P185. Comparison of bowel sonography and entero-MRI for diagnosis of small bowel Crohn's disease
A. Rispo1, P.P. Mainenti2, A. Testa3, G.D. De Palma4, M. Rea3, M. Diaferia3, D. Musto3, R. Vitale3, F. Sasso3, F. Castiglione3
1University Federico II Of Naples, Gastroenterologia Federico II, Naples, Italy; 2University Federico II Of Naples, Radiology, Naples, Italy; 3University Federico II Of Naples, Gastroenterology, Naples, Italy; 4University Federico II Of Naples, Surgery and Advanced Technologies, Naples, Italy
Background: Crohn's disease (CD) is frequently localised in the small bowel, with or without colonic involvement. The diagnosis of small bowel CD is mainly performed by ileo-colonoscopy (IC) while the assessment of its extension can be achieved by X‑ray studies (follow-through, enteroclysis, CT enterography/enteroclysis) or, not invasively, by using entero-MRI and bowel sonography (BS). However, comparative studies directly correlating the diagnostic accuracy of BS and MRI are scanty. Our aim was to evaluate the diagnostic accuracy of BS and MRI for the diagnosis of small bowel CD.
Methods: We prospectively studied 120 consecutive subjects who attended our third-level IBD Unit between September 2008 and May 2011 for suspected small bowel CD. All patients underwent IC (utilised as gold standard for CD diagnosis), BS and MRI in a random order and blind way. Entero-MRI was assumed as gold standard for defining the extension of small bowel CD in not operated patients. Bivariate correlation about CD extension between MRI and BS was calculated by Spearman's coefficient (r). To test the consistency between MRI and BS for CD location and complications (strictures, abscesses, fistula) the Cohen's k measure was applied. A p value of 0.05 was considered significant.
Results: CD diagnosis was made in 73 out of 120 subjects, whereas the remaining 47 subjects received a different diagnosis. Sensitivity, specificity, positive and negative predictive values for CD diagnosis were, respectively, 98%, 95%, 96%, 97% for MRI (TP 72, TN 46, FP 1, FN 1); 96%, 95%, 96%, 95% for BS (TP 61, TN 45, FP 2, FN 2). MRI was superior to BS in defining CD extension (r = 0.71) while the concordance in terms of CD location between the two procedures was high (k = 0.81). Also, MRI showed a fair concordance with BS about strictures (k = 0.71) and abscesses (k = 0.88), with a better detection of entero-enteric fistulas (k = 0.61).
Conclusions: Entero-MRI and BS are two accurate and not invasive procedures for the diagnosis of small bowel CD, even if MRI appears to be more sensitive in defining the extension of small bowel involvement. Considering the optimal accuracy of BS for CD diagnosis and the high costs of MRI, BS could be used to select the patients for a subsequent MRI examination, in order to confirm CD diagnosis with a better assessment of its extension and fistulising complications.