PTU-121 The value of endoscopic biopsies in routine upper gastrointestinal endoscopy

Introduction The BSG and AUGIS quality standards in upper gastrointestinal endoscopy recommend taking biopsies for histological examination to diagnose inflammatory, pre- or malignant lesions which may or may not be visible to the naked eye [Beg, Gut 2018]. This may increase the duration and cost of the procedure and potentially reduce the focus on mucosal inspection. This study aimed to investigate the diagnostic value (changes in diagnosis and contribution to management) added by histology and its cost. Methods Patients undergoing consecutive diagnostic OGDs performed by four fellows were recruited. Incomplete, surveillance or therapeutic procedures, procedures for dysphagia and those arranged in order to obtain histology (eg. positive coeliac serology) were excluded. Biopsies were performed according to BSG quality standards [Beg, Gut 2018]. The utility of biopsies (changes in diagnosis and contribution to management) were examined from endoscopic and histological findings. Results Of 509 patients examined, 314 were included. OGD was performed for dyspepsia (66%), anaemia (26%), weight loss (16%), diarrhoea (2%) and vomiting (1%). Biopsies occurred in 86% of patients with an mean of 1.5 samples per patient biopsied. Biopsy sampling provided additional pathological information not evident at endoscopy in 28% of cases. However, if benign pathologies are excluded (mostly chemical or Helicobacter gastritis or duodenitis), this falls to 6.7% (n=18: dysplasia n=3, adenocarcinoma n=1, gastric intestinal metaplasia (GIM) n=7, villous atrophy (VA) n=2, Barretts n=5). In total, management change occurred in 19% consequent on histological findings, but this falls to 12% (n=34: adenocarcinoma n=2, dysplasia n=3, GIM, n=10, Barretts n=15, VA n=3, peptic stricture n=1) where Helicobacter pylori (which can be detected by other non-invasive methods) eradication is excluded. Fifty seven focal lesions were biopsied where dysplasia or cancer was diagnosed in 7% (1/14 oesophageal, 2/35 gastric and 1/8 duodenal). Barretts, GIM and VA were found more commonly in abnormal compared to normal mucosa (See table 1). The overall cost of tissue sampling approximates to £17,100 (£82 per person). Routine biopsies of patients with only normal mucosa costed over £3,500. Conclusions Although 86% undergoing OGD have biopsies taken, contributing to aetiopathological understanding in 28%, only a small minority (6.7%) have pre-malignant or malignant pathologies and result in a change in management in only 12%. Non-targeted biopsies are of limited value and costly when sampled from normal mucosa.