Near-Focus NBI Classification of Villous Atrophy in Suspected Coeliac Disease: International Development and Validation

Coeliac disease (CD) is an autoimmune condition by the exposure to gluten in patients who are genetically susceptible. The global prevalence is thought to be 1%, and incidence rate of CD in the UK is approximately 13.8 per 100,000 person-years with greater than fourfold increase in incidence over the past two decades [1]. CD has variable clinical presentations including iron deficiency anaemia, diarrhoea, bloating, weight loss, osteoporosis, rash and rarely lymphoma. The ingestion of gluten triggers an immune response with the production of antibodies and an inflammatory destruction of small bowel (duodenal) mucosa (villous atrophy). The degree of villous atrophy is categorised based on duodenal biopsy analysis according to the Marsh classification [2]. Serological testing for antibody formation has varying sensitivity and specificity for CD with extent of villous damage [3]. At present the diagnosis of CD required confirmation on histopathological analysis of duodenal biopsy [4]. The British Society of Gastroenterology (BSG) requires the investigation of CD in all patients with iron deficiency anaemia as well as other symptoms above suggestive of CD [5]. Gastroscopy and duodenal biopsies are the current gold standard for exclusion or confirmation of CD [6]. Examination with standard white light endoscopy (WLE) can reveal patchy redness (erythema) of the small bowel mucosa and possibly appearance of flattening of the duodenal villi however due to limited sensitivity especially in milder cases of villous distortion and atrophy, biopsy is warranted [7]. Owing to this patchy nature of villous atrophy, orientation of biopsy is also imperative for accurate diagnosis of CD including biopsies from the duodenal bulb as well as the second part of the duodenum [8,9]. Although biopsy is the gold standard for diagnosis of CD, it is limited by inadequate quality and orientation of sampling contributing to false negative [10]. Various endoscopic imaging techniques have been studied for the diagnosis of CD by assessment of villous architecture. The water immersion technique involves filling the lumen of the duodenum with water has shown promising sensitivity and specificity for the diagnosis of CD. Further study to assess reproducibility are warranted [11,12]. Magnification endoscopy has demonstrated reasonable positive and negative predictive value with a fair degree of reproducibility within the remit of preliminary data [13]. The addition of agents to enhance villous architecture during endoscopic examination such a water soluble blue dye spray (chromendoscopy) or ascetic acid have been assessed. Chromendoscopy has shown promising sensitivity and specificity for CD, however it has not been consistently demonstrated to be superior to WLE alone and no further benefit achieved when combining chromendoscopy to magnification endoscopy [14,15]. Furthermore, chromendoscopy is limited by operator technical ability in achieving a uniform coating of dye, interpretation and prolongs time and cost [16]. The addition of ascetic acid increased sensitivity for detecting villous atrophy and demonstrating the patchy nature of this to aid target biopsy but there is insufficient evidence to support the use of this for endoscopic diagnosis of CD alone [17]. Preliminary data for digital imaging technology; virtual chromendoscopy provided by Fujinon Intelligent Chromo Endoscopy (FICE) and i-scan developed by Pentax Medical, Japan have been conducted for the diagnoses of CD. FICE with magnification has shown 100% accuracy for evaluating villous patterns and i-scan for the diagnosis of total villous atrophy using I-scan but studies demonstrating reproduction of this have not been conducted [18,19]. Narrow Band Imaging (NBI) provides an IEE technique unique to Olympus Medical Systems. NBI provides unique images based on penetrative properties of light in tissue which is directly proportional to wavelength with a use of a filter integral to the endoscope which is switched on using a button on the hand control at the time of examination. This filter produces two narrow bands at specific wavelengths; 415nm blue light and 540nm green light which is absorbed by superficial and deep mucosal vessels respectively. NBI provides a distinct contrast between vascular architecture and mucosa as NBI wavelength is reflected by mucosa but absorbed by blood vessels and has been shown to add diagnostic value in a variety of diagnostic settings most notably in the distinction between neoplastic and non-neoplastic lesions of the gastrointestinal tract [20]. It is the distortion of the villous vascular pattern as seen with varying degrees of villous atrophy that has the potential of making an optical diagnosis of coeliac disease even in its early stages, obviating the need for routine biopsy [9]. Clear endoscopic images of villous atrophy in comparison to healthy villi have been published using high resolution narrow band imaging in known CD [21]. Preliminary study of magnification in combination with NBI has shown feasibility for the detection of both total villous and partial villous atrophy, superiority to WLE with observer agreement [22,23]. A prospective observational study assessing the use of narrow band imaging in assessing duodenal villous atrophy in a variety of clinical scenarios has been performed however patients with a known diagnosis of coeliac disease were excluded [24,25]. These studies are limited by small sample size with only preliminary data for inter-observer variation. Dual focus NBI is a unique imaging system to Olympus endoscopic technology involving two-stage optical system operated by a single push of a button on the controller. The 'near' mode provides higher resolution power compared to previous NBI generations and holds enhanced brightness and contrast ability. This study will assess the ability of dual focus NBI for the endoscopic diagnosis of coeliac disease. To date there are no large prospective observational cohort studies evaluating conventional endoscopy and dual focus narrow band imaging for the endoscopic diagnosis and assessment of coeliac disease. Additionally there is limited data on inter-observer variation using NBI across degrees of endoscopic expertise; an important consideration when translating research findings to wide clinical practice. Additionally, NBI is readily available in most UK endoscopy centres making its implementation in practice relatively straightforward and of limited financial implication.