The assessment compared 4 intervention tests with 2 comparators.
3.1 The OC Sensor (Eiken Chemical/MAST Diagnostics) is a quantitative faecal immunochemical test. It comprises faecal sample collection tubes, latex reagent and buffer. The OC Sensor faecal sample collection tubes can hold 10 mg of faeces (Carroll et al. 2014) in 2 ml of buffer. The OC Sensor latex reagent contains latex particles coated with polyclonal antibodies for human haemoglobin. The antibodies bind with haemoglobin present in the faecal sample creating complexes that are detected using turbidimetry.
3.2 The test can be run on either the OC Sensor PLEDIA or the OC Sensor iO analyser. The OC Sensor PLEDIA can process up to 320 samples per hour, with a capacity of 200 samples per run. The OC Sensor iO can process up to 88 samples per hour with a maximum capacity of 20 samples per run. The performance of the assay varies according to the analyser used. The company states that a cut-off of 10 micrograms of haemoglobin (Hb)/g faeces (50 nanograms/ml) should be used for a symptomatic population.
3.3 The HM‑JACKarc system (Kyowa Medex/Alpha Laboratories) is a fully automated quantitative faecal immunochemical test system. It comprises faecal sample tubes, which incorporate a sample collection device (the Extel Hemo‑auto MC A device) and can hold 2 mg of faeces (Carroll et al. 2014) in 2 ml of buffer, and latex agglutination reagent (Extel Hemo‑Auto HS) and buffer (Extel Hemo‑auto). The reagent contains latex particles that are coated in antibodies specific to human haemoglobin. The antibodies bind to haemoglobin present in the faecal sample creating complexes that are detected using turbidimetry. The assay is compatible with the HM‑JACKarc analyser, which reports results as nanograms/ml. The user needs to convert the results to micrograms of Hb/g. However, because the test uses 2 mg of sample and 2 ml of buffer, results reported as nanograms/ml convert directly to micrograms of Hb/g faeces, that is 10 nanograms/ml equals 10 micrograms Hb/g faeces. The company suggests a cut-off of 10 micrograms Hb/g faeces for symptomatic populations. The HM‑JACKarc analyser can process up to 200 samples per hour, with a maximum capacity of 80 samples per run.
3.4 The FOB Gold system (Sentinel/Sysmex) is an automated quantitative faecal immunochemical test system. It comprises faecal sample collection tubes (the Sentifit pierce tube faecal collection device), which collect 10 mg of faeces (Carroll et al. 2014) in 1.7 ml of buffer, and latex agglutination reagent. The FOB Gold latex agglutination reagent contains polyclonal antibodies specific to human haemoglobin, which bind to haemoglobin present in the sample creating complexes that are detected using turbidimetry. The FOB Gold kit has CE‑marked applications for a range of clinical chemistry analysers, including the BioMajesty JCA‑6010/C, the SENTiFIT270 and those supplied by Siemens, Beckman Coulter and Abbott. The performance characteristics of the assay vary depending on which analyser is used. The company suggests that each laboratory should establish their own test cut-off according to the population the laboratory serves. The throughput of the test depends on the clinical chemistry analyser used to process the samples.
3.5 The RIDASCREEN haemoglobin test (R‑Biopharm Rhone) is an enzyme immunoassay (ELISA) for the quantitative determination of human haemoglobin in stool samples. The test is run on a microtitre plate using wells coated with polyclonal antibodies for human haemoglobin. The contents of each kit are enough for 96 tests. The instructions for the test suggest that it can be used with laboratory equipment other than the DSX automated ELISA system.
3.6 The test process incorporates 3 incubations and 2 wash steps. During the first incubation, any human haemoglobin present in the sample is captured by the polyclonal antibodies in the sample well. Unbound antigens are removed in the first wash step. Then peroxidase labelled monoclonal antibodies for human haemoglobin (conjugate) are added, which bind to the captured haemoglobin during the second incubation. In the final incubation, hydrogen peroxide and TMB (substrate) is added, which react with the peroxidase creating a colour change that is detected by a plate reader. The values from the plate reader are interpreted by the RIDA-SOFT Win.net software, which reports results as the concentration of haemoglobin per gram of stool (micrograms Hb/g faeces). The company states that 91 tests can be processed manually in 150 minutes, or 546 tests in 7 hours using an automated system. The company recommends a cut-off value of more than 2 micrograms Hb/g faeces to determine a positive sample.
3.7 The company also produces the RIDASCREEN haemoglobin/haptoglobin enzyme immunoassay that can be run in combination with the haemoglobin assay, using the same sample and processing on the same microtitre plate but with the addition of a well coated with polyclonal antibodies for human haptoglobin. Haptoglobin is a protein produced by the liver that binds to haemoglobin, making it less likely to break down as it moves through the gastrointestinal tract. The detection of haptoglobin is claimed to increase the likelihood of detecting lesions in the ascending and transverse colon. The company recommends a cut-off value of 2 micrograms Hb/g faeces to determine a positive test using the haemoglobin/haptoglobin assay.
3.8 The first comparator used in this assessment is guaiac-based faecal occult blood testing, as previously recommended in NICE's guideline on suspected cancer (see section 2.6). Guaiac-based tests detect the pseudoperoxidase activity of the haem component of haemoglobin in stool samples using guaiac-test paper and hydrogen peroxide developer. Unlike faecal immunochemical tests, they are not specific to human haemoglobin.
3.9 The second comparator is clinical assessment and referral for colonoscopy based on lower gastrointestinal symptoms alone.