P133. Fecal calprotectin and CRP as biomarkers of endoscopic activity in Crohn's disease: A meta-study
V. Bondjemah1, J.Y. Mary2, J. Jones3, W. Sandborn4, A. Schoepfer5, E. Louis6, T. Sipponen7, A. Vieira8, J.‑F. Colombel9, M. Allez10
1Beaujon, France; 2DBIM, Hôpital Saint Louis, Inserm U 717, Paris, France; 3Mayo clinic, Rochester, United States; 4University of California, San Diego, La Jolla, United States; 5Farncombe Family Institute of Digestive Health Research, McMaster University, Hamilton, Canada; 6University of Liège and CHU Liège, Department of Gastroenterology, Liège, Belgium; 7Helsinki University Central Hospital and Haartman Institute, Helsinki, Finland; 81Clinic of Gastroenterology, São Paulo, Brazil; 9Centre Hospitalier Universitaire de Lille, Hôpital Claude Huriez, Lille, France; 10Saint-Louis Hospital, Gastroenterology, Paris, France
Background: In patients with CD, endoscopy is useful (a) to confirm the presence of active lesions in symptomatic patients; (b) to confirm mucosal healing in patients in clinical remission. However, endoscopy is invasive and surrogate biomarkers may be useful in that setting. Aim: To assess the correlation between CRP and calprotectin and endoscopic activity in CD.
Methods: A meta-study including six published studies in CD was performed. All studies had endoscopic scores (CDEIS or SES-CD) and CRP and calprotectin measurements performed at endoscopic evaluation. The sensitivity, specificity, positive and negative predictive values (PPV and NPV) to anticipate mucosal healing (using 2 definitions: CDEIS ≤3 and CDEIS ≤6) were calculated for each test as well as their associations in two different situations: patients clinically active and patients in clinical remission. An optimal threshold value was proposed for each marker and situation.
Results: 551 patients were included. In patients clinically active (CDAI >220) (n = 97), the sensitivity of a CRP ≤5 mg/l or ≤ calprotectin 200 µg/g to anticipate a CDEIS ≤6 was 83%, the specificity 71%, the PPV ranged from 66 to 81% and NPV from 86 to 73% depending on a prevalence of CDEIS ≤6 between 40 and 60%. In this situation, 38 to 50 colonoscopies out of 100 could be avoided. In patients in clinical remission (CDAI ≤150) (n = 355), the sensitivity of the association of CRP ≤10 mg/l and calprotectin ≤200 µg/g to anticipate a CDEIS ≤3 was 78%, the specificity 58%, the PPV ranged from 88% to 65% and NPV from 40% to 73% depending on a prevalence of CDEIS ≤3 varying between 50 and 80%. In this situation, biomarkers could allow to avoid 30 to 40 out of 100 colonoscopies.
Conclusions: CRP and fecal calprotectin reflect endoscopic activity in CD. Appropriate use of these biomarkers could replace endoscopic evaluations in specific situations.