Gastrointestinal junction under white light endoscopy (WLE) showing intestinal metaplasia consistent with Barrett’s esophagus with a nodule
Chromoendoscopy uses the topical application of dyes or contrast agents which enhance mucosal abnormalities during endoscopic evaluation. Methylene blue chromoendoscopy studies have shown varied reports of accuracy of dysplasia detection. Indigo carmine , alternatively, is a topical coloring agent which is not absorbed by the mucosa and highlights mucosal irregularities during endoscopic evaluation. It has been used along with magnification endoscopy, which correlates villiform pit patterns and irregular mucosal patterns with the presence of intestinal metaplasia and dysplasia . A recent meta-analysis showed that use of chromoendoscopy was associated with an increased rate of dysplasia detection. A total of 843 patients were included in this meta-analysis, and the use of chromoendoscopy or electronic chromoendoscopy (i.e., NBI) increased the diagnostic yield for detection of dysplasia by 34% (95% CI, 20–56%; P < 0.001) when compared with WLE . Acetic acid is a commonly available dye and has been used in the detection of Barrett’s dysplasia. Another recent study comparing neoplasia detection rates in 982 BE patients showed statistically higher detection rates (12.5% vs. 2%; P = 0.001) using acetic acid vs. random biopsies. There was a 6.5-fold gain in neoplasia detection within the acetic acid cohort as compared with the biopsy cohort .
This term refers to imaging techniques that enhance the contrast between the squamous and columnar mucosa in the esophagus and allows for a detailed examination of the surface mucosa and vasculature. Narrowband imaging (NBI) manipulates light wavelengths which highlight the superficial capillary network and subepithelial vessels, allowing for identification of subtle mucosal abnormalities. In a recent meta-analysis of eight studies including 446 patients and 2194 lesions, NBI demonstrated a pooled sensitivity and specificity of 95% and 65%, respectively, for the detection of Barrett’s esophagus . Additionally, the sensitivity and specificity of NBI in detection of high-grade dysplasia were 96% and 94%, respectively. Particular attention should be paid to microvascular or pit patterns, as NBI pit pattern classification schemes for Barrett’s esophagus have a high sensitivity and specificity for detection of dysplastic Barrett’s . Recently a universal classification system for the vascular and mucosal patterns seen with NBI has been standardized by the Barrett’s International NBI Group (BING) . High-confidence readings, or when dysplasia was identified with a high level of confidence, had a 92% overall accuracy, 91% sensitivity, 93% specificity, 89% positive predictive value, and 95% negative predictive value and lends promise to identifying high-grade dysplasia and esophageal adenocarcinoma more accurately . Electronic chromoendoscopy techniques including FICE and I-Scan have also been evaluated in patients with BE with promising results. A subgroup analysis comparing chromoendoscopy to electronic chromoendoscopy showed that both techniques increased the diagnostic yield of dysplasia detection when compared to WLD, though there was no significant difference between the two chromoendoscopic techniques . From a practical standpoint, however, electronic chromoendoscopy may be an easier tool to use, as it does not require the topical application of a chemical substance.
Confocal laser endomicroscopy (CLE) allows for histologic evaluation of GI mucosa during endoscopy using magnification up to 1000-fold and up to 250 μm below the mucosal surface . This level of magnification enables visualization of goblet cells and specialized intestinal metaplasia. Two platforms have been evaluated, an endoscope-based system (eCLE) which is integrated into the tip of a standard endoscope and a probe-based system (pCLE) which is passed through the accessory channel of the endoscope. Both use blue laser light and require intravenous fluorescein as a contrast agent. A prospective, multicenter, international study using pCLE in conjunction with HD-WLE and NBI-enabled identification of additional high-grade dysplasia/esophageal adenocarcinoma patients compared to HD-WLE or NBI demonstrated a sensitivity and specificity of identification of neoplasia with pCLE of 68.3% and 87.8%, respectively, compared to 34.2% and 92.7% with HD-WLE . Another study found that eCLE with targeted biopsies almost doubled the diagnostic yield for neoplasia (33%) compared to the standard biopsy protocol for BE (17%) . With the use of eCLE, there was a 59% statistically significant decrease in mucosal biopsies needed for diagnosis and avoidance of mucosal biopsies by patients undergoing surveillance endoscopies . While the advantages of CLE are apparent, such as real-time diagnostic capability with more accurate, targeted biopsies, there is an associated increase in procedure length, cost of procedure and equipment, and additional training involved to correctly interpret images.
Endoscopic Therapies for Barrett’s Esophagus
The AGA recommends endoscopic eradication therapy in cases of high-grade dysplasia, though it can be used in selected cases of low-grade dysplasia, as well . There is no recommendation to endoscopically treat non-dysplastic Barrett’s mucosa at this time. Although several therapeutic endoscopic techniques for eradication of dysplasia exist, as of yet, there are no head-to-head trials comparing the efficacy of each method. These techniques can be broadly categorized into those that provide tissue and those that ablate tissue. EMR and ESD provide viable tissue for histologic review, which can provide information about length and depth of dysplasia.
Endoscopic Mucosal Resection (EMR)
EMR (Fig. 13.2 ) removes the mucosal and submucosal layers of tissue by raising the targeted segment with submucosal injection followed by resection (cap EMR) or the suction, banding, and cut method (multi-band ligation). The band ligation technique has the advantage of allowing for multiple resections within a single endoscopic session and does not always require submucosal fluid injection. Both EMR techniques have been compared, showing that the cap EMR technique provides larger tissue samples, while band ligation is faster and less expensive when performing multiple resections .
Post-resection view of nodular lesions removed by endoscopic mucosal resection (EMR)
Use of EMR may be applied in the definitive therapy of dysplastic and early-stage (T1 N0) neoplasias with limited submucosal invasion and can be used for staging prior to endoscopic resection . In a prospective study of 107 patients with suspected HGD or adenocarcinoma who underwent complete EMR with a mean follow-up time of 40.6 months, BE was e radicated completely in 80.4% of patients with 71.6% of patients with clearance of intestinal metaplasia and 100% in complete remission from HGD . Recurrence rates for both HGD and cancer were 1.4%.
EMR has been shown to be relatively safe, as well. The rate of significant bleeding post-EMR in a single-center study of 681 patients who underwent 2513 EMRs was 1.2% . In physicians experienced in performing EMR, reported perforation rates are less than 0.5% .
EMR and RFA can be combined, and recent data suggest that this can be done safely within a single session. A recent retrospective analysis of 40 patients with short-segment (median C1M2), early BE neoplasia, of which 68% of patients had invasive carcinoma, who were treated with combined EMR followed by RFA in a single session demonstrated complete remission of all neoplasia and intestinal metaplasia in an intention-to-treat analysis of 95% . Most patients underwent subsequent focal RFA sessions every 2–3 months until a median follow-up of 19 months but one single-session treatment resulted in complete histologic remission intestinal metaplasia in 43% of patients. Esophageal stricture occurred in 33% of cases and was successfully treated with a median of two dilations.
Endoscopic Submucosal Dissection (ESD)
Endoscopic submucosal dissection is a technique mostly used in Japan for the treatment of gastric neoplasia. It involves submucosal injection of fluid, followed by an incision using various cutting devices, and finally submucosal dissection of the segment . The theoretical advantage of this method is that it allows the opportunity to remove a large area of cancerous mucosa en bloc with determination of lateral and vertical margins. However, this is technically difficult, especially in the esophagus, requiring several hours to complete with the potential for serious complications such as perforation.
EMR vs. ESD
A recent randomized controlled trial looking at efficacy and safety in patients with HGD or early esophageal adenocarcinoma who underwent EMR or ESD observed high rates of complete resection in those in the ESD group (58%, P = 0.01) than in the EMR group (11%) . Though no difference in complete remission of intestinal metaplasia (CRIM) was seen in either arm at 3 months, recurrent EAC was seen in one case in the ESD group during a mean follow-up of 23.1 ± 6.4 months . Two esophageal perforations were noted in the ESD group (11%) while none were observed in the EMR group . This study showed that EMR is both safe and effective in eradicating dysplasia without any significant clinical advantage over ESD .
Radiofrequency Ablation (RFA)
RFA employs radiofrequency energy that is delivered either by an endoscopic balloon catheter or a focal ablation device (Fig. 13.3 ) to eradicate intestinal metaplasia. The balloon catheter spans 3 cm and the focal ablation device can ablate non-circumferential segments up to 2 cm at a time. A study comparing RFA vs. sham over 1 year showed high rates of complete eradication in both high- (81%) and low-grade dysplasia (90.5%) patients in the ablation group . Results from the US registry and UK registry have shown improved clearance of dysplasia (CRD) and CRIM in patients undergoing EMR/RFA as well as efficacy and safety of using both treatments in tandem. The US patient registry study treated patients with nodular BE with EMR before RFA and patients with non-nodular BE with RFA alone. Between the two groups, complication rates (i.e., bleeding, stricture, and hospitalizations) were not significantly different (8.4% in the EMR/RFA group vs. 7.2% in the RFA-only group) . CRIM was achieved in 84% of patients treated in both groups (P = 0.96) and CRD was achieved in 94% and 92% of patients in the EMR/RFA and RFA-only group, respectively (P = 0.17) . In the UK patient registry study, HGD was eradicated in 86% of patients, all dysplasia in 81%, and BE in 62% by 12 months after a mean of 2.5 RFA sessions . After 19 months of therapy, 94% of these patients did not experience recurrence . It was also noted that shorter segments of BE responded better to RFA and complete reversal of dysplasia was 15% less likely per each 1 cm increment of BE .