Non-enhanced CT versus contrast-enhanced CT in integrated PET/CT studies for nodal staging of rectal cancer

Abstract

The purpose of the present study was to determine the diagnostic accuracy of non-enhanced CT and contrast-enhanced CT in integrated PET/CT studies for preoperative nodal staging of rectal cancer. Retrospective analysis was performed in 53 patients with pathologically proven rectal cancer who had been referred for preoperative staging. All patients underwent integrated PET/CT consisting of non-enhanced and contrast-enhanced CT followed by whole-body fluorine-18-fluorodeoxyglucose ([(18)F]FDG) PET. Both non-enhanced and contrast-enhanced PET/CT images were evaluated separately by two observers in consensus. The reference standard was histopathologic results. For nodal staging of rectal cancer, we compared diagnostic accuracy on a per-patient basis between the two modalities. Nodal staging was correctly determined with non-enhanced studies in 37 patients (70%) and with contrast-enhanced studies in 42 patients (79%). On a per-patient basis, sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of regional lymph node staging were 85%, 68%, 83%, 72%, and 79%, respectively, with contrast-enhanced studies, and 85%, 42%, 73%, 62%, and 70%, respectively, with non-enhanced studies. The difference in the accuracy of nodal staging between the two modalities was not significant (p = 0.063). Compared with non-enhanced studies, contrast-enhanced studies determined more correctly the status of pararectal lymph nodes (p = 0.002), internal iliac lymph nodes (p = 0.004), and obturator lymph nodes (p < 0.0001). Contrast-enhanced PET/CT is superior to non-enhanced PET/CT for precise definition of regional nodal status in rectal cancer.