Abstract Background Intestinal Ultrasound (IUS) is a key diagnostic and monitoring tool in paediatric IBD; however, transmural changes in young children have not been clearly defined and differ from established adult norms. We aim to define normal transmural parameters based on colonoscopy findings in children 12 yrs. Methods Single centre retrospective review of patients 12yrs who underwent both IUS and colonoscopy, regardless of indication. The primary outcome was evaluating the bowel wall thickness (BWT) threshold correlating to any macroscopic change. Secondary outcomes included evaluating the utility of other transmural parameters in predicting macroscopic inflammation (hyperaemia, submucosal prominence, lymphadenopathy, bowel wall stratification, inflammatory fat), evaluating transmural parameters associated with moderate/severe macroscopic changes, and evaluating the reliability of existing IUS scores for detecting macroscopic change (IBUS-SAS, Milan). Descriptive statistics (median IQR) summarized the data; ROC analysis with Youden’s index evaluated relationship between continuous variables and the presence of macroscopic inflammation. Results Thirty-five patients (46% female, 9.6yrs 6.9-11.3 at IUS) met inclusion, with median IQR 1.0 days -27,17 from IUS to colonoscopy. Indication for colonoscopy included IBD (n = 30) and non-IBD (n = 5). Of the 164 matched segments (colon =134, ileum = 30), median IQR BWT corresponding to normal macroscopic appearance (n = 90) was 1.0mm 0.8-1.3 vs 2.7mm 2.1-3.5 for abnormal (n = 73; p 0.001). Moderate or severe macroscopic appearance (n = 31) had median BWT 3.3mm 2.6-3.8 (p 0.001 vs. normal). BWT correlated with macroscopic inflammation (AUC 0.94 0.90-0.98) with an optimal cutoff 2.0mm, corresponding to a sensitivity of 85% and a specificity of 93%. Using the previously recommended paediatric cut-off of 2.5mm, 46/74 segments had macroscopic inflammation, giving a false negative rate of 62%. Hyperaemia, submucosal prominence, and lymphadenopathy were all commonly seen in the absence of macroscopic inflammation (present in 13%, 14%, 11% respectively); however, inflammatory fat, loss of bowel wall stratification, and loss of colonic haustration strongly correlated with macroscopic inflammation (PPV 0.97, 1, 1). In addition to BWT, IBUS-SAS and MUC were accurate at differentiating macroscopic inflammation (AUC 0.90 for both). Conclusion In this cohort of young children, a BWT 2.0mm was the optimal cutoff for detecting macroscopic inflammation, significantly less than the established adult threshold; additionally, hyperaemia, submucosal prominence and lymphadenopathy were all commonly seen. This clearly shows the urgent need to establish normal parameters in this cohort, unique from older patients. Conflict of interest: Kullmann, Hannah: No conflict of interest Celiberto, Larissa: No conflict of interest Foster, Alice: No conflict of interest Lawrence, Sally: No conflict of interest Dr. Smyth, Matthew: Speaker/advisory board: Abbvie, Celltrion Consulting/Shareholder: Dova Health
Kullmann et al. (Thu,) studied this question.