|
|
||||||||
|
|
|||||||||
|
|
Oncology-Clinical Diagnosis: Solid TumorsGI Cancers - Esophagus, Gastric and Pancreas |
1 Department of Radiology; 2 Department of PET, Kinki University, Osaka-Sayama, Japan
233
Objectives: Standardized uptake values(SUV) are affected by the partial volume effect(PVE). SUV measurement have been previously reported with or without correction for PVE. The aim of this study was to evaluate diagnostic ability of SUV on PET and size on CT in the assessment of lymph node(LN) metastases in patients with esophageal cancer. Correction of SUV using count recovery coefficients(RCs) was conducted and validated.
Methods: RCs were calculated using a NEMA phantom with 6 spheres(
10 to 37 mm) filled with FDG solution on a PET/CT scanner.Fourteen patients with confirmed esophageal cancer underwent FDG PET/CT as a preoperative staging test. SUVmax of LNs was corrected for size with a RC method assuming 2 different background SUV of 0 and 1.5,defined as cSUVmax and cSUVmax'. Size on CT,SUVmax,cSUVmax,and cSUVmax' of LNs on PET/CT were correlated with histopathology,and analyzed with a ROC method.
Results: Thirty-four mediastinal and abdominal LNs were included and histopathologically determined;18 positive(
9.6±5.1 mm,SUVmax of 3.0±2.2) and 16 negative LNs(
7.7±2.4 mm,SUVmax of 1.7±0.5). For the detection of metastasis,the size on CT provided the area under the ROC curve(Az) of 0.60 and cutoff value of 8.15 mm,while SUVmax,cSUVmax,and cSUVmax' gave Az of 0.78,0.70,and 0.77,cutoff of 1.86,1.74,and 1.71,respectively.
Conclusions: SUVmax and cSUVmax showed higher diagnostic ability for LN status than the size criteria alone on CT.Six of 8 LNs,falsely negative on CT,were correctly diagnosed as positive by SUVmax. Uncorrected SUVmax and cSUVmax' presented similar accuracy,thus,PVE correction is not mandatory for the assessment of LNs on PET/CT on this condition of scanner properties.
| ||||||||||||||||||||||||||||||||||||||