Objective: The clinical course of Crohn's disease is often unpredictable. The aim of this study was to select the most useful parameters able to predict clinical relapses.
Methods: One hundred-thirty Crohn's disease patients in clinical remission were followed every 4 months for 2 yr or until clinical relapse. Demographic and clinical data were recorded and intestinal permeability (lactulose/mannitol [L/M] test) and biochemical tests (white blood cell count, erythrocyte sedimentation rate, C-reactive protein, alpha1 acid glycoprotein, and serum iron) were performed at study entry. A subgroup of 54 patients had clinical follow-up and repeated tests every 4 months.
Results: Fifty-two patients (40%) relapsed during the 2-yr follow-up. A significant correlation was found between relapse and gender (p = 0.030) but not between relapse and age, extent and type of disease, previous surgery, or therapy. Increased L/M test (p = 0.0001) and decreased serum iron level (p = 0.0057) were associated with clinical relapse. Time-dependent analysis, performed on patients receiving serial evaluation, showed that L/M test alteration was the only variable that could predict a relapse (RR 8.84, 95% confidence interval [CI] 1.41-53.37; p < 0.05).
Conclusions: The L/M test identifies Crohn's disease patients in apparent remission, but with a high risk of clinical relapse, better than clinical and biochemical indices. Different treatment strategies might be suggested for this subgroup of patients.