Abstract
In patients with Crohn’s disease (CD), following an ileocolic resection (ICR), colonoscopy is the gold standard for the detection of endoscopic recurrence (ER). Colonoscopy, however, is invasive and not easily accepted by patients for repeated monitoring. In contrast, transabdominal intestinal ultrasound (IUS) is non-irradiating, non-invasive, and easy to repeat. The goal of this study was to assess the accuracy of IUS for ER in CD.
This was a cross-sectional study of CD patients who underwent point-of-care IUS during a postoperative follow-up clinic visit within 30 days of a planned colonoscopy. Parameters on IUS included bowel wall thickness (BWT), bowel wall hyperemia (BWH), layer stratification, inflammatory fat, lymphadenopathy, and complications. C-reactive protein (CRP), fecal calprotectin (FC), endoscopic healing index (EHI) and Harvey Bradshaw Index (HBI) were also measured. ER was defined as a Rutgeerts Score (RS) ≥ i2. Primary outcome was the association between IUS parameters and ER. Secondary outcomes were the association of IUS parameters with other markers of disease activity. Univariable analysis – Fisher’s exact, Wilcoxon Rank Sum and Spearman correlation coefficient tested associations with ER. Area under the receiver operator curve (ROC) was used to determine optimal cut-off values for BWT to accurately identify ER.
Eighteen CD patients (9 female; 29 [19-40] years old), underwent IUS examination 45 [29-99] months post-ICR. All patients underwent endoscopy within 30 days of IUS examination, four (22%) for the first time post-ICR. Seven (39%) patients were on ustekinumab, four (22%) on adalimumab, two (11%) on infliximab, one (6%) on vedolizumab, and four (22%) on no therapy. ER was found in eight (44%) patients. BWT and BWH in the neo-terminal ileum and BWH at the ileocolic anastomosis were the only IUS parameters associated with ER (Table 1). BWT was 4.0 [3.2-4.8] mm in patients with ER vs. 2.0 [1.5-2.6] mm without (p=0.04). Neo-terminal ileum BWH was abnormal in six (75%) patients with ER vs. 0% without (p=0.007). BWT of 3.2 mm was the optimal cut point for predicting ER with an: AUROC of 0.82, positive predictive value of 100%, negative predictive value of 97.3%, sensitivity of 75%, and specificity of 100% (Figure 1) vs. a CRP of 10.4 mg/L (AUROC = 0.54) or FC of 1146 μg/g (AUROC = 0.56). Significant correlations were observed between neo-terminal ileum BWT and RS (ρ=0.51, p=0.04) and between BWT and CRP (ρ=-0.56, p=0.023), but not BWT and FC (ρ=-0.04, p=0.91), BWT and EHI (ρ=-0.04, p=0.91), or BWT and HBI (ρ=0.09, p=0.75).
IUS is a feasible, accurate, non-invasive monitoring tool for detection of postoperative CD recurrence. Larger prospective studies are needed to determine how IUS can be integrated in the monitoring of CD patients after surgery.
