DTB | What is the faecal test?

DTB CME/CPD* BNF 4.6 What is the test?

Calprotectin is a protein released by white blood cells involved in of the bowel.1 It can be detected in faeces using laboratory or point of care tests. Although an elevated calprotectin level indicates inflammation in the bowel, it cannot identify the cause.1 Faecal calprotectin testing is mainly used in distinguishing between ‘functional’ disorders such as and ‘organic’ disorders, such as inflammatory bowel disease, a group of conditions including Crohn’s disease and ulcerative that require referral to specialist services. Here we explain the role of faecal calprotectin testing in adults in primary care.

About faecal calprotectin and testing Cut-off levels In patients with intestinal inflammation, recruited to the site Studies that compared ELISA tests with endoscopy as the reference of inflammation release calprotectin into the stool; this binds to standard for distinguishing irritable bowel syndrome (IBS) and to form a stable compound which is not broken down in the intestines.1 inflammatory bowel disease (IBD) in adults found that the sensitivity of the test ranged from 83% to 100% at a cut-off of 50µg/g (the cut-off level recommended by manufacturers of the ELISA and most POCTs), but Sample collection specificity varied (51–100%) at lower cut-off levels (see Box 1).1 Some manufacturers suggest re-testing samples if results are between 30 and Faecal calprotectin measurements require a small sample of faeces which 70µg/g, as there is a ‘grey zone’ between positive and negative tests due can be obtained relatively easily using universal containers with a spoon to imprecision around the cut-off value of 50µg/g.1 attached to the interior of the screw cap, or other commercially available devices. Sampling of the first bowel motion of the day, when the patient is most symptomatic, is recommended and could increase the diagnostic Box 1: Definitionsii yield for distal colitis (proctitis).2 Information for patients and guidance on i stool sampling can be found on the nhs.uk website. Sensitivity: ability to correctly identify people with the disease Specificity: ability to correctly identify people who do not have Faecal calprotectin tests the disease Tests to detect or measure the amount of calprotectin in faeces include Positive predictive value: probability that the person has the disease laboratory tests (e.g. -linked immunosorbent assay [ELISA]) or if the test result is positive lateral flow chromatographic immunoassay point of care tests (POCTs). In Negative predictive value: probability that the person does not have the ELISA test, an to calprotectin is attached to a microtitre plate. the disease if the test is negative When the sample is added, any calprotectin present attaches onto the antibody on the plate. The plate is exposed to an enzyme, and the Diagnostic test accuracy: overall proportion of correct results amount of enzyme product (detected by colorimetry or fluorimetry) reflects the amount of calprotectin present in the sample.1,3,4 In a lateral flow chromatographic immunoassay POCT, the sample is diluted and Detecting inflammation with calprotectin added to the sample port of a test cassette or device. The sample flows across a pad containing conjugate particles that bind to the target In a meta-analysis of 30 studies involving a total of 5,983 patients, molecule (calprotectin) and the target/conjugate complex then binds to comparing faecal calprotectin levels against histological diagnosis, immobilised and produces a visible line. For some POCT patients with IBD had a significantly higher calprotectin level than systems, an optical reader is needed to provide a quantitative readout of patients with normal findings (weighted mean difference 219µg/g, 95% calprotectin level. With other POCT systems a semi-quantitative estimate CI 174 to 264, p<0.001), giving a sensitivity of 95% and specificity of 91% 5 of calprotectin level is read directly from the test device. for the diagnosis of IBD. ▶ ELISA methods are relatively time consuming and expensive and require laboratory facilities, while some POCTs can provide faster results, within about 30 minutes.1,3,4 However, some POCTs may require a degree of * DTB CME/CPD sample preparation (e.g. centrifugation). The results are usually presented A CME/CPD module based on this article is available for completion online via BMJ as µg calprotectin per g of faeces (or mg/kg). Learning (learning.bmj.com) by subscribers to the online version of DTB. If prompted, subscribers must sign into DTB with their username and password. All users must The current evidence base does not suggest any preference for any test also complete a one-time registration on BMJ Learning and subsequently log in (with over the others on diagnostic grounds; relative cost and availability may a BMJ Learning username and password) on every visit. The answers to the multiple be more important in the choice of test.1 choice questions will be freely available on dtb.bmj.com on publication of the next issue of DTB. i See “How should I collect and store a stool (faeces) sample?” http://www.nhs.uk/chq/ ii Please see Understanding statistical terms 4: diagnostic tests (DTB 2009; 47: 71-2) for Pages/how-should-i-collect-and-store-a-stool-faeces-sample.aspx?CategoryID=69. further information on sensitivity and specificity.

102 | DTB | Vol 52 | No 9 | September 2014 dtb.bmj.com DTB | What is the faecal calprotectin test?

However, an important issue about faecal calprotectin testing and rather than secondary care relate to the different prevalence of disease interpretation is that it can detect inflammation but is not disease- in the two patient populations and the objectives of the test.2,12 specific. Elevated levels may be associated with conditions other than The emphasis in specialist care is usually on ‘ruling in’: increasing the IBD (e.g. symptomatic diverticular disease),6 certain medications probability of IBD to carry out more expensive, time consuming and (e.g. NSAIDs, proton pump inhibitors),7,8 or an artefact of within-patient invasive procedures, establish a firm diagnosis and start appropriate variation. It has been recommended to stop NSAID use several weeks treatment, so a diagnostic test with a high positive likelihood ratio is before measuring faecal calprotectin.7 Some authors have suggested preferred.12 In primary care, where the prevalence of IBD is low, the stopping alcohol for 1 week prior to testing.9 In patients presenting for emphasis is on ‘ruling out’: lowering the probability of the target the first time in primary care who have an initial faecal calprotectin disease to provide reassurance, or to adopt a ‘watchful waiting’ level in the range 50–150µg/g, a repeat sample may be warranted to strategy.12 In these instances, tests with a low negative likelihood ratio reduce the false-positive rate due to within-patient variation.7 are preferred.12 Primary care healthcare professionals will be looking for parameters such as sensitivity for IBD and negative predictive value to provide a basis for a decision not to refer.1 Differentiating IBS and IBD A retrospective study has been published based on consecutively The majority of younger adult patients (<45 years) seen with lower collected faecal calprotectin data from 962 patients aged 18–45 years abdominal symptoms (e.g. change in bowel habit, abdominal pain, presenting to their GP with persistent gastrointestinal symptoms, bloating for over 1 month and no red-flag symptoms) in general excluding patients older than 45 years with alteration in bowel habit or practice have IBS.1 Red-flag symptoms (e.g. anaemia, rectal bleeding, those with alarm symptoms (weight loss, rectal bleeding and unexplained weight loss, abdominal masses, change in bowel habit in anaemia).2 In the study, 686 (71%) patients had a negative test patients over 60 years old, family history of bowel cancer) are (<50µg/g) and 276 (29%) had a positive test; 28% (77/276) of the indications for rapid referral.1 There are no biological markers to define patients testing positive and 3% (17/686) of those testing negative had IBS, so a consensus definition and criteria were developed, the Rome a diagnosis of organic disease (e.g. , Crohn’s disease, criteria (see Box 2).10 other colitis or infective diarrhoea, , NSAID-induced or alcohol-related problems, Meckel’s diverticulum or rectal adenocarcinoma) on further investigation.2 At 50µg/g, the sensitivity of Box 2: The Rome III Diagnostic Criteria* for the test for organic disease was 82% (95% CI 73% to 89%) and the irritable bowel syndrome11 specificity was 77% (95% CI 74% to 80%), with negative predictive value (NPV) and positive predictive value (PPV) of 98% and 28%, Recurrent abdominal pain or discomfort† at least 3 days/month in respectively. the last 3 months associated with two or more of the following: With the lower prevalence of disease in primary care, the cut-off value 1. Improvement with defecation that is used for the test may need to be higher than that used in secondary care, which may lead to missing organic pathology.2 An 2. Onset associated with a change in frequency of stool increase in cut-off to 150µg/g in the aforementioned study reduced the 3. Onset associated with a change in form (appearance) of stool NPV by 1% whilst increasing the PPV to 71%. This would have reduced *Criterion fulfilled for the last 3 months with symptom onset at least 6 months prior the number of and flexible sigmoidoscopies by 10% to diagnosis. (potential endoscopic cost savings of £36,625 and specialist referral 2 †’Discomfort’ means an uncomfortable sensation not described as pain. savings of £26,433) at the cost of four missed cases of IBD. The choice of the cut-off value depends on the acceptability of risk of missing disease and also an evaluation of cost-effectiveness.2 A pragmatic approach may be to re-test faecal calprotectin (e.g. after 3 months) in Faecal calprotectin vs. other patients with an initial value between 50–150µg/g and whose symptoms inflammatory markers persist despite treatment for suspected IBS, and if calprotectin is persistently raised to refer at that point.2 The indirect inflammatory markers erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) may be measured, but have poor sensitivity and/or specificity because they can be affected by a number Cost-effectiveness of non-intestinal diseases.1,10 For example, in a study of 602 new referrals to a gastroenterology clinic, faecal calprotectin had a Costs for ESR or CRP tests are around £5 each; faecal calprotectin testing sensitivity of 89% to detect IBD (specificity 79%) compared with a around £25 and £550–£750.1,13 sensitivity of only 50% for CRP (specificity 81%) and a sensitivity of A health technology assessment in the UK included seven studies in the 10 58% for ESR (specificity 72%). comparison of IBS vs. IBD in adults, most using ELISA tests; these gave a pooled sensitivity of 93% (range 83–100%) and specificity of 94% (range 60–100%) at the cut-off level of 50µg/g.14 Although few studies used POCT, Use in general practice it seemed as reliable as ELISA, though perhaps less specific. The evidence Without faecal calprotectin testing, GP clinical assessment with the did not provide any clinical effectiveness grounds for preferring one test aid of ESR/CRP is highly sensitive in referring the majority of people over others. The assessment concluded that faecal calprotectin testing in who have IBD, but with low specificity, incorrectly identifying around primary care could reduce the need for referral and colonoscopies; any 20% of people who do not have IBD as false positives being referred quality-adjusted life-year gains are likely to be small because of the low to colonoscopy.1,10 prevalence of IBD and the high sensitivities of all of the tests, resulting in few false negatives with IBD, but considerable savings could accrue; areas This raises the question of whether faecal calprotectin testing could be of uncertainty include the optimum management of people with used in primary care to reduce unnecessary referrals and endoscopies borderline results (50–150µg/g), most of whom do not have IBD; and (reducing costs and also unwanted effects of unpleasant invasive repeat testing may be appropriate before referral.14 procedures).1 However, nearly all the evidence for the clinical effectiveness of calprotectin testing comes from ELISA testing in Based on the health technology assessment,14 the National Institute for specialist care settings, with little data from primary care.1 The key Health and Care Excellence (NICE) concluded that faecal calprotectin tests issues concerning the use of faecal calprotectin testing in primary care would reduce costs compared with GP current practice: £83 for the POCT dtb.bmj.com Vol 52 | No 9| September 2014 | DTB | 103 DTB | What is the faecal calprotectin test?

(CalDetect) and £82 for the ELISA test per patient, mainly because of the A diagnosis of IBS rather than IBD would not be made on the basis of lower number of referrals and colonoscopies for false-positive results.13 calprotectin results alone, but GPs may find calprotectin useful to help rule out IBD.1 The rates of false positives incorrectly referred to A cost-effectiveness analysis from the USA also found that a faecal colonoscopy after POCT or ELISA testing would be much lower than GP calprotectin screening strategy in patients with suspected IBD, with a cut off clinical assessment alone (5.1% or 5.6% respectively).1 of 50µg/g, was cost-effective.15 Screening at a cut-off of 100µg/g saved $417/ adult patient but delayed diagnosis for 2.2/32 patients with IBD among 100 screened patients.15 The cut-off level of 50µg/g cost an additional $55 but it Issues to address yielded 2.4 additional accurate diagnoses of IBD per 100 screened adults.15 The use of the faecal calprotectin test relies on locally agreed care pathways, appropriate training and quality assurance processes that What do guidelines say? have yet to be fully developed.13 Faecal calprotectin testing has the potential to falsely reassure GPs when used in people suspected of Faecal calprotectin testing is recommended by NICE as an option to having bowel cancer, and so these people should be excluded from support clinicians with the differential diagnosis of IBD or IBS in adults the recommendations; when uncertainty remains in primary care with recent onset lower gastrointestinal symptoms for whom specialist around whether to refer a patient for specialist assessment based on assessment is being considered, if cancer is not suspected, having faecal calprotectin testing, the clinician will benefit from further considered the risk factors, and if appropriate quality assurance processes specialist input (clinical or laboratory) before making a decision.13 and locally agreed care pathways are in place for the testing.13 Robust evidence is needed on the comparative performance of Faecal calprotectin testing is both more effective and less costly than the different faecal calprotectin tests, including the performance of POCTs ESR and CRP strategy.1 The absence of inflammation, as indicated by a compared with laboratory-based tests.13 Further research is needed on negative calprotectin test, means that IBD is very unlikely.1 Patients with the impact of testing on clinical decision-making at different cut-off a negative calprotectin test can be managed in primary care and spared levels in primary care.13 Research is also required into repeat testing further investigations.1 strategies in people with intermediate levels of faecal calprotectin.13

Conclusion Faecal calprotectin is a sensitive and specific non-invasive marker of intestinal inflammation that has been shown to reliably exclude IBD in a selected patient population with no red-flag symptoms. NICE recommends that health care professionals in primary care should use it as part of an established diagnostic pathway and ideally as a test prior to the decision on a patient’s referral to secondary care for further investigations. It is potentially cost-effective in reducing unnecessary invasive procedures. However, its introduction will require development of appropriate care pathways, agreement over the appropriate testing system (laboratory-based or point of care) and local referral processes. Training for healthcare professionals and information for patients will be needed to ensure that tests are targeted appropriately. There is currently limited evidence for the test in older people and future studies are warranted to assess the potential of faecal calprotectin testing in older people.

1. National Institute for Health and Care Excellence, 2013. Faecal calprotectin 9. Tibble JA et al. Surrogate markers of intestinal inflammation are predictive of testing for differentiating amongst inflammatory and non-inflammatory bowel relapse in patients with inflammatory bowel disease. Gastroenterology 2000; diseases: a systematic review and economic evaluation [online]. Available: 119: 15-22. http://www.nice.org.uk/guidance/dg11/resources/faecal-calprotectin- 10. Tibble JA et al. Use of surrogate markers of inflammation and Rome criteria to diagnostic-tests-for-inflammatory-diseases-of-the-bowel-diagnostics- distinguish organic from nonorganic intestinal disease. Gastroenterology 2002; assessment-report-dap122 [Accessed 20 August 2014]. 123: 450-60. 2. Pavlidis P et al. Diagnostic accuracy and clinical application of faecal 11. Rome Foundation, 2014. Rome III disorders and criteria [online]. Available: calprotectin in adult patients presenting with gastrointestinal symptoms in http://www.romecriteria.org/criteria/ [Accessed 20 August 2014]. primary care. Scand J Gastroenterol 2013; 48: 1048-54. 12. van Rheenen PF et al. Faecal calprotectin for screening of patients with 3. Otten CMT et al. Diagnostic performance of rapid tests for detection of fecal suspected inflammatory bowel disease: diagnostic meta-analysis. BMJ 2010; calprotectin and lactoferrin and their ability to discriminate inflammatory from 341: c3369. irritable bowel syndrome. Clin Chem Lab Med 2008; 46: 1275-80. 13. National Institute for Health and Care Excellence, 2013. Faecal calprotectin 4. NHS Purchasing and Supply Agency, 2010. Value of calprotectin in screening out diagnostic tests for inflammatory diseases of the bowel. Diagnostics guidance irritable bowel syndrome [online]. Available: http://www.alphalabs.co.uk/asp/ 11 [online]. Available: http://www.nice.org.uk/guidance/DG11 [Accessed db/documents/cep09026_Calprotectin%20Evidence_Jan10.pdf [Accessed 20 20 August 2014]. August 2014]. 14. Waugh N et al. Faecal calprotectin testing for differentiating amongst 5. von Roon AC et al. Diagnostic precision of fecal calprotectin for inflammatory inflammatory and non-inflammatory bowel diseases: systematic review and bowel disease and colorectal malignancy. Am J Gastroenterol 2007; 102: 803-13. economic evaluation. Health Technol Assess 2013; 17: 1-203. 6. Tursi A et al. Faecal calprotectin in colonic diverticular disease: a case-control 15. Yang Z et al. Effectiveness and cost-effectiveness of measuring fecal study. Int J Colorectal Dis 2009; 24: 49-55. calprotectin in diagnosis of inflammatory bowel disease in adults and children. 7. Sherwood RA. Faecal markers of gastrointestinal inflammation. J Clin Pathol Clin Gastroenterol Hepatol 2014; 12: 253-62. 2012; 65: 981-5. 8. Poullis A et al. Proton pump inhibitors are associated with elevation of faecal calprotectin and may affect specificity. Eur J Gastroenterol Hepatol 2003; 15: DOI: 10.1136/dtb.2014.9.0276 573-4.

104 | DTB | Vol 52 | No 9 | September 2014 dtb.bmj.com