid and bile injury. The sequence opens with refluxed acid and pepsin destroying squamous epithelium at the gastroesophageal junction. That exposes progenitor cells that repopulate the ulcerated surface under conditions that reprogram their fate. Several progenitor candidates exist. Basal cells of the squamous epithelium, glands of the cardiac mucosa, submucosal gland duct cells, and a transitional basal cell population at the gastroesophageal junction have all been mapped. Ongoing reflux induces transcription-factor reprogramming, and the wound heals not as squamous mucosa but as specialized intestinal metaplasia with goblet cells. Cardiac, non-goblet columnar mucosa is increasingly seen as an acquired metaplasia in many cases. It expresses CDX2 and villin and is best understood as a precursor lesion to intestinal metaplasia rather than a normal variant.
The histologic sequence the boards test runs squamous to columnar to intestinal metaplasia to low-grade dysplasia to high-grade dysplasia to intramucosal carcinoma to invasive adenocarcinoma. Bile acids and acid in combination drive the metaplasia and the dysplasia by inducing oxidative DNA damage, double-strand breaks, p53 mutations, and a Th2-leaning cytokine environment. Two carcinogenetic routes have been described. The first is a slow, traditional pathway of progressive tumor-suppressor loss involving CDKN2A, SMAD4, and p53, followed by oncogene amplification. The second is a faster genome-doubled pathway, in which an early p53-mutant clone undergoes whole-genome doubling to a tetraploid state that tolerates further mutation. Roughly sixty percent of Barrett-derived cancers travel the genome-doubled route. That route can accelerate from non-dysplastic mucosa to invasive cancer faster than the conventional three-to-five-year surveillance interval can catch, which is part of why surveillance ever fails.
Dysplasia is the histologic read that grades the cancer risk. Pathologists call dysplasia when they see nuclear enlargement, pleomorphism, hyperchromatism, stratification, atypical mitoses, loss of cytoplasmic maturation, and crowding of tubules and villiform surfaces. Low-grade dysplasia preserves surface maturation. High-grade dysplasia loses it. The category is imperfect because interobserver agreement among general pathologists is poor, especially for low-grade dysplasia. The same biopsy can be read as non-dysplastic, indefinite for dysplasia, or low-grade dysplasia by three different pathologists. The cohort that anchors this point is Duits, who looked at two hundred ninety-three community-diagnosed low-grade dysplasia biopsies and had them re-read by an expert gastrointestinal pathologist. Only twenty-seven percent of the community low-grade calls were confirmed. Fourteen percent were downgraded to indefinite. Fifty-nine percent were re-read as non-dysplastic.
The clinical translation is non-negotiable and the boards test it relentlessly. Any dysplasia call must be confirmed by a second pathologist with gastrointestinal expertise before management changes. The candidate who orders radiofrequency ablation on a community low-grade dysplasia call without expert confirmation gets the question wrong, because the underlying read is wrong in over half of those cases.
Now the cancer-risk ladder, which is the number set to memorize, because the surveillance intervals are pulled off it. Non-dysplastic Barrett carries an annual progression rate to esophageal adenocarcinoma of roughly zero point two to zero point five percent. Confirmed low-grade dysplasia carries an annual progression rate over zero point five percent, with most cohorts clustering around one percent per year. The behavior of low-grade dysplasia approaches that of high-grade dysplasia as the inter-pathologist agreement strengthens. More confirmation, more progression. High-grade dysplasia in the absence of treatment carries an annual progression to cancer of roughly five to eight percent, with some older series reaching ten percent.
That ladder continues into T1 disease, which the boards test alongside dysplasia because the management lanes are continuous. Intramucosal carcinoma, T1a, is divided into m one, m two, and m three. The m one lesion is high-grade dysplasia confined to the epithelium. The m two lesion is intramucosal carcinoma invading lamina propria but sparing the muscularis mucosae. The m three lesion invades but does not penetrate the muscularis mucosae. Nodal metastasis risk is under two percent for m one and m two and four to seven percent for m three. T1b crosses into the submucosa and changes the calculus, and here the nodal risk climbs fast. Sm one, under five hundred micrometers of submucosal invasion, carries roughly thirteen percent nodal risk. Sm two carries roughly twenty-six percent. Sm three carries up to sixty-seven percent. We will use that ladder in episode two to decide who stays endoscopic and who goes to surgery.
A few modulators of progression within a Barrett patient are worth holding. Male sex, tobacco smoking, every additional centimeter of Barrett length, and confirmed low-grade dysplasia all push the annual progression rate upward in the Parasa model. Low-risk groups run around zero point one to zero point two percent per year. High-risk groups run around two percent per year. Tissue biomarkers are increasingly relevant. Aberrant p53 immunostaining is consistently associated with progression to high-grade dysplasia or adenocarcinoma and is the single most validated tissue biomarker. The Tissue Systems Pathology Test 9, marketed as TissueCypher, integrates nine biomarkers including p53 into a quantitative score. A high-risk TSP-nine result carries a positive predictive value of twenty-five to fifty percent and a negative predictive value around ninety-six percent. The 2022 AGA Clinical Practice Update endorses it as a risk-stratification tool in non-dysplastic Barrett.
One epidemiologic point belongs here because it explains why screening alone will not solve the problem. Roughly half of patients diagnosed with esophageal adenocarcinoma have no detectable Barrett mucosa at the time of cancer diagnosis. Cohort data suggest the cancer arises from a Barrett field in roughly forty-six percent of cases. The intestinal-metaplasia-positive cancers behave better than the intestinal-metaplasia-absent cancers, with longer survival in matched cohorts. That fits the field-cancerization model. The visible Barrett extent reflects the size of the metaplastic field and the surveillance opportunity, while metaplasia-absent cancers have likely overgrown or replaced the precursor lining by the time of presentation.
Chemoprevention has been tested formally. The AspECT trial randomized two thousand five hundred fifty-seven Barrett patients to high-dose esomeprazole forty milligrams twice daily or low-dose esomeprazole twenty milligrams daily, with or without aspirin three hundred milligrams daily. Median follow-up was roughly nine years. High-dose proton pump inhibitor protected against a composite of all-cause mortality, high-grade dysplasia, and esophageal adenocarcinoma. Aspirin protected against the same composite when patients on other NSAIDs were censored, with apparent additive effect. The results were not robust enough to mandate routine high-dose proton pump inhibitor plus aspirin for every Barrett patient. The calibrated board answer is once-daily proton pump inhibitor for all Barrett patients per ACG, with aspirin and high-dose proton pump inhibitor as individualized considerations. Antireflux surgery is not advised for cancer prevention alone, because the data do not support fundoplication over proton pump inhibitor for adenocarcinoma protection.
Now the surveillance algorithm, which is where the dysplasia ladder becomes an interval. The principle is straightforward. The higher the dysplasia grade, the higher the annual progression rate, and the shorter the interval the algorithm permits before the next look. The candidate should be able to walk a vignette through the algorithm without hesitation, because the interval is what the question is testing.
Non-dysplastic Barrett under the 2022 ACG framework is surveyed every five years for short segments under three centimeters and every three years for long segments of three centimeters or more. The earlier guidance combined the two into a three-to-five-year window, but ACG 2022 separates them based on the higher progression risk of long segments. The first surveillance interval after a new Barrett diagnosis is typically one year, even when the index biopsies are non-dysplastic. The reason is the highest-yield interval for finding prevalent dysplasia or cancer that was missed at the index endoscopy is the first follow-up. A meta-analysis of fifteen non-dysplastic Barrett series found that twenty-four percent of cancers were diagnosed within the first year of the index endoscopy. Roughly eighty percent of cancers and high-grade dysplasia were detected at or within the first year. Once that first surveillance pass is clean, the patient enters the three-year long-segment or five-year short-segment cycle.
Indefinite for dysplasia is a real diagnostic category and the management is mechanism-driven. The indefinite label reflects pathologist uncertainty between reactive change from acid injury and true dysplasia. The acid-suppressed mucosa looks bland; the inflamed mucosa looks atypical in ways that mimic dysplasia. The maneuver is to optimize acid suppression with twice-daily proton pump inhibitor for three to six months and then repeat the EGD with biopsies. Most indefinite reads resolve to non-dysplastic on the repeat sampling under good acid control. If the indefinite call persists, surveillance shifts to a twelve-month interval. A vignette with mild esophagitis and an indefinite-for-dysplasia biopsy is asking for twice-daily proton pump inhibitor for three to six months and a repeat EGD. It is not asking for ablation.
Low-grade dysplasia is the shared-decision branch, which is the heart of the surveillance episode and the most-tested decision point in this material. The first move after a community low-grade dysplasia call is expert confirmation. A second gastrointestinal pathologist confirms or downgrades the read, because most community low-grade calls do not survive expert review. Once an expert pathologist confirms low-grade dysplasia, the patient has two acceptable management paths under the 2022 ACG guideline. The first is endoscopic eradication therapy, meaning resection of any visible lesion followed by radiofrequency ablation of the field. The second is continued surveillance every twelve months. The data favor endoscopic eradication in patients with reasonable life expectancy, because randomized trials show radiofrequency ablation reduces progression to high-grade dysplasia and cancer. Surveillance is acceptable for older patients, patients with limited life expectancy, or patients who decline ablation after counseling. The first repeat EGD after a low-grade dysplasia call is typically at six months on twice-daily proton pump inhibitor, because low-grade dysplasia that disappears on a clean acid-suppressed repeat exam was likely overcalled. Low-grade dysplasia that persists across two examinations is the disease that earns ablation. Patients on the surveillance arm get EGD at six and twelve months and then yearly.
High-grade dysplasia is treated with endoscopic eradication therapy in essentially all patients with reasonable life expectancy. The mechanism is two-fold. The annual progression rate to invasive cancer is five to eight percent. Ten to twenty percent of high-grade dysplasia biopsies in surveillance series already harbor occult cancer at the time of diagnosis. The pathway opens with a careful re-examination by an experienced endoscopist using chromoendoscopy and four-quadrant biopsies at one-centimeter intervals. Any visible nodule is resected by endoscopic mucosal resection, with endoscopic submucosal dissection in larger or higher-suspicion lesions, for staging. The remaining flat high-grade dysplasia is ablated. Esophagectomy was the historical answer and remains an option for multifocal high-grade dysplasia plus deep submucosal invasion that cannot be cleanly resected endoscopically, or for failures of endoscopic eradication. An eighty-year-old with a six-centimeter segment of high-grade dysplasia confirmed by an expert pathologist and no visible nodule does not need esophagectomy. The answer is endoscopic eradication.
Intramucosal carcinoma, T1a, follows the same endoscopic-first logic in most patients. T1a m one and m two have under two percent lymph node metastasis risk. They are reliably handled by endoscopic mucosal resection or submucosal dissection of the visible lesion, followed by ablation of the flat segment. T1a m three carries four to seven percent nodal risk and is still managed endoscopically in most cases, with multidisciplinary review for borderline patients. T1b submucosal invasion is the threshold where surgical resection enters the conversation, and we will pick that up in episode two.
Two final points sit at the edges of the surveillance ladder. The first is when to stop. Modeling work by Omidvari suggests stopping non-dysplastic Barrett surveillance at age eighty-one in healthy men, seventy-three in men with severe comorbidity, seventy-five in healthy women, and sixty-nine in women with severe comorbidity. The principle is that surveillance only benefits patients with enough life expectancy to develop and treat a cancer detected on a surveillance pass. Patients beyond those thresholds gain nothing from continued endoscopy. The second point is BOSS, the trial the boards have started to reward. BOSS randomized three thousand four hundred fifty-two Barrett patients to two-yearly endoscopic surveillance versus EGD on demand for symptoms. Over a minimum ten-year follow-up, there was no improvement in overall or cancer-specific survival and no difference in stage at cancer diagnosis. The translation is that the surveillance dogma has limits, and a candidate who can articulate them is rewarded over one who insists on rigid adherence.
Pull the arc together. Barrett is the conjunction of a visible columnar segment one centimeter or more above the gastroesophageal junction and intestinal metaplasia on biopsy. The Prague C and M criteria quantify the segment. The Seattle protocol samples it, with at least one minute of inspection per centimeter and four-quadrant biopsies at two-centimeter intervals in non-dysplastic disease and at one-centimeter intervals once dysplasia is present. Visible nodules are not forceps-biopsied; they are resected. Pathogenesis is reflux-driven metaplasia, then a histologic dysplasia ladder, then cancer. The annual progression numbers are roughly zero point two to zero point five percent for non-dysplastic Barrett, around one percent for confirmed low-grade dysplasia, and five to eight percent for high-grade dysplasia. Those numbers set the surveillance intervals. Non-dysplastic short segments get every five years. Non-dysplastic long segments get every three years. Indefinite and surveillance-pathway low-grade dysplasia get every twelve months after acid-suppression confirmation. Confirmed low-grade dysplasia in patients with life expectancy and essentially all high-grade dysplasia enter endoscopic eradication. Any dysplasia call needs a second gastrointestinal pathologist before management changes.
The next two episodes pick up where this one stops. Episode two takes the patient who needs endoscopic eradication through a stack of techniques with a fixed sequence. Visible disease is resected first, by endoscopic mucosal resection or submucosal dissection, because the resection specimen is the staging test. The flat Barrett field is ablated second, with radiofrequency ablation as the workhorse. Episode three takes the cancer side, which picks up at the T1a versus T1b boundary. That is the line where lymph node risk crosses from under two percent to roughly thirteen percent, and where the patient stops being endoscopic and starts being surgical.