Endoscopic ultrasound‐guided needle‐based confocal laser endomicroscopy in gastrointestinal subepithelial lesions: Feasibility study

Abstract

Background and Aim: Needle-based confocal laser endomicroscopy (nCLE) allows for real-time optical biopsies during endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA). Little is known about nCLE imaging of gastrointestinal subepithelial lesions (GI-SEL); therefore, we determined its feasibility.Methods: We carried out EUS, nCLE, and finally FNA in 25 patients with GI-SEL between November 2015 and December 2018. We retrospectively compared nCLE findings with pathological findings of EUS-FNA or surgical specimens. For concordance analysis, two endoscopists independently validated representative nCLE images 5 months or more after examinations.Results: Adequate sample acquisition rate of EUS-FNA was 67% per needle pass and 96% per patient. EUS-FNA was diagnostic in 80% (20/25), suspicious in 4% (1/25), and nondiagnostic in 16% (4/ 25). nCLE image acquisition rate was 100% and its concordance rate with final pathology was 88% (22/25), which was not significantly different from diagnostic and suspicious EUS-FNA. nCLE could differentiate GI stromal tumors (GISTs) from leiomyoma, in that GISTs were characterized by contrast-enhanced densely populated spindle cell tumors with unenhanced rodshaped nuclei in 93% of 14 patients, whereas leiomyomas were characterized by narrower spindle cell tumors with fewer and smaller unenhanced nuclei in 100% of three patients. In rectal metastasis from lung adenocarcinoma, some pleomorphic dark nests were observed. At concordance analysis between the two endoscopists' validation results, j value was 0.560 (P < 0.001), indicating moderate agreement. There were no adverse events associated with nCLE and EUS-FNA.Conclusion: Needle-based confocal laser endomicroscopy can be safe and useful for on-site detection of abnormalities of GI-SEL (UMIN 000013857).

Current Status of Needles in the Optimization of Endoscopic Ultrasound-Guided Procedures

“…A recent prospective multicenter study of 206 patients with pancreatic cystic lesions showed thay sensitivities and specificities of nCLE for the diagnosis of serous cystadenomas, mucinous cystic neoplasms, and premalignant cystic lesions were all 0.95 or higher; the area under the receiver operating characteristic curve was significantly larger for nCLE than for cyst fluid carcinoembryonic antigen or EUS [ 80 ]. A number of studies also evaluated the feasibility and safety of nCLE for the assessment of solid organs (i.e., solid pancreatic lesions, lymph nodes, and SELs) [ 81 , 82 , 83 ]. Although in vivo real-time imaging using nCLE is highly concordant with pathology and may replace ROSE, it is important to keep in mind that EUS-FNA alone has a sensitivity around 90% for malignancy.…”