PillCam COLON 2 for investigation of the colon through direct visualisation: systematic review and economic evaluation

Background Colorectal cancer is the fourth most common cancer and the second most common cause of cancer deaths in England. Most cases of colorectal cancer arise from a prior adenomatous polyp in the bowel lining. Colonoscopy is the gold standard investigation for people with symptoms suggestive of colorectal cancer. During a colonoscopy, polyps can be removed or a biopsy can be taken. However, waiting times for colonoscopy can be long and the procedure can be unpleasant. Colon capsule endoscopy may provide an alternative diagnostic procedure to rule out polyps or colorectal cancer. Objectives To evaluate the clinical effectiveness, acceptability and cost-effectiveness of colon capsule endoscopy using PillCam COLON 2 for detecting colorectal polyps and colorectal cancer. Methods and data sources A systematic review searched six bibliographic databases and eight conference proceedings in August 2024. Studies of PillCam COLON 2 in symptomatic or polyp surveillance patients were included if they were randomised controlled trials, or if they reported data on diagnostic test accuracy, yield or patient preference. Bayesian pooling of sensitivity and specificity was performed. The economic analysis included a review of existing models and development of an independent model to assess the cost-effectiveness of colon capsule endoscopy versus colonoscopy and computed tomography colonography in three main populations (symptomatic patients with a faecal immunochemical test score of 10–100 μg/g, symptomatic faecal immunochemical test < 10 μg/g and surveillance patients). Subgroup analyses were conducted in patients who are able and willing to undergo colonoscopy and those who are not (denoted ‘COL-eligible’ and ‘COL-ineligible’). Results Among the diagnostic test accuracy studies (11–64% patients ‘in-scope’), for polyps of any size (two studies), ≥ 6 mm (four studies) and ≥ 10 mm (four studies), pooled sensitivities were 0.78 (95% credible interval 0.51 to 0.90), 0.83 (95% credible interval 0.70 to 0.91) and 0.85 (95% credible interval 0.70 to 0.94), respectively. Specificities were 0.60 (95% credible interval 0.27 to 0.88), 0.69 (95% credible interval 0.52 to 0.81) and 0.90 (95% credible interval 0.82 to 0.95), respectively. Among yield studies, colonoscopy was spared in 37–50% of symptomatic patients (three studies). Data on colonoscopy spared in surveillance patients were available from one study; however, these data are confidential and cannot be reported here. In patients unwilling/unable to undergo colonoscopy, colon capsule endoscopy completed 70–98% of incomplete colonoscopies. Patient preference studies indicated general satisfaction with PillCam COLON 2, but some conflicting information on patient preference for colon capsule endoscopy compared to colonoscopy and computed tomography colonography. For colonoscopy-eligible patients, within all three main analysis populations, the External Assessment Group’s model suggests that colon capsule endoscopy is expected to lead to small quality-adjusted life-year losses and higher costs than colonoscopy; hence, colon capsule endoscopy is dominated by colonoscopy. For colonoscopy-ineligible patients, colon capsule endoscopy either dominated by computed tomography colonography or has an incremental cost-effectiveness ratio which is markedly higher than £30,000 per quality-adjusted life-year gained. Despite these findings, colon capsule endoscopy is predicted to lead to substantial reductions in the number of colonoscopies required, particularly for symptomatic patients who are able to undergo colonoscopy. Limitations The generalisability of the diagnostic test accuracy data to symptomatic and polyp surveillance patients was unclear. Conclusions Colon capsule endoscopy is expected to be less effective and more expensive than colonoscopy. However, it could help to free up constrained colonoscopy services, particularly in people with symptoms suggestive of bowel cancer. Study registration The systematic review protocol is available on the PROSPERO website (registration number CRD42024586405). Funding This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: NIHR136010) and is published in full in Health Technology Assessment ; Vol. 30, No. 58. See the NIHR Funding and Awards website for further award information.