Introduction Chromoendoscopy is increasingly being used to detect, localise and characterise mucosal abnormalities seen at gastrointestinal endoscopy. The endoscopic features of coeliac disease may be difficult to recognise and are reported to lack sensitivity and/or specificity. Thus many UK centres undertake routine duodenal biopsy or have a low threshold for duodenal biopsy in order to ensure detection of patients with coeliac disease. Other than one Italian investigator group there has been limited work evaluating the role of duodenal dye spray in patients undergoing endoscopy. We aimed to determine if dye spray improved identification of characteristic endoscopic markers of coeliac disease and whether this would enhance a biopsy avoidance strategy.
Methods Patients undergoing oesophogastroduodenoscopy (OGD) with duodenal biopsies were prospectively recruited between January and November 2011. Four experienced endoscopists undertook the endoscopic examinations, with endoscopic findings reported both before and after the use of indigo carmine dye spray in the second part of the duodenum (D2). Endoscopic markers reported suggestive of coeliac disease included reduction or absence of duodenal folds, scalloping, mosaic pattern, visible blood vessels and nodularity of the duodenal folds. Thereafter, in accordance with the current gold standard four duodenal biopsies were taken and histology compared with reported endoscopic findings.
Results 83 patients were recruited (55 female: 28 male, median age 49 years). Of these, 33 (40%) had coeliac disease (24 newly diagnosed, nine previously treated) and 50 were controls. Three of the treated coeliac patients had persistent Marsh 3a–3c changes. In patients with coeliac disease (n=33), endoscopic features of coeliac disease were identified in 16/33 (48%) of patients. The addition of dye spray in D2 accentuated these features but only highlighted endoscopic markers in two further cases (18/33, 55%), which was not statistically significant (p=0.81). However, a significant difference was identified when comparing endoscopic markers in the coeliac group with the control group (p<0.001), both before and after the use of dye spray (Abstract PWE-121 table 1). Sensitivity, specificity, positive and negative predictive values of chromoendoscopy to detect coeliac disease were 55%, 100%, 100% and 77% respectively.
Conclusion The preliminary data from this study suggests there is no additional benefit of using dye spray in patients with suspected coeliac disease. Our data suggests that our current practice of a low threshold for duodenal biopsy may still be the optimal way of diagnosing patients with coeliac disease due to the low sensitivity of endoscopic markers.
Competing interests None declared.