O26 A multicentre, prospective blinded comparison of magnetic- controlled capsule endoscopy (MACE) and oesophagogastroduodenoscopy (OGD) for upper gastrointestinal tract assessment

Introduction

MACE is a new, non-invasive technology which approaches the diagnostic sensitivity of OGD in gastric focal lesion detection in patients with dyspepsia. [Li et al., CGH 2016] This study compared the two in the detection of focal and diffuse lesions in patients referred for the investigation of anaemia.

Methods

MACE was performed within one week before OGD using the AnX Robotica magnet system and Navicam capsule (single camera capsule, frame rate 6/sec: Wuhan, China). Following a fast, simethicone and gastric distension using 0.5–1L water, patients swallowed the capsule semi-recumbent on the left lateral side before lying supine to allow approximation of the magnet to the chest. The examination protocol included a sequence of magnet movements and patient position changes. Inspection of an anatomical location was considered complete if sufficient to exclude any significant pathology. Endoscopists performing OGD were blinded to the MACE outcome and all findings were recorded from a pre-identified diagnostic list. Data was analysed using SPSS.

Results

45 of 81 (55.5%; median age 63 (IQR 45.5- 72), 60% male) patients had pathology identified by MACE, OGD or both. Complete views were obtained by MACE in the oesophagus (72.4%), proximal stomach (93.8%), distal stomach (93.8%) and duodenum(79.6%, p<0.001) and mean examination time was 22 minutes. Findings were concordant in 17 (37.8%: gastric erosions (n=6), ulcers (n=5), oesophagitis (n=2), eroded polyps (n=2) and gastric antral vascular ectasia (n=2)). MACE alone identified pathology in 24 (53.3%: gastric erosions (n=17), angioectasia (n=4), ulcers (n=2, oesophageal and duodenal), eroded polyp (n=1) and both angioectasia and erosions (n=1). OGD identified a distal body ulcer (n=1), small (n=1) and large (n=1) oesophageal varices and oesophageal ulcers (n=1) missed by MACE and in one patient MACE identified only an oesophageal ulcer and OGD only a duodenal ulcer, each missing the pathology detected by the other. Diagnostic yield was higher for MACE than OGD (P<0.001).

Conclusions

The improved diagnostic yield of MACE compared to OGD is accounted for mainly by minor mucosal abnormalities which could, nonetheless, contribute to occult bleeding and anaemia. Although speculative this could be because of better tolerance of MACE and a longer examination time. An increased frame capture rate, with or without a second camera, is needed to improve oesophageal visualisation.