C-reactive protein (CRP) is used as a biomarker of ulcerative colitis (UC) activity, but CRP levels are sometimes insufficient to reflect UC activity. Therefore, a simple noninvasive biomarker assay with sufficient sensitivity and specificity to accurately reflect UC activity is desired. Since prostaglandin E2 production and colonic inflammation are associated, we evaluated whether prostaglandin E–major urinary metabolite (PGE-MUM) can be used as such a biomarker.
Patients with UC (n = 99) were enrolled from March 2011 to February 2012. UC activity was evaluated using the simple clinical colitis activity index in 99 patients, Mayo endoscopic scoring (Mayo) in 79 patients, and Matts' grading (Matts) in 64 patients. PGE-MUM levels were measured by radioimmunoassay kit and compared against CRP levels as a control.
Both PGE-MUM and CRP levels correlated with UC activity (P < 0.01). Areas under the receiver operating characteristic curves of simple clinical colitis activity index, Mayo, and Matts for PEG-MUM were each higher than for CRP (0.93 > 0.73, 0.90 > 0.77, and 0.89 > 0.75, respectively). In multivariate logistic regression models, PGE-MUM was a significant independent predictor of histologic remission (sensitivity/specificity, 0.82/0.82) when the cutoff value was set to 17.0 μg/g creatinine, but CRP was not (0.69/0.69) (P < 0.01).
Compared with CRP level, PGE-MUM level demonstrated better sensitivity for reflecting UC activity, especially in cases of histologic inflammation, and thus seems to be a better evaluator of mucosal healing. Because this method is simple, quick, and noninvasive, PGE-MUM seems to be a useful biomarker of UC.