Role of image-enhanced endoscopy in the assessment of inflammatory bowel disease.

Abstract Introduction: Ulcerative colitis (UC) which is a form of inflammatory bowel disease (IBD) is characterised by a relapsing and remitting disease course. Clinical disease activity indices (DAIs) are used to assess the severity of the disease activity relying solely on the clinical symptomatology of the patients. Non-invasive biomarkers help in assessment and possibly predicting the disease relapse. Although faecal calprotectin (FCP) is one such biomarker that is extensively researched, its accuracy in assessment and prediction of relapse is only modest. Similarly endoscopy in IBD with white light examination (WLE) alone is not accurate in either the assessment of disease activity or the prediction of disease course. Narrow band imaging (NBI) allows examination of the vasculature and pit pattern of the mucosa in greater detail than WLE. Patients with colonic IBD also have a higher risk of developing dysplasia or colorectal cancer (CRC). Chromoendoscopy (CE) provides a contrast enhancement and aids in highlighting the dysplastic areas. Aims: Primary aim of the research is to assess the role of advanced endoscopy, NBI and Chromoendoscopy (CE) in assessment of disease activity and dysplasia detection respectively in UC. The secondary aim is to assess the role of DAIs in assessment of disease activity, their correlation with endoscopic & histological markers and overall outcomes during the follow up period. Methods We performed two different experiments using advanced endoscopic techniques for this research project; one is in assessment of inflammatory activity and second is in detection of dysplasia in UC. We performed retrospective analysis of our practice to identify if white light alone predicts relapse in patients with quiescent UC. Based on our findings we devised a prospective observational study to look at the effect of adding NBI to WLE in assessment of disease activity in patients with UC of varying grades of severity. As newer generation of NBI (H290 series of Olympus KeyMed®) endoscopes were being introduced into the UK market at the time of the study, we compared the effect of NBI in three generations of endoscope (Q240, H260 and H290 series). We also assessed the use of Raman spectroscopy in endoscopic and histological assessment of inflammation in UC. In another retrospective study we looked at the uptake of chromoendoscopy in surveillance colonoscopies in UC. A randomised controlled study (RCT) was also designed to compare high definition WLE (HDWLE) to high definition CE (HDCE) in detecting dysplasia in UC surveillance. As part of relapse-prediction work we also conducted a meta-analysis of published RCTs on FCP to analyse its predictive capability in IBD. Results: In the retrospective analysis, we found that the presence of either Mayo Endoscopic Subscore >1 or Geboes score ≥2.1, increases the risk of relapse up to 6 times in the subsequent twelve months period. In our comparative study of NBI in three different generations of endoscopes, we demonstrated that NBI is superior to WLE in the assessment of the presence of blood. We also noticed a significant improvement in NBI in the newer generation of endoscopes (H290 and H260) compared to the earlier endoscopes (Q240). From the meta-analysis of RCTs we found that the FCP can predict disease flare with an accuracy of up to 75% only. In the observational study we determined that addition of NBI to WLE did not provide additional value in either assessment of disease activity or predicting relapse. Among the clinical disease activity indices (DAIs), the simple clinical colitis index or Walmsley index with score of ≥3 correlated well with endoscopy and histological findings. From the Raman spectroscopy study we identified the intensities of peaks (carotenoid and the phospholipids) that were statistically significantly different between the Raman spectra of the inflamed and quiescent colonic tissue. In our second retrospective analysis CE was found to be superior to WLE in detecting all dysplastic lesions and the detection of endoscopically visible flat non-polypoid lesions. However CE was performed only in one third of the study population. In the RCT we found that HDCE has an incremental yield of about 12.7% with a NNT of about 8, suggesting that HDCE would detect one additional patient with a dysplastic lesion for every 8 patients on whom this procedure is done. Conclusion: The thesis has shown that endoscopic biomarkers and FCP do not reliably predict relapse in UC. Addition of NBI does not confer added benefit in assessment of disease activity. HDCE is superior to HDWLE and should be adapted as a standard practice in surveillance of dysplasia in UC.