Faecal calprotectin for the diagnosis of inflammatory bowel diseaseBMJ 2010; 341 doi: https://doi.org/10.1136/bmj.c3636 (Published 15 July 2010) Cite this as: BMJ 2010;341:c3636
For the family doctor presented with a patient who gives a short history of bloody diarrhoea the clinical diagnosis of ulcerative colitis may be straightforward, particularly if there is a family history of inflammatory bowel disease or recent cessation of smoking. However, making a diagnosis of Crohn’s disease can be challenging, especially in younger patients, when symptoms of recurrent abdominal pain and intermittent diarrhoea may be indistinguishable from those of irritable bowel syndrome (a more likely diagnosis). Referral to specialist services for diagnosis usually involves either colonoscopy or flexible sigmoidoscopy, which may be uncomfortable for patients with irritable bowel syndrome who have visceral hypersensitivity. The linked meta-analysis by van Rheenen and colleagues (doi:10.1136/bmj.c3369) highlights a new approach to the diagnosis of inflammatory bowel disease that might reduce the number of patients referred for endoscopy.1
Calprotectin is a non-covalently associated complex of two S100 (A8 and A9) proteins that is released from phagocytes and inflamed epithelia as part of the initial innate immune response. It is resistant to intestinal degradation and is distributed throughout the stool, where it can be detected readily using standard enzyme linked immunosorbent assays (ELISAs). Since the initial reports from Fagerhol’s group in Norway,2 several groups have published data on the accuracy of testing for faecal calprotectin using different assays. The findings have been variable, but with standardisation and experience results are now more consistent. Three assays are commercially available, two of which are based on immunorecognition of the same molecular epitope, either by monoclonal or polyclonal antibodies.
The meta-analysis by van Rheenen and colleagues is timely and notable for two reasons. Firstly, meta-analysis is an unusual method for assessing studies of a diagnostic test but using the QUADAS (QUality Assessment of studies of Diagnostic Accuracy included in Systematic reviews) checklist resolves initial inconsistencies in the published data by prespecifying standards of quality for including studies in the meta-analysis. In this case the authors included only studies in which biopsies had been taken from the right colon or ileum to ensure that microscopic inflammation was not missed. Secondly, the main finding, that the ELISA test for faecal calprotectin has a sensitivity of 93% and a specificity of 96% for the diagnosis of inflammatory bowel disease in adults, is remarkable considering the diverse and complex antigenic environment of faeces.
Interestingly, the authors found that the test had a lower specificity in children. They suggest that the case mix of the study populations may have been responsible for this, but the false positives could actually have been true positives resulting from undetected small bowel pathology only detectable using capsule endoscopy. Alternatively, given that the study protocol allowed for a delay of four weeks between stool sample collection and endoscopy, the stool samples may have come from children with “infectious” diarrhoea that had resolved by the time of endoscopy.
What are the implications of van Rheenen and colleagues’ study for routine clinical practice? Our own experience, and that from other centres,3 4 5 supports more widespread use of the test in secondary care, not only to reduce the need for colonoscopy in patients referred with symptoms suggestive of inflammatory bowel disease, but also to help tailor immunosuppressive treatment in patients with established disease. The NHS recently commissioned a review of faecal calprotectin from the Centre for Evidenced based Purchasing (CEP) in an attempt to find out what effects major new technologies might have on the NHS.6 The comprehensive review includes studies of the other commercial assays (the meta-analysis included only the original Phical assay or an in-house assay) and covers technical aspects such as intra-assay variation and limits of detection. Its main findings were that for the NHS, even at current costs, faecal calprotectin was cheaper and more accurate than measuring C reactive protein or erythrocyte sedimentation rate in a diagnostic algorithm to differentiate inflammatory bowel disease from irritable bowel syndrome in primary care.
However, we cannot yet recommend the use of faecal calprotectin as a diagnostic test for inflammatory bowel disease in primary care, mainly because the results of the current study apply only to patients referred to secondary care. We have no good evidence of how the test performs in primary care, where patient characteristics and populations are probably different, thereby affecting the negative and positive predictive value of the test. This is why van Rheenen and colleagues emphasise the importance of study populations and pretest probabilities when discussing their findings. Certainly, our data from a large number of patients from primary and secondary care suggest that cut-off points might need to be raised to maintain the high negative predictive value.3 In addition, data on head to head comparisons of the different faecal calprotectin assays are scant. We also need to consider how to avoid overinvestigating patients with infectious diarrhoea, in whom falling faecal calprotectin titres offer a quantitative test that may provide a diagnosis when stool cultures are negative.
If studies conducted in primary care find a high diagnostic accuracy of the faecal calprotectin test it will be an important step forward in how inflammatory bowel disease is diagnosed.
Cite this as: BMJ 2010;341:c3636
Competing interests: The author has completed the Unified Competing Interest form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declares: (1) No financial support for the submitted work from anyone other than his employer; (2) No financial relationships with commercial entities that might have an interest in the submitted work; (3) No spouse, partner, or children with relationships with commercial entities that might have an interest in the submitted work; (4) No non-financial interests that may be relevant to the submitted work.
Provenance and peer review: Commissioned; not externally peer reviewed.