The accuracy of noninvasive markers for the detection of endoscopically active ulcerative colitis (UC) according the Rachmilewitz Score is so far unknown. The aim was to evaluate the correlation between endoscopic disease activity and fecal calprotectin, Clinical Activity Index, C-reactive protein (CRP), and blood leukocytes.
UC patients undergoing colonoscopy were prospectively enrolled and scored independently according the endoscopic and clinical part of the Rachmilewitz Index. Patients and controls provided fecal and blood samples for measuring calprotectin, CRP, and leukocytes.
Values in UC patients (n = 134) compared to controls (n = 48): calprotectin: 396 ± 351 versus 18.1 ± 5 μg/g, CRP 16 ± 13 versus 3 ± 2 mg/L, blood leukocytes 9.9 ± 3.5 versus 5.4 ± 1.9 g/L (all P < 0.001). Endoscopic disease activity correlated closest with calprotectin (Spearman's rank correlation coefficient r = 0.834), followed by Clinical Activity Index (r = 0.672), CRP (r = 0.503), and leukocytes (r = 0.461). Calprotectin levels were significantly lower in UC patients with inactive disease (endoscopic score 0–3, calprotectin 42 ± 38 μg/g), compared to patients with mild (score 4–6, calprotectin 210 ± 121 μg/g, P < 0.001), moderate (score 7–9, calprotectin 392 ± 246 μg/g, P = 0.002), and severe disease (score 10–12, calprotectin 730 ± 291 μg/g, P < 0.001). The overall accuracy for the detection of endoscopically active disease (score ≥4) was 89% for calprotectin, 73% for Clinical Activity Index, 62% for elevated CRP, and 60% for leukocytosis.
Fecal calprotectin correlated closest with endoscopic disease activity, followed by Clinical Activity Index, CRP, and blood leukocytes. Furthermore, fecal calprotectin was the only marker that reliably discriminated inactive from mild, moderate, and highly active disease, which emphasizes its usefulness for activity monitoring. Inflamm Bowel Dis 2009