Search in the Abstract Database

Search Abstracts 2012

* = Presenting author

P152. Magnetic Resonance Enterography for the diagnosis and grading of postsurgical recurrence of Crohn's disease


A. Echarri1, J.C. Gallego1, V. Ollero1, A. Porta1, J. Castro1

1Arquitecto Marcide Hospital, Ferrol, Spain



Background: Two thirds of Crohn's disease CD patients will require an intestinal resection. Postoperative recurrence is common. The severity of endoscopic recurrence is associated with early clinical recurrence and the need for reoperation. Postoperative endoscopic revision from 6–12 months after surgery has been protocolized with the purpose of detecting early recurrence and of introducing the best possible treatment.

Methods: The aim was to compare Magnetic Resonance Enterography MRE and endoscopy in the detection of postsurgical recurrence. Thirty-six CD patients (females 21; mean age 41.2 years) with intestinal resection and ileocolic anastomosis underwent ileocolonoscopy and MRE. Diagnosis and grading was made according to the Rutgeerts score. MRE studies were performed at 1.0 T and evaluated by two radiologists with experience in abdominal radiology, blinded to endoscopy. A global assessment of disease activity at the anastomosis was performed using the MRE score for CD of our Hospital.

MRE activity score (0–12)
  12
Wall thickness (mm)<22–4>4
Relative enhancement (%)<7070–100>100
Motility↓ ↓ ↓
Stenosis (%)<60>60
Bowell wall edemaNoYes
Mucosal abnormalitiesNoYes
FistulaeNoYes
Inflammatory massesNoYes
AdenopathiesNoYes

The diagnostic accuracy of each single parameter and the MRE score for the detection of recurrence were calculated measuring the area under the ROC curves. Concordance between the MRE score and the Rutgeerts score was assessed calculating Cohen's kappa value.

Results: Ileocolonoscopy detected anastomotic recurrence in 29 patients. Seven patients had no recurrence (i0). MRE correctly identified 31 patients (5 inactive, 7 low grade and 19 high grade recurrences). For the detection of the presence or absence of recurrence, single parameters showed heterogeneous values for accuracy, sensitivity and specificity. The MRE score yielded accuracy 94.4%, sensitivity 100%, specificity 71.4%, positive predictive value 93.5%, and negative predictive value 100%. In reference to the ROC curve, the MRE score showed high accuracy (AUC = 0.981) and very high specificity (100%) in discriminating “high grade recurrence”. The agreement between MRE and endoscopy was excellent (κ = 0.83).

Conclusions: MRE is useful for the diagnosis of postsurgical recurrence in CD patients. Furthermore, identification of high grade recurrences is possible, in excellent agreement with endoscopy.