features into repeating patterns or irregular and distorted areas and (ii) vascular patterns into regular or irregular. NBI has been extensively studied in BE. A cohort of international expert physicians forming the BING (Barrett’s International NBI Group) working group examined 50 NBI images followed by an additional 120 images to determine the validity of the simple binary classification of pit and vascular patterns as regular or irregular [10]. The group found a sensitivity of 80.4% and specificity of 88.4%. A recent report using blue‐light imaging (BLI) (Fujifilm, Tokyo, Japan) confirmed similar findings in 100 patients undergoing evaluation of suspected BE dysplasia; careful inspection of the color, mucosal pits, and crypts and vessels was able to reliably identify areas of neoplasia [11].
The i‐SCAN optical enhancement (OE) system (Pentax, Tokyo, Japan) has been studied in a recent prospective clinical trial of 41 patients with BE comparing high‐definition white light magnified using still images marked by seven trainees and seven experts [12]. The design of the study addressed the bias of many previous clinical trials using only experienced physicians. Mucosal and vascular changes were recorded in a standardized fashion, and neoplasia was defined by two unblinded experts by consensus. The OE system improved sensitivity, specificity, and positive predictive value in both trainees and experts. The authors also noted increased inter‐observer agreement, and these changes appeared clinically significant.
The Fuji Intelligent Chromoendoscopy System (FICE, Japan) has limited data but also shows improvement in detecting neoplasia in BE compared to high‐definition and AA [13].
Confocal laser endomicroscopy (CLE)
Confocal laser endomicroscopy (CLE) utilizes a low‐power laser light‐emitting source, pinhole filter, and lens to literally focus on cellular features in the most superficial layers of tissue. The narrowly focused parallel light penetrating the tissue and reflectance from the tissues are ideally suited to examine tissue perpendicular to the objective lens; however, this can be difficult in the tangential alignment we often find ourselves in with the esophagus. This technology can be termed an “optical biopsy” due to its ability to provide special resolution comparable to histologic examinations. The earliest attempts to visualize tissue without contrast agents found poor image resolution compared to contrast‐enhanced CLE. Fluorescein 10% contrast dosing is not standardized and varies from 2.5 to 10 mL intravenously immediately before the imaging. This technology has developed through two strategies: in probe (pCLE) catheter‐based equipment from Cellvizio (Mauna Kea Technologies, Paris, France), and integrated into specialized endoscopes (eCLE) from OptiScan (Victoria, Australia; and Pentax, Tokyo, Japan). The integrated endoscope system may not be commercially viable. Reusable endoscopic probes for the esophagus and stomach provide resolution to 1 μm and have a maximum of 20 uses. Other commercially available probes are designed for colonic, biliary pulmonary, urinary epithelium, and pancreatic cysts. Probe‐based CLE education is provided with online interactive programs in BE, gastric diseases, pancreatic cysts, and inflammatory bowel disease at www.Cellvizio.net. Real‐time imaging with CLE requires a dedicated infrastructure to support the time, equipment, and expertise needed for this advanced imaging. The payoff appears to be high yield of neoplasia with CLE‐targeted biopsy (34%) compared to with high‐definition random biopsy (7%) [14]. In this prospective randomized study of 192 patients using endoscope‐based CLE, the authors suggested that up to 65% of patients referred for surveillance would not need a traditional histologic biopsy. The ASGE Technology Committee reviewed the data and recommended CLE as an emerging technology with the potential to significantly reduce the number of biopsies in BE; however, further studies are necessary to evaluate the clinical efficacy and cost‐effectiveness before utilization in both academic and community settings [15].
Figure 7.1 VLE Barrett’s esophagus case 1: 63‐year‐old with random surveillance biopsies reported indefinite for dysplasia and no focal abnormalities on high‐definition white‐light endoscopy or NBI; had VLE targeted endoscopic mucosal resection revealing high‐grade dysplasia. (A) NBI representative image; (B) VLE revealing abnormalities at 3 o’clock; (C) VLE targeted area detail; (D) laser markings at 3 o’clock.
Figure 7.2 VLE Barrett’s esophagus case 2: 70‐year‐old with random surveillance biopsies reported low‐grade dysplasia and no focal abnormalities on high‐definition white‐light endoscopy or NBI; had VLE targeted endoscopic mucosal resection revealing low‐grade dysplasia. (A) White light representative image; (B) NBI representative image; (C) VLE revealing abnormalities at 5 o’clock; (D) VLE targeted area detail; (E) laser markings at 4 o’clock.
Figure 7.3 VLE Barrett’s esophagus case 3: 60‐year‐old with random surveillance biopsies reported high‐grade dysplasia and no convincing abnormalities on high‐definition white‐light endoscopy and equivocal changes on NBI; had VLE targeted endoscopic mucosal resection revealing high‐grade dysplasia. (A) and (B) White light representative images; (C) VLE representative image revealing abnormalities at 11 o’clock; (D), (E), and (F) VLE targeted area detail; (G) NBI image with laser markings at 12 o’clock.
Volumetric laser endomicroscopy (VLE)
Volumetric laser endomicroscopy (VLE) uses second‐generation optical coherence tomography (OCT) technology (NvisionVLE, Merit Medical Systems, South Jordan, Utah, USA), a relatively new imaging modality that provides real‐time wide‐field endomicroscopy over the entire circumference of 6 cm of the esophagus in 90 seconds. The resolution up to 7 μm can reach a depth of 3 mm and allows for detailed microscopic imaging of the mucosa and submucosa. Unlike CLE, where microscopic intracellular detail is observed, VLE provides more detailed patterns for squamous epithelium, BE, and neoplasia. VLE allows for placement of superficial mucosal laser marks in abnormal areas that are identified, and thus targeted sampling or mucosal resection, as shown in Figures 7.1–7.3. The addition of this laser marking to the technology has been associated with increased dysplasia yield [16, 17]. Criteria for dysplasia detection in VLE have been developed in ex vivo; models and are currently used in practice for dysplasia detection [18]. New to VLE is artificial intelligence image enhancement software that can color‐code the VLE abnormalities associated with dysplasia to help the endoscopist identify abnormal areas to be subsequently laser marked [19]. The software color‐codes a hyper‐reflexive