A Point-of-Care Faecal Test Combining Four Biomarkers Allows Avoidance of Normal Colonoscopies and Prioritizes Symptomatic Patients with a High Risk of Colorectal Cancer

Abstract

Most colonoscopies performed to evaluate gastrointestinal symptoms detect only non-relevant pathologies. We aimed to evaluate the diagnostic accuracy of a qualitative point-of-care (POC) test combining four biomarkers (haemoglobin, transferrin, calprotectin, and lactoferrin), a quantitative faecal immunochemical test (FIT) for haemoglobin, and a quantitative faecal calprotectin (FC) test in symptomatic patients prospectively recruited. Colorectal cancer (CRC), adenoma requiring surveillance, inflammatory bowel disease (IBD), microscopic colitis, and angiodysplasia were considered significant pathologies. A total of 571 patients were included. Significant pathology was diagnosed in 118 (20.7%), including 30 CRC cases (5.3%). The POC test yielded the highest negative predictive values: 94.8% for a significant pathology and 100% for CRC or IBD if the four markers turned negative (36.8% of the patients). Negative predictive values of FIT, FC, and its combination for diagnosis of a significant pathology were 88.4%, 87.6%, and 90.8%, respectively. Moreover, the positive predictive value using the POC test was 82.3% for significant pathology when all biomarkers tested positive (6% of the patients), with 70.6% of these patients diagnosed with CRC or IBD. The AUC of the POC test was 0.801 (95%CI 0.754-0.848) for the diagnosis of a significant pathology. Therefore, this POC faecal test allows the avoidance of unnecessary colonoscopies and prioritizes high risk symptomatic patients.

The performance of FIT-based and other risk prediction models for colorectal neoplasia in symptomatic patients: a systematic review

“…All included studies were of an observational study design, with 21 cross-sectional studies, 19 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 17 retrospective studies, 18 , 30 , 46 , 47 , 48 , 49 , 50 , 51 , 52 , 53 , 54 , 55 , 56 , 57 , 58 , 59 , 60 15 prospective studies, 39 , 61 , 62 , 63 , 64 , 65 , 66 , 67 , 68 , 69 , 70 , 71 , 72 , 73 , 74 and eight case–control studies. 62 , 75 , 76 , 77 , 78 , 79 , 80 One study design was unclear, as it was an abstract only.…”

“…Settings were primary care (n = 21), 30 , 31 , 35 , 47 , 50 , 51 , 52 , 53 , 54 , 56 , 59 , 62 , 64 , 66 , 68 , 71 , 73 , 77 , 78 , 80 primary and secondary care (n = 12), 18 , 19 , 30 , 33 , 34 , 49 , 60 , 63 , 65 , 67 , 72 secondary care (n = 24), 29 , 32 , 36 , 37 , 38 , 39 , 41 , 42 , 43 , 44 , 46 , 48 , 55 , 57 , 58 , 61 , 62 , 69 , 70 , 74 , 75 , 76 , 79 , 82 secondary and tertiary care (n = 3), 27 , 28 , 40 and tertiary care (n = 1). 45 One study was unclear regarding the setting.…”

“…One study examined patients from two different countries: Scotland and Spain. 18 A further 24 studies assessed patients from the UK, 30 , 31 , 38 , 41 , 43 , 44 , 45 , 46 , 47 , 51 , 56 , 57 , 59 , 60 , 61 , 62 , 66 , 70 , 71 , 77 , 80 , 81 eight from Denmark, 35 , 39 , 48 , 69 , 74 , 76 , 79 , 82 seven from Spain, 19 , 30 , 34 , 40 , 49 , 63 , 65 five from the Netherlands, 50 , 52 , 53 , 54 , 64 five from Sweden, 67 , 68 , 78 , 83 , 84 four from Australia, 27 , 28 , 37 , 72 two from China, 32 , 55 one from the USA, 73 one from Canada, 42 one from New Zealand, 58 one from Egypt, 75 one from Italy, 36 one from Malaysia 33 and one from Nigeria. 29 For further...…”

“… CRC = 349 Herrero 2018 (Spain) Retrospective cohort Primary and secondary Overall = 1572 Uses COLONPREDICT cohort, see Cubiella 2016. NR NR NR Colonoscopy CRC = 214 Hijos-Mallada 2023 65 (Spain) Prospective cohort Primary and secondary 571 Median (IQR) Significant pathology = 70 (59.5–80.5) Non-significant findings = 60 (48.5–71.5) Significant pathology Male = 67 Female = 51 Non-significant findings Male = 205 Female = 248 NR Colonoscopy CRC = 30 Adenoma = 53 Hippisley-Cox 2012 66 (UK) Prospective cohort Primary Overall = 3,880,944 Development = 2,351,052 Validation = 1,236,601 QResearch database (v.30). All practices in England and Wales that had been using their EMIS (Egton Medical Information System) computer system for at least a year were included.…”

“…Twenty-three of the studies included FIT (n = 22) or gFOBT (n = 1) combined with one or more other variables ( Table 2 ). 18 , 19 , 30 , 32 , 34 , 36 , 40 , 43 , 44 , 45 , 46 , 49 , 51 , 60 , 63 , 65 , 67 , 68 , 70 , 71 , 81 Of these, ten studies reported model development only, 30 , 34 , 40 , 43 , 44 , 45 , 60 , 65 , 70 , 81 four studies presented validations of models, 30 , 46 , 49 three studies presented both development and validation, 18 , 19 , 63 and six were classed as PPV only studies (i.e. they reported PPVs for FIT in combination with at least one other factor).…”