Predicting relapse in patients with inflammatory bowel disease: what is the role of biomarkers?

V arious laboratory biomarkers have been studied in inflammatory bowel disease (IBD) as diagnostic aids, indicators of disease activity or severity, and to predict the risk of relapse in those patients in remission. These biomarkers have enormous potential implications for patient management. For example, therapeutic decisions could be directed more appropriately if a marker could reliably distinguish active IBD from other inflammatory or non-inflammatory causes of symptoms, or if one could distinguish Crohn’s disease from ulcerative colitis. In addition, a simple, inexpensive, sensitive, specific marker could help monitor response in the clinic and in clinical trials. Finally, the ability to reliably predict the risk of recurrence would help direct appropriate therapy to those who would most likely benefit from it and avoid the expense and potential toxicity of chronic maintenance therapy in those who have a low risk of recurrence. In the current issue of Gut, Costa and colleagues addressed the latter issue by studying the role of faecal calprotectin as a marker of risk of relapse in ulcerative colitis and Crohn’s disease (see page 364). Calprotectin represents 50–60% of neutrophilic cytosolic protein, is stable in faeces for several days after excretion, and has a relatively easy to perform assay which is available commercially and correlates well with the more difficult and more expensive indium 111-labelled leucocyte excretion Costa and colleagues studied 38 patients with Crohn’s disease and 41 with ulcerative colitis in remission for a mean of five months. A baseline faecal calprotectin level greater than 150 mg/g had a sensitivity for predicting relapse within the next year of 89% in ulcerative colitis and 87% in Crohn’s disease. The specificity in ulcerative colitis was 82% but only 43% in Crohn’s disease. After multivariate analysis, patients with Crohn’s disease with a baseline faecal calprotectin greater than 150 mg/g had a non-significant twofold increased likelihood of relapse whereas those with ulcerative colitis and an elevated faecal calprotectin had a significant 14-fold increase risk. C reactive protein (CRP) and erythrocyte sedimentation rate (ESR), either as single tests or combined with calprotectin, were not useful for predicting relapse. The results of this study add to the growing body of literature on biomarkers in IBD in general and faecal calprotectin in particular. In previous studies, faecal calprotectin has been shown to be a sensitive marker of activity in Crohn’s disease and to correlate well with endoscopic and histological disease activity in ulcerative colitis. 4 Faecal calprotectin also normalises along with endoscopic healing in Crohn’s disease and is a very sensitive and specific marker for distinguishing IBD from irritable bowel syndrome (IBS). The Costa study suggests that calprotectin is a good marker of relapse risk in ulcerative colitis, but not in Crohn’s disease, related primarily to poor specificity, and therefore poor positive predictive value. However, a previous study of faecal calprotectin as a marker of relapse risk in IBD found a faecal calprotectin level of 50 mg/g to be a sensitive and specific marker of recurrence risk in both ulcerative colitis and Crohn’s disease. In this study, patients with a faecal calprotectin level greater than 50 mg/g had a 13-fold increased risk of relapse, and the specificity for predicting relapse in Crohn’s disease was 83%, compared with 43% in the Costa study. The reason(s) for the poorer specificity of calprotectin in the current study are not clear. A different cut off value was used, but the higher cut off in the Costa study would be expected to increase specificity rather than decrease it. The definition of remission and relapse were similar in the two studies, and both used an ELISA assay (although the assays were probably at least slightly different). In the previous study, patients were in remission for 1– 4 months while in the Costa study they were in remission for 1–12 months. Perhaps the value of calprotectin for predicting relapse in Crohn’s disease decreases the longer a patient is in remission. Several other clinical indices and biological markers have been studied in IBD. These can be divided into clinical disease activity indices, endoscopic indices, serum markers, faecal markers, and miscellaneous tests. Clinical indices, including the Crohn’s disease activity index (CDAI) and other disease activity indices for Crohn’s disease and ulcerative colitis, are largely subjective and typically cumbersome and have significant interobserver variability. Furthermore, they are not valid in patients with fistulas, stomas, and non-inflammatory causes of symptoms (for example, stenosis, post surgical anatomy). There are also several endoscopic indices of severity. The Crohn’s disease endoscopic index of severity has a poor correlation with clinical activity and treatment induced remission, although complete mucosal healing may predict a favourable course. However, this index is complex and not suitable for routine clinical practice. The most well known endoscopic index of severity is the Rutgeerts score which is simple and accurately predicts the risk of recurrence after surgical resection. A multitude of serum markers have been studied in IBD. Many of these are acute phase reactants (acid a1 glycoprotein (orosomucoid)), CRP, fibrinogen, lactoferrin, serum amyloid A, and a1 antitrypsin. In an early study, Brignola and colleagues found that combining acid a1 glycoprotein, a2 globulin, and ESR into an index resulted in 88% accuracy in predicting relapse over 18 months in patients with Crohn’s disease in remission. They found no value in baseline levels of haemoglobin, white blood cell count, albumin, iron, or CRP. Other studies have also found limited value for CRP in predicting relapse. 6 12 Data on other markers such as serum and urinary neopterin levels, serum tumour necrosis factor (TNF) a levels, TNF-a receptor levels, and various interleukins (IL-1, IL-6, IL-8), interleukin receptors (IL-2R, IL-6R), or interleukin receptor antagonists (IL-1ra) are limited or contradictory. 13 As serum markers of inflammation can be elevated in a variety of conditions, it seems likely that faecal markers COMMENTARIES 321