Search in the Abstract Database

Search Abstracts 2016

* = Presenting author

P251 Comparison of 3 magnetic resonance enterography indices for detection of activity and severity in Crohn’s disease

I. Alfaro*1, I. Ordás1, S. Rodríguez2, J. Castro1, A. López1, T. Pérez3, A. Cofiño4, A. Ramírez-Morros1, M. Gallego1, R. Barastegui1, A. Giner1, M. Masamunt1, E. Ricart1, J. Rimola2, J. Panés1

1Hospital Clinic de Barcelona, Department of Gastroenterology, Barcelona, Spain, 2Hospital Clinic, Radiology Department, Barcelona, Spain, 3Hospital San Borja Arriaran, Department of Gastroenterology, Santiago, Chile, 4Fundación Hospital de Jove, Department of Radiology, Gijon, Spain

Background

Magnetic resonance enterography (MRE) is an accurate technique for assessment of disease activity in Crohn’s disease (CD). Various MRE indices have been derived for use in clinical research. The aims of this study were to compare the accuracy of 3 MRE indices for detecting activity and severity in CD.

Methods

MRE and ileocolonoscopies performed within 1 month in 43 patients with CD were reviewed. MRE images were interpreted with proper blinding. MaRIA, Clermond, and London scores for each colonic segment and terminal ileum were calculated. Simplified Endoscopy Score for CD (SES-CD) was considered the gold standard.

Results

In total, 224 segments were included in the analysis. According to the established cut-off points for detecting activity using MaRIA, Clermond, and London indices, the sensitivity was 0.88, 0.90, and 0.71, with specificities of 0.97, 0.78, and 0.98, respectively. Sensitivity for detecting ulcerations was 0.90 and 0.83, for MaRIA and Clermond indices, with specificities of 0.92 and 0.89, respectively (no cut-off point for detecting ulcers were defined for London index). The AUROC curve for MaRIA, Clermond, and London indices were 0.93, 0.94, and 0.89 for detecting activity, and 0.94, 0.92, and 0.90 for detecting ulcerations. There were no statistical differences (p > 0.05) between the 3 indices for detecting activity or ulcers, but performance index was superior for MaRIA. Correlations between MaRIA, Clermond, and London indices and SES-CD were 0.68, 0.68, and 0.80 respectively (p < 0.01)

Conclusion

The 3 MRE-based indices evaluated in the current study have similar accuracy for the diagnosis of activity and severity in CD, although the MaRIA showed a higher diagnostic performance.