Probe-Based Confocal Laser Endomicroscopy For The Diagnosis Of Gastric Intestinal Metaplasia,Intraepithelial Neoplasia,and Carcinoma

Background and AimGastric cancer is one of the most common cancers of mankind. The prognosis for patients with this cancer clearly depends on the efficiency of diagnosis. Both pathological and epidemiological studies prove that gastric cancer is caused by a series of factors through several steps, from chronic nonatrophic gastritis, atrophic gastritis, gastric intestinal metaplasia(GIM), gastric intraepithelial neoplasia(GIN), finally to carcinoma. GIM and GIN are the precancerous lesions of gastric cancer. These lesions need to be followed closely, and identified by the endoscope and received treatment timely. This is conducive to the prevention and control of gastric cancer. However, the diagnosis of GIM through the conventional endoscopy usually has high inter-observer variability and a poor correlation with histopathological finding.As for GIN, since the lesion is superficial and within a narrow range, it is hard to detect the minor changes of gastric mucosa under the conventional endoscope whose resolution is low. As a result, GIN is likely to be misdiagnosed or never diagnosed. Histopathological examination is the golden standard for diagnosing GIM, GIN and carcinoma. The five-part biopsy suggested by the newly revised Sydney Standards is proven to be not efficient in detecting GIM, with such demerits as inaccuracy, causing injuries, requiring longer examination time and higher cost. More importantly, the biopsy may lead to the fibrosis and adhesion of diseased parts, so as to harm the effect of the treatment of early gastric cancer under the endoscope.Confocal laser endomicroscopy (CLE) can conduct real-time observation on micro-structures such as the gastric epithelial cells and blood vessels in the living body, so as to realize the optical biopsy. There are two kinds of CLE used in clinic practice, namely, the endoscopy-based confocal laser endomicroscopy (eCLE) and probe-based confocal laser endomicroscopy (pCLE). Since the pCLE can be put flexibly into the working channel of the ordinary endoscope, it can be used with other gastrointestinal endoscopes. Recently, the current application of pCLE in the research for diagnosing of GIM, GIN or carcinoma is still less.The aims of this study were:1. To determine whether pCLE with optical biopsy and targeted mucosal biopsy improves the diagnostic yield of GIM、GIN、carcinoma in high risk populations compared to vitual chromoendoscopy with standard biopsy protocol.2. To determine whether pCLE with optical biopsy and targeted biopsy, as compared to vitual chromoendoscopy with standard biopsy, can reduce the number of biopsies needed per patient for detection of GIM、GIN、carcinoma without the loss of corresponding diagnostic yield.MethodsThis study is a prospective, randomized controlled trial for the specific people. Consecutive patients scheduled for pCLE examination were enrolled in this study at Qilu Hospital, Shandong University, from June 2014 to April 2015, according to the inclusion and exclusion criteria.Inclusion Criteria were:(i) having histologically verified gastric intestinal metaplasia, intraepithelial neoplasia, and atrophic gastritis; (ii) having a history of gastric endoscopic submucosal dissection; ((iii) written informed consent.Exclusion Criteria were:(i) Patients with gastrectomy, acute gastrointestinal bleeding, and advanced gastric cancer; (ii) patients under conditions unsuitable for performing CLE including coagulopathy, impaired renal function, severe heart and lung disease, and known allergy to fluorescein sodium; (iii) pregnancy or breastfeeding; (iv) inability to provide informed consent and other situations that could interfere with the examination protocol. Before into the group, patients were given a comprehensive clinical evaluation. They were randomly allocated using computer-generated codes at a 1:1 ratio divided into two groups(A group and B group). Group A:pCLE with optical biopsy and targeted mucosal biopsy; Group B: vitual chromoendoscopy with standard biopsy. Group A:Macroscopic lesions and five standardized locations were carefully and randomly examined using the pCLE system, and continue to recording until the confocal images can clearly show the mucous membrane and vascular structure. pCLE videos obtained from different lesions and from the standardized regions were stored separately. Finally, targeted biopsies were taken from the pCLE-scanned areas where were diagnosed as GIM, GIN or carcinoma. 1 to 3 months after endoscopic examination,the quality of separately stored pCLE videos were evaluated in randomized order by a endoscopic physician. Group B: Macroscopic lesions and five standardized locations were carefully and randomly examined using the vitual chromoendoscopy examinations. Biopsies were first taken from macroscopic lesions in the stomach and then from the five standard biopsy sites. The biopsy specimens were first diagnosed by one experienced pathologist, and then all of the biopsy specimens were unity diagnosed by an experienced gastrointestinal pathologist. Finally, compare the diagnosis efficiency and the number of biopsies of GIM、GIN、carcinoma in two groups.ResultsA total of 148 patients were finally enrolled (group A 74,group B 74) according to the inclusion and exclusion criteria. On a per-patient analysis, there were no significant differences between group A and group B on the diagnostic yields of positive lesions、GIM、GIN and carcinoma. On a per-biopsy analysis, the diagnostic yields of positive lesions for group A and group B were 66.30% and 37.17%, respectively(P<0.001); the diagnostic yields of GIM for group A and group B were 50.55% and 33.09%, respectively(P<0.001); the diagnostic yields of GIN for group A and group B were 19.05% and 8.83%, respectively(P<0.001); the diagnostic yields of carcinoma for group A and group B were 66.30% and 37.17%, respectively(P<0.001). pCLE with targeted biopsy gave a significantly higher diagnostic yield of positive lesions、GIM、GIN、carcinoma compared with vitual chromoendoscopy with standard biopsy. On a per-lesion analysis, there were no significant differences between group A and group B on the diagnostic yields of positive lesions、GIM、GIN and carcinoma. The mean number of biopsies required per patients in group A and group B was 3.7 and 6.5 respectively. The number of biopsies needed by group A is significantly less than that needed by group B(P<0.001).ConclusionspCLE with optical biopsy and targeted mucosal biopsy can improve the diagnostic yield of positive lesions、GIM、GIN and gastric carcinoma, and can reduce the number of biopsies.