P125 Investigating the small bowel in paediatric Crohn’s disease: prospective comparative study between small intestine contrast ultrasonography and magnetic resonance imaging
F. Civitelli*1, F. Maccioni2, S. Oliva1, M. Aloi1, A. Dilillo1, E. Casciani3, F. Viola1, S. Cucchiara1
1Sapienza University, Paediatric Gastroenterology and Liver Unit, Rome, Italy, 2Sapienza University, Department of Radiological Sciences, Oncology, and Pathology, Rome, Italy, 3University of Rome Sapienza, Department of Emergency Radiology, Rome, Italy
Magnetic resonance imaging (MRI) is considered the gold standard for evaluation of small bowel (SB) in Crohn’s disease (CD). However, MRI is expensive, requires a strong compliance and a considerable amount of oral contrast to distend intestinal lumen. SICUS is non-invasive, low cost, and generally well tolerated by paediatric patients (pts). We aimed to compare the diagnostic accuracy of SICUS and MRI in detecting presence, site, and extension of SB disease and assessing strictures in paediatric CD.
Children with suspected CD or relapse of a known CD were prospectively enrolled. All underwent SICUS, MRI, and ileo-colonoscopy, performed by different operators blinded to other results. The SB was subdivided into jejunum, ileum, and terminal ileum (TI). The concordance (k) between the 2 techniques for presence and site of lesions was calculated.1 For TI, sensitivity (SE) and specificity (SP) were also assessed, with ileo-colonoscopy as reference standard. One-way ANOVA with Kruskal–Wallis post-test was applied to compare the extension (cm) of disease in the different segments
Included were 66 pts (median age 13; range 7–18): 23 suspected and 43 known CD. The overall k for presence of SB lesions was 0.94 (ES 0.06; 95% CI 0.8–1). The k for segments was jejunum, 0.67 (ES 0.1, 95% CI 0.4–0.8), ileum, 0.91 (ES 0.06, 95% CI 0.76–1), and TI 0.91 (ES 0.06; 95% CI 0.8–1). SE and SP (%) of SICUS and MRI for TI lesions were 98,100 and 93, 92, respectively. There was no difference in the assessment of disease extension between SICUS and MRI (p NS). The overall k for strictures was 0.62 (ES 0.1, 95% CI 0.4–0.8). SE and SP (%) of SICUS and MRI for TI strictures were 100, 100 and 92, 87, respectively. MRI provided 7 false-positive results, not detected at SICUS nor confirmed at endoscopy.
The diagnostic performance of SICUS is comparable to that of MRI in paediatric CD. SICUS might represent a first-line tool in paediatric CD, able to reduce costs and to postpone or even avoid more invasive and expensive investigations.1
 Kramer SM, Feinstein AR. Clinical biostatistics: the biostatistics of concordance. Clin Pharmacol Ther 1981;29(1)111–23, 29.