P148 Simultaneous positron emission tomography with 18F-fluorodeoxyglucose with magnetic resonance imaging enterography for monitoring activity of inflammation in patients with Crohn’s disease: a pilot study
K. Beiderwellen1, B. Gomez2, P. Heusch3, G. Gerken4, M. Ruenzi5, A. Bockisch2, T. Lauenstein6, J. Langhorst*7
1University of Duisburg-Essen, Department of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital Essen, Essen, Germany, 2University Duisburg-Essen, Clinic for Nuclear Medicine, Essen, Germany, 3University of Dusseldorf, Department of Diagnostic and Interventional Radiology, Dusseldorf, Germany, 4University of Duisburg-Essen, Department of Gastroenterology and Hepatology, University Hospital Essen, Essen, Germany, 5Universität of Duisburg-Essen, Department of Gastroenterology, Kliniken Essen-Sued, Essen, Germany, 6University Duisburg-Essen, Department of Diagnostic and Interventional Radiology and Neuroradiology, Essen, Germany, 7University of Duisburg-Essen, Integrative Gastroenterology, Essen, Germany
To assess the performance of combined positron emission tomography (PET)/magnetic resonance imaging (MRI) enterography in the non-invasive assessment of inflammatory bowel lesions in patients with Crohn’s disease (CD).
The first 20 patients with confirmed CD (female: n = 11; mean age: 46 ± 11 years; n = 7 reported previous surgery; n = 7 with current stenosis) underwent PET/MRI enterography using an integrated PET/magnetic resonance (MR) scanner (Biograph mMR, Siemens Healthcare, Erlangen, Germany) with [18F] fluorodeoxyglucose (FDG) in this ongoing study. For small bowel distension, an oral contrast solution (1500 cc of mannitol and locust bean gum) was ingested. All patients received a dose of 20 mg scopolamine IV to minimise bowel motion artefacts. The following sequences were acquired: TrueFISP coronar; T2w HASTE with fat saturation coronar; T1w VIBE coronar post gadolinium; and T1w FLASH 2D coronar and axial post gadolinium. PET was acquired simultaneously in list mode with 8 min per bed. All datasets were reviewed by 2 readers in consensus with regard to the presence of active inflammation. For each segment SUVmax, as well as SULmax (SUVmax/SUVmaxLiver), was determined. For a segment-based analysis, the lower gastrointestinal tract was divided into 7 segments (rectum, sigmoid, descending, transverse and ascending colon, caecum, and terminal ileum). A colonoscopy with a segment-based analysis served as standard of reference. Receiver operating characteristic (ROC) analysis for diagnostic accuracy as area under the curve for SUVmax and SULmax was performed to determine an optimal cut-off value. Further diagnostic accuracy, sensitivity, and specificity for PET, MRI, and PET/MRI were calculated.
In total, 122 ileocolonic segments were evaluated. According to the reference standard, active inflammation was present in 15 segments. A cut-off value for SULmax of > 1 was associated with the highest accuracy (SULmax > 1, 0.77; SULmax > 1.5, 0.69; SULmax > 2, 0.75; SULmax > 3, 0.62) for the detection of inflammation. Using a cut-off value of SULmax < 1, sensitivity for PET was 87%, specificity 78%. MRI alone was associated with a higher specificity (sensitivity 73% and specificity 97%). Compared with MRI alone, the combination of PET and MRI lead to an increase in sensitivity as well as decrease in specificity (sensitivity 86%, and specificity 88%).
A combined PET/MRI enterography is feasible in the treatment of CD. Moreover, 18F-FDG PET, as well as MR enterography, provides complementary approaches for the non-invasive assessment of inflammatory bowel lesions. The combination leads to an increase in sensitivity at a decrease of specificity in this ongoing study. Further studies investigating the integration of the different MR sequences as part of the PET/MR enterography protocol, as well as the performance in the different segments, should be performed.