MRI in the diagnosis of small bowel disease: use of positive and negative oral contrast media in combination with enteroclysis

Abstract.

The aim of the study was to evaluate the additional findings of MRI following small bowel enteroclysis and to compare the efficacy of negative and positive intraluminal contrast agents. Fifty patients with inflammatory or tumorous small bowel disease were investigated by small bowel enteroclysis and consecutive MRI using breathhold protocol (T1-weighted fast low-angle shot, T2-weighted turbo spin echo). Patients were randomly assigned to either receiving a positive oral (Magnevist, Schering, Berlin, Germany) or a negative oral MR contrast media (Abdoscan, Nycomed, Oslo, Norway). The pattern of contrast distribution, the contrast effect, presence of artifacts, as well as bowel wall and extraluminal changes, were determined and compared between the contrast type using Fischer's exact test. Sensitivity, specificity, and diagnostic accuracy for MRI and enteroclysis were calculated. Twenty-seven patients had clinically proven Crohn's disease and two patients surgically proven small bowel tumours. Magnetic resonance imaging had important additional findings as abscesses and fistulae in 20 patients. Surgically compared sensitivities were 100 and 0 % for MRI and enteroclysis, for the detection of abscesses, and 83.3 and 17 % for the diagnosis of fistulae, respectively. Bowel wall thickening was more reliably detected with use of positive oral contrast media without intravenous enhancement (p < 0.001), whereas postcontrast negative oral contrast media allow for a superior detection (p < 0.001). T2-weighted sequences were necessary with use of negative oral contrast media, because loop abscesses may be masked. Magnetic resonance imaging should be performed in all patients with suspicion of extraintestinal complications, because the complications are more reliably detected by MRI. Negative oral contrast media show advantages with the use of intravenous contrast but can mask loop abscesses using only T1-weighted imaging.