Intestinal Ultrasound and Its Advanced Modalities in Characterizing Strictures in Crohn's Disease

Crohn's disease (CD) strictures exhibit varying levels of inflammation and chronicity. We evaluate whether intestinal ultrasound (IUS) is accurate in characterizing stricture composition.

This prospective, cross-sectional study included patients with CD who underwent IUS prior to a small bowel segment resection. Histology in the resection specimen was assessed in location-matched tissue sections. Following a consensus session, 2 pathologists blindly assessed the predominant phenotype (inflammatory [IP], chronic [CP], or mixed [MP]) in the strictures. The primary objective of the study was to find a parameter to identify strictures with IP.

A total of 36 patients were included (IP, 7 [20%]; CP, 16 [44%]; MP, 13 [36%]). Univariate analysis showed a positive association with an IP for loss of wall layer stratification (WLS) (odds ratio [OR], 7.86; P = .029) and higher contrast-enhanced ultrasound (CEUS) parameters. Bowel wall thickness (5.74 vs 7.46 mm; P = .002) was lower and color Doppler signal (OR, 0.14; P = .03) and loss of WLS (OR, 0.14; P = .027) were less present in CP. In multivariate analysis, loss of WLS and wash-in area under the curve at CEUS indicated an IP, whereas lower bowel wall thickness and color Doppler signal ≤2 indicated a CP. With these parameters, the Stricture Score Amsterdam was constructed and demonstrated accuracy for IP (area under the receiver operating curve, 0.88; P = .002) and CP (area under the receiver operating curve, 0.90; P < .0001), respectively. Inter-observer agreement for the score was good (intraclass correlation coefficient, 0.73; P < .0001).

A combination of IUS and CEUS is accurate to differentiate between inflammatory and chronic strictures in CD. The Stricture Score Amsterdam needs external validation and confirmation in its potential as a diagnostic decision aid when choosing between surgical and available medical treatments.