Rectal dialysate and fecal concentrations of neutrophil gelatinase-associated lipocalin, interleukin-8, and tumor necrosis factor-alpha in ulcerativecolitis

OBJECTIVE: Neutrophil gelatinase-associated lipocalin (NGAL) is a newly des cribed neutrophil lipocalin that may bind the proinflammatory bacterial tri peptide N-formylmethionyl-leucyl-phenylalanine. In situ hybridization and i mmunohistochemical studies have shown a strong NGAL expression in colonocyt es and neutrophils in ulcerative colitis (UC). Because NGAL is highly prote ase resistant, it should be ideal for in vivo fecal and dialysate studies. Our aim was to investigate the potential of NGAL as a disease activity mark er in UC and to compare it with IL-8 and TNF-alpha. METHODS: Twenty-three patients with UC, 14 with Crohn's disease (CD), 19 pa tients with acute infectious enterocolitis, and 20 healthy controls were in cluded. The disease activity of UC and CD was scored semiquantitatively. Co ncentrations of NGAL, IL-8, and TNF-alpha were determined in rectal dialysi s fluid, feces, and serum using sandwich enzyme-linked immunosorbent assays . The total protein concentration in feces and dialysate fluid was measured , and the amount of markers was expressed as ng/mg protein. RESULTS: In healthy controls and non-IBD (irritable bowel disease) colitis, the median values for NGAL in feces were 183 ng/mg protein and 546 ng/mg p rotein (p < 0.01), respectively. When separating UC into clinical activity groups (remission, mild/moderate, and severe disease activity) the correspo nding values of NGAL were 442 ng/mg (p > 0.05), 605 ng/mg (p < 0.02), and 3 646 ng/mg (p < 0.001, compared with controls), respectively, and in quiesce nt colonic CD 368 ng/mg (p > 0.05) and in active stages 751 ng/mg (p < 0.01 ). NGAL levels in dialysis fluid listed in the same order were: 11 ng/mg fo r controls, 71 ng/mg p > 0.05) for non-IBD colitis, 100 ng/mg (p < 0.02), 1 79 ng/mg (p < 0.01), and 2053 ng/mg (p < 0.001) for UC, and 14 ng/mg (p > 0 .05) and 121 ng/mg (p < 0.02) for CD, respectively. Serum NGAL concentratio ns did not differ between UC and CD in quiescent versus active stages. A si gnificant increase of NGAL in both feces and dialysate with increasing dise ase activity of UC was found (p = 0.02 and p = 0.003, respectively). CONCLUSIONS: The NGAL content in rectal dialysate and particularly in feces seems to be a reliable marker for severe disease activity in UC, whereas s erum NGAL concentrations do not reflect disease activity. (C) 1999 by Am. C oll. of Gastroenterology.