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 two of three of the Barrett Esophagus and Esophageal Cancer chapter, in the Esophageal Disorders module. In this episode we cover endoscopic eradication therapy: visible disease is resected first because the specimen is the staging test, the flat Barrett field is ablated second with radiofrequency ablation as the workhorse, and complete eradication of intestinal metaplasia is the endpoint that still requires structured surveillance. The esophageal cancer staging and treatment that follow come in episode three.
Start with the principle that organizes endoscopic eradication therapy, because everything else is downstream of it. The Barrett segment contains two kinds of tissue that need two different treatments. Visible nodules, plaques, and any irregular mucosal abnormality represent neoplasia of unknown depth. Flat metaplasia represents the field from which future neoplasia will arise. Visible disease gets resected, because resection produces a histologic specimen that tells you how deep the tumor invaded. Flat metaplasia gets ablated, because ablation removes the precursor mucosa across a wide area without needing to deliver a specimen.
The sequence is not optional. Resect first, ablate second. Ablating a visible nodule destroys the specimen and removes the staging information you need to decide whether the patient stays endoscopic or goes to surgery. A vignette with a Barrett segment, a one-centimeter nodule, and confirmed high-grade dysplasia on biopsy is not a case for upfront ablation. The next step is endoscopic mucosal resection of the nodule, and only after the pathology returns does the field get ablated.
EMR is the workhorse resection technique for Barrett neoplasia. Two methods predominate. The suck-and-cut method elevates the lesion with a submucosal saline injection, suctions the elevated mucosa into a cap on the scope, and deploys a snare around the captured tissue. The band-and-cut method ligates the mucosa with an elastic band, much like a variceal band, and snares the resulting pseudopolyp. Either way the specimen carries the dysplastic or neoplastic mucosa together with the lamina propria and superficial submucosa. That is enough tissue for the pathologist to read depth of invasion, lateral and deep margins, and lymphovascular invasion.
The reason EMR is the staging step, not just a treatment step, is the rate at which it changes the diagnosis. EMR alters the histologic read in roughly a third to a half of cases compared with the prior forceps biopsy. The shift is asymmetric. Roughly two-thirds of the changes are downgrades, where high-grade dysplasia on the forceps biopsy turns out to be low-grade dysplasia or non-dysplastic on the resection specimen. Roughly one-third are upgrades, where the forceps biopsy underestimated the disease and the resection specimen shows deeper invasion. That asymmetry matters in two directions. It explains why so many patients referred for esophagectomy on a forceps high-grade read end up needing only ablation after EMR. It also explains why a small but real number of patients turn out to have submucosal invasion that the forceps biopsy missed.
EMR comfortably handles lesions up to one and a half to two centimeters in a single piece. Larger lesions can be removed piecemeal, but piecemeal resection sacrifices margin assessment because the lesion is reassembled across multiple specimens with fragmented borders. That limitation is where endoscopic submucosal dissection enters. ESD uses specialized diathermic knives to lift, mark, and dissect a target lesion en bloc, no matter the size, with clean lateral and deep margins. The advantage is the unbroken specimen, which gives the pathologist a definitive margin read and is critical when the lesion is over two centimeters or carries Paris features suggesting deeper invasion. Paris zero-I-Ic, the slightly depressed lesion, and Paris zero-I-Ia plus Ic, the elevated lesion with a central depression, are the morphologies that should push toward ESD over piecemeal EMR. The disadvantages of ESD are longer procedure times, higher perforation risk, and a steeper operator learning curve. That is why the AGA guidance still positions EMR as adequate for most Barrett neoplastic lesions and reserves ESD for the larger or higher-suspicion subset.
Once visible disease is resected and the pathology confirms an endoscopically curable stage, the flat Barrett field gets ablated. Radiofrequency ablation is the field-ablation workhorse and the preferred ablative modality in current ACG guidance. The mechanism is a controlled thermal injury delivered at approximately twelve joules per square centimeter in the standard balloon protocol. That energy density is calibrated to destroy the columnar epithelium and the immediate subepithelium without injuring the muscularis propria, which lets the wound heal as squamous mucosa under proton pump inhibitor cover. Two device formats are used. A three-hundred-sixty-degree balloon handles circumferential disease. A focal device, available in ninety-degree or sixty-degree variants, handles residual tongues, islands, and the post-ablation gastroesophageal junction. Patients typically need two to four sessions at two to three month intervals to clear the entire field, with sessions continued until no Barrett remains on inspection or biopsy.
The randomized evidence behind RFA is the Shaheen sham-controlled trial, which is the canonical citation the boards expect a candidate to recognize. The trial randomized one hundred twenty-seven patients with dysplasia in Barrett to RFA versus sham. Here are the two numbers to carry. At one year, complete eradication of low-grade dysplasia was achieved in roughly ninety percent of RFA patients versus twenty-three percent of sham patients. Complete eradication of high-grade dysplasia was achieved in roughly eighty-one percent of RFA patients versus nineteen percent of sham patients. Progression to higher-grade neoplasia and incident cancer were both reduced in the RFA arm. The translation for the boards is that RFA after EMR of visible disease is the modern standard for high-grade dysplasia and most intramucosal cancer. Real-world series confirm dysplasia eradication rates in the high eighties to low nineties. Intestinal metaplasia eradication rates run in the upper seventies to upper eighties, with cancer risk reduced roughly tenfold compared with surveillance alone.
The randomized evidence for RFA in low-grade dysplasia specifically is the SURF trial, which required expert pathologist confirmation as the inclusion criterion. SURF showed an absolute reduction in progression to high-grade dysplasia or adenocarcinoma of about twenty-five percentage points over three years compared with surveillance. The condition was expert confirmation, which is the procedural point. Low-grade dysplasia called by a community pathologist downgrades to non-dysplastic in over half of cases on expert review. A community low-grade dysplasia read is not adequate grounds for ablation.
Cryotherapy is the alternative ablative modality. Two device formats are in use. Spray cryotherapy delivers liquid nitrogen or carbon dioxide directly onto the mucosa through a catheter. Balloon cryotherapy delivers nitrous oxide through a controlled balloon contact. The mechanism in either format is rapid freezing and thawing that causes intracellular ice crystal formation and microvascular thrombosis with tissue necrosis. Cryotherapy achieves dysplasia eradication around ninety-seven percent and intestinal metaplasia eradication around ninety-one percent in published series. The role on the boards is as a salvage option after RFA failure. It is also used in patients with esophageal stenosis or prior RFA stricture, in patients with very long segments that ablate poorly, and in patients who do not tolerate RFA. Cryotherapy tends to produce less post-procedural pain and less stricturing. Argon plasma coagulation has a residual role for isolated islands of metaplasia after RFA. Photodynamic therapy with porfimer sodium was the original Barrett ablation modality and is now obsolete because of high stricture rates and prolonged photosensitivity.
The endpoint of the eradication pathway is complete eradication of intestinal metaplasia. CEIM means no visible Barrett at endoscopy and no intestinal metaplasia on biopsies of the neosquamous epithelium and the gastroesophageal junction. CEIM is not cure. Recurrent intestinal metaplasia after CEIM occurs at roughly nine to ten percent per patient-year in pooled series. Most recurrences sit at or within one to two centimeters above the gastroesophageal junction, where ablation reach is limited and where ongoing acid exposure favors recolonization by columnar mucosa. Recurrent low-grade dysplasia runs around two percent per patient-year. Recurrent high-grade dysplasia or cancer runs around one percent per patient-year. Those numbers justify the post-CEIM surveillance schedule.
The schedule is more intensive than pre-treatment Barrett surveillance, because the highest recurrence yield is in the first year after eradication. For patients who entered eradication with high-grade dysplasia or intramucosal carcinoma, surveillance is at three months, six months, and twelve months in the first year, then every six months in year two, then annually thereafter. For patients who entered with low-grade dysplasia, ACG recommends endoscopy at one year and three years before reverting to longer intervals. At each post-CEIM exam the endoscopist takes four-quadrant biopsies from the gastroesophageal junction and from the distal two to three centimeters of neosquamous mucosa, plus targeted biopsies of any visible irregularity. The gastroesophageal junction is sampled because it is where recurrence concentrates. The neosquamous mucosa is sampled because of the buried-gland phenomenon, where residual intestinal metaplasia persists beneath regenerated squamous epithelium and is invisible on surface inspection. Buried glands on biopsy are not an artifact and not a normal post-CEIM finding. They represent incomplete eradication of the field and require directed therapy, usually endoscopic resection of the involved segment, because their continued cancer risk cannot be monitored by surface surveillance alone.
Failure of EET to achieve CEIM occurs in roughly ten to twenty percent of patients. The risk factors are long segments, persistent reflux, esophageal narrowing, and high-grade dysplasia or intramucosal carcinoma at presentation. Salvage options after RFA failure include cryotherapy, repeat EMR or ESD of any residual disease, and esophagectomy in selected patients with multifocal disease that cannot be cleanly resected. The complications the candidate should hold are predictable. Stricture is the most common, around five to six percent of RFA cases overall and rising to seven to eleven percent when EMR is performed before RFA. The risk rises when the resection covers more than half the circumference. Strictures respond to sequential balloon dilation in almost all cases. Bleeding occurs in roughly one percent and is managed endoscopically. Perforation is rare at around six tenths of a percent for RFA and somewhat higher for ESD and large EMR. Chest pain in the days after RFA is expected and is managed with topical anesthetics, viscous lidocaine, sucralfate, and high-dose acid suppression. Twice-daily PPI is mandatory throughout EET because acid exposure inhibits squamous re-epithelialization and increases recurrence.
Pull the eradication side together. The Barrett segment holds two kinds of tissue that need two different treatments. Visible disease is resected first, by EMR or ESD, because the resection specimen is the staging test that decides whether the patient stays endoscopic or goes to surgery. The flat field is ablated second, with radiofrequency ablation as the workhorse, backed by the Shaheen trial for dysplasia and the SURF trial for expert-confirmed low-grade dysplasia. Cryotherapy is the salvage modality after RFA failure. The endpoint is complete eradication of intestinal metaplasia, which is not cure, because recurrence concentrates at the gastroesophageal junction and in buried glands. That is why post-CEIM surveillance is most intensive in the first year and samples both the junction and the neosquamous mucosa, all under twice-daily PPI cover.
Episode three picks up where eradication ends, with the patient whose resected lesion turns out on EMR to have invaded past the mucosa, or who presents with cancer outside a known Barrett field. That is the esophageal cancer arc: the histology split between adenocarcinoma and squamous cell carcinoma, TNM staging, the T1a-versus-T1b boundary that separates the endoscopic patient from the surgical one, and the CROSS and FLOT regimens for locally advanced disease.