Diagnostic accuracy of Faecal Immunochemical Testing for patients with symptoms of colorectal cancer: a retrospective cohort study of 14,487 consecutive test requests from English primary care (preprint)

ABSTRACT

Objective To ascertain the diagnostic performance of faecal immunochemical test (FIT) in symptomatic primary care patients, to provide objective data on which to base referral guidelines. Design Stool samples from routine primary care practice in Oxfordshire, UK were analysed using the HM-JACKarc FIT method between March 2017 to March 2020. Clinical details described on the test request were recorded. Patients were followed up for up-to 36 months in linked hospital records for evidence of benign and serious (colorectal cancer, high-risk adenomas and bowel inflammation) colorectal disease. The diagnostic accuracy of FIT is reported by gender, age, and FIT threshold. Results In 9,896 adult patients with at least 6 months of follow-up, a FIT result ≥10 μg/g had an overall sensitivity for colorectal cancer of 90.5% (95% CI 84.9%-96.1%), women 90.0 (80.7-99.3), men 90.8 (83.7-97.8); overall specificity 91.3 (90.8-91.9), women 92.4 (91.8-93.1), men 89.8 (88.8-90.7); overall Positive Predictive Value (PPV) 10.1 (8.15-12.0), women 7.64 (5.24-10.0), men 12.5 (9.52-15.5)); and an overall Negative Predictive Value (NPV) 99.9 (99.8-100.0), women 99.8 (99.7-100.), men 99.9 (99.9-100.0). The PPV and specificity of FIT were higher for serious colorectal disease combined and the sensitivity and NPV were lower than for colorectal cancer alone. The Area Under the Curve (AUC) for all patients did not change substantially by increasing the minimum age of testing. In this population, 10% would be further investigated to detect 91% of the cancers at 10ug/g and 3% further investigated to detect 54% of the cancers at 150ug/g. The number needed to scope to detect one cancer was ten using FIT at 10ug/g. Conclusion A FIT threshold of 10 µg/g is appropriate to triage adult patients presenting to primary care with symptoms of serious colorectal disease. FIT may provide an appropriate approach to reprioritising patients colorectal cancer symptoms whose tests have been delayed by the COVID-19 pandemic. What is already known on this subject? Faecal Immunochemical Testing (FIT) is recommended by NICE to triage symptomatic primary care patients into further investigation for serious colorectal disease, including colorectal cancer. Almost no real-world data exists documenting the diagnostic accuracy of low FIT thresholds associated with colorectal cancer or serious colorectal disease in primary care with symptoms of colorectal cancer. What are the new findings? In 9,896 consecutive FITs submitted by English General Practitioners to a large English laboratory, using a threshold of 10ug/g, FIT had a sensitivity and specificity of 91% for colorectal cancer, a sensitivity of 53% and a specificity of 92% for serious colorectal disease. Of the population tested with FIT, 10% would be further investigated to detect 91% of the cancers at 10ug/g, 4% would be further investigated to detect 74% of the cancers at 50ug/g, and 3% further investigated to detect 54% of the cancers at 150ug/g. How might it impact on clinical practice in the foreseeable future? Low threshold FIT could be used as a triage test without overburdening endoscopy resources, supporting widespread implementation of the NICE recommendations for its use in low-risk patients in primary care. FIT may be an effective test for re-prioritizing patients whose endoscopy test have been deferred due to the COVID-19 pandemic to match available endoscopy resources to those at highest risk of colorectal cancer.