Diagnostic accuracy of one or two faecal haemoglobin and calprotectin measurements in patients with suspected colorectal cancer. (2nd December 2018)

• Record Type: Journal Article
• Authors: Turvill, James; Mellen, Samantha; Jeffery, Laura; Bevan, Sarah; Keding, Ada; Turnock, Daniel
• Is Part Of: Scandinavian journal of gastroenterology. Volume 53:Number 12(2018)
• Page Start: 1526
• Page End: 1534
• Publication Date: 2018-12-02
• Subjects: Colorectal cancer -- faecal biomarkers -- faecal immunochemical test -- faecal calprotectin -- diagnostic accuracy; Gastroenterology -- Periodicals; Digestive organs -- Diseases -- Periodicals; 616.33
• DOI: 10.1080/00365521.2018.1539761
• Languages: English
• ISSNs: 0036-5521

Abstract

Abstract: Background: The role of faecal biomarkers in patients at 'high risk' of colorectal cancer (CRC) is not yet defined. Pre-analytical factors, such as heterogeneity of biomarker distribution within faeces, may influence their optimisation in clinical practice. We undertook to determine whether repeat or combined biomarker testing improves diagnostic accuracy for CRC or clinically significant disease. Methods: Patients referred with suspected CRC provided two separate faecal samples each for faecal immunochemical testing (FIT) and faecal calprotectin (FC) prior to investigation. Diagnostic accuracy of FIT and FC were evaluated based on final diagnoses. Results: Five hundred fifteen patients completed a full colorectal evaluation. The optimal cut-off for CRC using a single FIT was ≥12 µgHb/g faeces (84.6% sensitivity, 88.5% specificity). For two FIT, the cut-off was ≥43 µgHb/g faeces if either and ≥2 µgHb/g faeces if both were positive. There was no advantage in their diagnostic accuracy compared with a single FIT. FC had a lower diagnostic accuracy for CRC than FIT, which was not improved by repeat FC. No benefit was identified with FIT-FC combined. For CRC, significant adenomatous polyps and organic enteric disease combined, FIT and FC performed similarly to each other but were poorer predictors (AUC 0.677 and 0.660). There was no uplift in diagnostic accuracy when the tests were repeated or combined. Conclusion: This study supports using a single FIT at a cut-off close to that recommended by NICE DG30 to improve diagnostic accuracy for 'two-week wait' patients referred with suspected CRC.