Background: Fecal calprotectin (FC) is a widely used noninvasive marker of gut inflammation that is associated with endoscopic severity in Crohn's disease (CD). However, FC has been inconsistent in predicting postoperative recurrence of CD, and its utility in the postoperative setting remains unclear.
Materials and methods: Blood and fecal samples were collected in consecutively recruited patients with CD who had undergone ileocolonic resection and required a colonoscopy to assess postoperative recurrence, as defined by the Rutgeerts score (RS).
Results: A total of 86 patients were prospectively recruited at five centers. Overall, 49 (57%) had CD recurrence (RS≥i2). FC concentrations trended to increase with RS severity; FC median (interquartile range) was significantly higher in patients with endoscopic recurrence than those in endoscopic remission [172.5 (75-375) vs. 75 (36.5-180.5) μg/g, respectively]. The same occurred for C-reactive protein (CRP) [0.5 (0.1-0.95) vs. 0.1 (0.02-0.27)] mg/dl and the Harvey-Bradshaw index (HBI) [4 (2-7) vs. 1 (0-3.5)]. The three variables significantly correlated. The area under the curve to discriminate between patients in endoscopic remission and recurrence was 0.698 for FC, with 62 μg/g being the optimal cut-off point. This indicated FC would have 85.7% sensitivity and 45.9% specificity in detecting any recurrence, having positive predictive value and negative predictive value of 67.7 and 70.8%, respectively. Area under the curve for CRP and HBI were both 0.710. The combination of CRP and HBI provided a positive predictive value 95.7 and a diagnostic odds ratio of 30.8.
Conclusion: FC is not better than CRP combined with HBI to predict endoscopic postoperative recurrence of CD.