FDG-PET can distinguish vital tumor tissue and scar tissue, and therefore alters the GTV especially in patients with previous local treatment. In addition, it reduces the interobserver variability significantly compared to MRI. However, respiratory gated PET is necessary for good co-registration of PET and MRI.

The use of multiplanar reconstructions (MPRs) generated from multislice spiral CT (MSCT) data sets in the preoperative assessment of vascular invasion in pancreatic cancer was evaluated. Forty patients underwent biphasic high-resolution MSCT prior to surgery for pancreatic head cancer. Image reconstruction included thin-slice axial, sagittal and coronal MPRs as well as an MPR perpendicular to the course of a major peripancreatic vessel in proximity to the tumor. CT criteria for vascular invasion were: (1) circumferential involvement >180 degrees and (2) vessel narrowing. Imaging findings of 52 vessels were correlated with surgical and histopathological reports. Regarding the CT criterion circumferential involvement, vascular invasion was demonstrated on axial MPRs with a sensitivity and specificity of 58 and 97%. For the assessment with coronal and sagittal MPRs sensitivity was only 47%. Vascular invasion was recognized best on perpendicular MPRs with a sensitivity, specificity and accuracy of 74, 97 and 88%, respectively. Vessel narrowing was a less reliable CT criterion for vascular invasion, mainly due to the lower specificity of 91% obtained with each available MPR. Thin-slice MPRs oriented perpendicularly to a possibly invaded vessel exactly depict the grade of circumferential involvement and thus have the capability to improve the assessment of vascular invasion in pancreatic cancer.

No abstract