Does enhanced CT influence the biological GTV measurement on FDG-PET images?

To test the influence of media injection in PET/CT on the functional or gross tumour volume measurement.

Thirty-three patients (56 ± 19 years) with non-Hodgkin's lymphoma (n=22) or Hodgkin's disease (n=11) were prospectively studied at staging. PET/CTs were performed 60 min after injection of FDG. Iopamiron 300 (Iopamidol, 1.5cc/kg) was injected immediately after, followed 50s later by a second craniocaudal CT (CT+). PET images were successively reconstructed using the unenhanced CT (PET-) and the CT+ (PET+) for attenuation correction using iterative reconstruction (4 iterations, 8 subsets, 5mm post-filtering). The SUVmax, SUVmean, SUVpeak and functional tumoural volume were measured in tumoural lymphadenopathies or malignant tissues (n=56 VOIs) using 5 3D-thresholding methods on PET- and PET+ images: absolute SUV value of 2.5; 40% of SUVmax, and 3 adaptative thresholding methods (Vauclin, Black and Schaefer methods).

The SUVmean and the volume measurement were significantly different (p<0.001) for the five segmentation methods for PET- (p<0.001) and PET+ (p<0.001). The SUVmax, SUVmean and SUVpeak increased significantly in PET+ compared to PET- (2-5%). The SUVpeak was not significantly different for the five segmentation methods. The functional volume measurements were significantly different between PET- and PET+ only for the 2.5 segmentation method (+3%; p=0.001), but not for the 40%, Vauclin, Black and Schaefer methods.

The functional volume could be measured in PET/CT when CT was performed with enhanced media. Caution should be taken when using the volume delineation method. Volume delineation methods using absolute threshold may artefactually increase the functional volume when enhanced CT is used for attenuation correction. The delineation volume using the relative or adaptative method should be preferred when contrast media are used for PET/CT.