A PROSPECTIVE ASSESSMENT OF COLON CAPSULE ENDOSCOPY AS A COLONIC POLYP SURVEILLANCE TECHNIQUE IN IRELAND, AND FUTURE DEVELOPMENTS

Abstract

Colorectal cancer (CRC) has a significant global impact in terms morbidity and mortality. Our knowledge of its natural course, and its growth from a precursor lesion as a polyp, gives us the opportunity to prevent CRC development. Colonoscopy is currently the reference standard procedure to identify and treat colonic polyps, however, due to its invasive nature, it comes with risks. Polyp surveillance is endorsed by national and international guidelines aiming to reduce the risk of CRC. Surveillance colonoscopies carry a significant burden on endoscopy services. Most CRCs develop from a polyp, but not all polyps lead to a cancer. Therefore, polyp surveillance is nuanced, and alternative techniques could streamline this practice and filter in only those patients that truly would benefit from a colonoscopy. Chapters 1&2 explore the current yield of surveillance and its barriers.

My thesis primarily focuses on colon capsule endoscopy (CCE), and its application as a filter test in polyp surveillance, described in chapter 3. In addition, we assess if faecal immunochemical testing (FIT) has a place in the surveillance algorithm, as discussed in chapter 4. As part of the main study which received ethical approval from the St James Hospital/Tallaght University Hospital Ethics Committee, we recruited patients from our endoscopy surveillance waiting list and offered a CCE and FIT instead. If significant lesions were seen on CCE (3 or more polyps or any measuring 6mm or more) or FIT was positive, patients were referred for colonoscopy. The significance of this prospective study is that by applying this strategy, more than half of patients were spared a full colonoscopy and over a third avoided any invasive procedure. In contrast, FIT was not useful in selecting high risk polyps. The use of FIT as well as other available biomarkers in detecting polyps and CRC is covered in chapter 4.

Further published work in chapter 3 explores optimal bowel preparation and booster regimens in CCE. Bowel cleanliness is one of the main barriers to CCE, impacting polyp detection. We prospectively studied the benefits of adding 15ml of castor oil to CCE bowel preparation. We found it was safe, acceptable by patients and significantly improved completion rates and polyp detection in an unselected cohort. We also identified a lack of published data on CCE false negative rates. Absence of this performance parameter is not only important when consenting patients for a CCE but is also a barrier to cementing CCE as a valid alternative to colonoscopy. By retrospectively identifying patients who had both tests within a reasonable time frame, we matched reported polyps seen on colonoscopy to those not seen on CCE. Each study with a missed polyp was re-read by an experienced CCE reader and each polyp reviewed by a panel of experts. We found a reader error of 14% and capsule miss rate of 2% for significant lesions (6mm or more). The false negative rates of CCE were comparable to that of colonoscopy highlighting its value in polyp detection and its use in surveillance.

Chapter 5 covers polyp characterisation in CCE and how we select adenomatous polyps for resection. We study flexible spectral imaging colour enhancement (FICE), which is a digital chromoendoscopy technique that helps exaggerate polyp surface patterns, and its application in CCE. In a pilot study we found that FICE had a good negative predictive value and enhanced adenoma detection. Building on this work, we developed an imaging bundle that utilised FICE and additional polyp characteristics seen on white light. We found that our imaging bundle performed better in identifying adenomas compared to current CCE guidelines that focus on size and number of lesions alone. Artificial intelligence (AI) is also discussed in chapter 5 as this technology will likely enhance and/or replace certain aspects of CCE. If applied correctly, AI could further establish CCE in polyp surveillance.

Finally, in chapter 6, we explore the patient CCE experience in a symptomatic cohort and in surveillance. We found that comfort levels were higher with CCE given its less invasive nature. Patients preferred CCE over colonoscopy in both, the symptomatic and surveillance studies. Interestingly, this preference was largely unchanged when patients were reminded of the possibility of needing both tests if significant lesions were seen on CCE. In addition, CCE also had a lesser impact on days missed from work and daily activities, highlighting its superiority over colonoscopy which typically takes longer, requires sedation and more time for recovery, not to mention the need for a nominated individual to collect the patient post procedure. The thesis is summarised in chapter 7, and we discuss what the future holds for colon capsule endoscopy and polyp surveillance.