Single and double faecal immunochemical test strategies are effective in risk stratification for patients with symptoms of per rectal bleeding suggestive of colorectal cancer.

Abstract

Background: Faecal immunochemical test (FIT) results triage urgent suspicion of colorectal cancer (USoC) referrals to investigation. As FIT detects microscopic blood, its role in patients with per rectal bleeding (PRB) is controversial. Patients are encouraged to submit sample stools without evident bleeding. The positivity rate, colorectal cancer (CRC) detection accuracy, and benefits from repeated FITs in patients with rectal bleeding are unknown.

Methods: A prospective dataset of USoC referrals for CRC was interrogated for referral symptoms, FIT results, and colorectal investigation outcomes. These were linked to South-East Scotland Cancer Network data to ensure complete CRC outcome data. A FIT result of 10 µg Hb/g or more was considered positive. The primary outcome of interest was diagnostic performance of FIT in patients with PRB compared with symptoms excluding PRB, including sensitivity, specificity, and negative predictive value (NPV). Secondarily, the impact of double FITs in these cohorts was investigated.

Results: A total of 5686 patients completed a FIT and subsequent colorectal investigation, and 2130 (37.5%) of these had PRB as a referral symptom. FIT positivity was higher in patients with PRB compared with no PRB (34.7% versus 18.6%; P < 0.001). When two successive FITs were completed, the positivity rate rose to 43.5%. Significant bowel pathology (CRC, advanced adenoma, inflammatory bowel disease (IBD)) was more prevalent in patients with PRB. The majority of CRCs in the PRB group were located distally (PRB 94.1% versus no PRB 51.5%; P < 0.001). The sensitivity for CRC was significantly greater in those with PRB compared with no PRB (98.0% (95% confidence interval (c.i.) 95.1-99.2) versus 82.5% (95% c.i. 74.6-88.9)), with respective NPVs of 99.8% and 99.4%. Double FITs increased CRC sensitivity in the non-PRB group, removing the difference in sensitivity between the two groups observed with one test (PRB 100% (95% c.i. 92.3-100) versus no PRB 92.9% (95% c.i. 79.4-97.8)). The NPV for CRC in PRB when two FITs were complete was 100% (99.0-100).

Conclusion: Rectal bleeding makes up one-third of USoC referrals to secondary care. The FIT positivity rate is 34.7% and it has a high sensitivity for CRC. Patients with PRB with two negative FITs have a negligible CRC prevalence.