Article Text


PTU-080 Correlation Of Fdg Pet Scanning With Endoscopic Findings In Patients With Crohn’s Disease
  1. E Russo1,
  2. S Khan2,
  3. AP Brown3,
  4. N Keat3,
  5. W Hallett3,
  6. R Janisch3,
  7. RN Gunn3,
  8. EA Rabiner3,
  9. PM Matthews1,
  10. TR Orchard1
  1. 1Medicine, Imperial College London, UK
  2. 2Radiology, Imperial College Healthcare NHS Trust, UK
  3. 3Imanova Centre for Imaging Sciences, London, UK


Introduction 18F- Fluoro-deoxy-glucose Positron Emission Tomography (FDG-PET) scanning is gaining popularity in the assessment of several inflammatory disorders including Crohn’s Disease (CD).

Methods 11 patients with established CD,underwent 18F- FDG-PET scanning that was compared with recent endoscopic data, provided treatment had been unchanged between the two tests. Patients were fasted for 6 h and received 185 MBq of iv FDG followed by 800 ml of oral 2.5% mannitol. A low dose CT scan of the abdomen was performed, followed by PET, at 60 min post FDG injection. PET data was acquired over a maximum of 3 bed positions (10 min/ bed position).

Analysis involved dividing the gut into 6 segments on CT (terminal ileum, ascending, transverse, descending and sigmoid colon and rectum). Segmental maximum standardised uptake value (SUVMAX) and SUV intestine-to-liver ratio (SUVITL= SUVMAX/Liver SUVMEAN) were calculated. A segment is defined as abnormal (PET +ve) when its SUVMAX > Liver SUVMEAN as per previous literature.1–3 SUVMAX and SUVITL of endoscopically abnormal versus endoscopically normal PET +ve segments were compared using the Mann-Whitney test.

Results 11 patients (52 gut segments) had PET within a median of 1 month of endoscopy. 21/52 segments were active on endoscopy. Of these 20/21 were also PET +ve. However, 17/31 of endoscopically negative segments were alsoPET+ve suggesting a sensitivity of 95% and a specificity of 45% in our cohort.

Raising the SUVMAX threshold for defining a PET +ve segment from the existing (>SUVLIVER) to > 3.5x SUVLIVER reduced sensitivity from 95% to 86%, but improved specificity from 45% to 82% compared to the gold-standard of endoscopy.

Abstract PTU-080 Table 1

SUVMAX and SUVITL in endoscopically normal (17) versus endoscopically abnormal (20) PET +ve segments

Conclusion FDG-PET appears to be up to 95% sensitive in identifying segments with endoscopically active CD.

Several ‘false positive’ segments are also observed conferring a low specificity.

A threshold of segmental SUVMAX signal > 3.5 x SUVLIVER greatly improves sensitivity with a minimar reduction in specificity.

Segments which demonstrate FDG signal but are negative on endoscopy may reflect disease undetected by endoscopy, or may be false positives. A comparison with histological activity is required to clarify this.


  1. Louis E, JNM 2007

  2. Meisner, IBD 2007

  3. Das CJ, Eur J Nucl Med Mol Imaging 2007

Disclosure of Interest None Declared.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.