Crohn's disease: a comparative prospective study of transabdominal ultrasonography, small intestine contrast ultrasonography, and small bowel enema

Abstract

Background: Small intestine contrast ultrasonography (SICUS), when performed after distention of the small bowel lumen with an iso-osmolar polyethylene glycol electrolyte-balanced solution, shows high sensitivity (100%) and specificity (97%) in detecting small bowel abnormalities in patients who have not received a diagnosis but in whom there is a suspicion of intestinal diseases. The diagnostic yield of SICUS remains to be established in detecting small bowel lesions in patients with proven Crohn's disease (CD) in comparison with transabdominal ultrasonography (TUS), and in relationship to the experience of the operator, using small bowel enema (SBE) as the "gold standard."

Aim: The aim of this study was to evaluate the diagnostic value of SICUS, when performed by a sonologist with 1 year of experience, and TUS, when performed by a sonologist with 10 years of experience, compared to SBE in the assessment of the site, extension, and stenosis of small intestinal lesions in CD patients.

Patients and methods: A total of 28 consecutive patients (men, 16; women, 12; age range, 21 to 60 yr) with a diagnosis of CD underwent TUS and SICUS, which were performed by 2 sonologists who were unaware of the radiologic findings, on the same day. SICUS was performed after the ingestion of 375 mL of a polyethylene glycol contrast solution. A standard SBE was performed on a different day by an expert radiologist who was unaware of the sonographic findings.

Results: Sensitivities in the detection of small bowel lesions were 96% for TUS and 100% for SICUS. Compared with SBE, SICUS detected the presence of 4 lesions in the jejunum that had been missed by TUS. The mean (+/-SD) extent of the ileal disease was 22 +/- 12.5 cm when measured during SBE, 14.5 +/- 8.6 cm when measured during TUS, and 19.5 +/- 12.5 cm when measured during SICUS [P = 0.05 (SICUS versus SBE)]. The correlation of the extension of the lesions between SICUS and SBE (r = 0.88) was better than that between TUS and SBE (r = 0.64). The sensitivities of TUS and SICUS in the detection of at least 1 stricture were 76% and 94%, respectively. Sensitivity and specificity in assessing prestenotic dilatation were 50% and 100%, respectively, at TUS, and 100% and 90%, respectively, at SICUS.

Conclusion: In inexperienced hands, SICUS is a more accurate technique for assessing CD lesions, and the accuracy is better than that of TUS performed by an expert sonologist. The use of SICUS, instead of SBE, could be indicated for the follow-up of patients with CD.