Diagnostic Yield of Combining CT Colonoscopy and Endoscopy to Investigate Colorectal Cancer

The number of urgent referrals for suspected colorectal cancer in the NHS is constantly on the rise placing significant strain on endoscopy services. Alternatives or adjuncts to colonoscopy are the need of the hour. CT colonography (CTC) is gaining popularity demonstrating comparable sensitivity albeit concerns regarding missed lesions especially in the anorectum. In theory, this can be overcome by combining flexible sigmoidoscopy with CTC, and this is still in practice in some centres in the UK. The UK Bowel Cancer Screening Programme (UK BCSP) guidelines however suggest that additional endoscopy is not required alongside CTC to improve diagnostic yield. This study primarily aims to determine the need to perform flexible sigmoidoscopy/endoscopy in conjunction with CTC, in a view to reduce the number of unwarranted endoscopic investigations.

A single tertiary centre analysis was performed from 2019 to 2022 comprising of symptomatic/high risk patients referred to the colorectal clinic on the 2-week wait pathway, patients under polyp/cancer surveillance and patients undergoing endoscopic investigations following inpatient admission. Findings on endoscopy were compared with those on CTC with significant findings being defined as polypoidal lesions ≥ 5 mm, benign and malignant. Classification matrices were generated for each outcome, and sensitivity analysis was performed.

A total of 480 patients were included with a median age of 71 years and a male-to-female ratio of 1:1.02. The incidence of histologically proven malignancy was 30/480 (6.3%). The sensitivity of CTC for detection of malignancy in relation to histology was 93.3% (confidence interval (CI) 77.9%, 99.2%) while that for endoscopy was higher at 96.7% (CI 82.8%, 99.9%). For the detection of polyps ≥ 5 mm, the relative sensitivity and specificity of CTC with respect to endoscopy were 77.7% (CI 68.9%, 85.0%) and 98.9% (CI 97.3%, 99.7%). Overall, 2 cancers and 25 polyps were missed by CTC. In total, 70.3% of these missed lesions were in the transverse colon, descending and the sigmoid colon (left-sided lesions).

Among patients undergoing investigation for suspected colorectal cancer, CTC demonstrated high sensitivity for malignancy but was less reliable than endoscopy for detecting significant polyps (≥ 5 mm). Our findings suggest that although CTC is effective as a standalone investigation, selective endoscopic assessment may still be necessary to optimise diagnostic yield.