Staging for locoregional extension of rectal cancer by transrectal ultrasonography and computed tomography

The preoperative assessment of rectal cancer wall invasion and regional lymph node metastasis is essential for planning optimal therapy.
The purpose of this study was to evaluate usefulness of and limitation on staging of transrectal ultrasound and computed tomography in rectal cancer. This study was performed to determine the accuracy, sensitivity, specificity, positive predictive value and negative predictive value for perirectal tumor invasion and lymph node metastasis for both of these methods. A total of 120 patients with rectal cancer were examined using transrectal ultrasonography (n=67), computed tomography (n=91) and both of these methods at the same time (n=38). The results obtained in these diagnostic modalities (cT, cN) were compared to the histopathologic staging of specimens (pT, pN). A tumor extending beyond the bowel wall (T3, T4) and lymph node involvement (N1) were considered to be ”positive”. A tumor within the bowel wall (T1, T2) and without lymph node involvement (N0) were considered to be ”negative”. In staging the depth of tumor invasion (T), the overall accuracy was 65.75% in transrectal ultrasonography and 79.12% in computed tomography. Overstaging was 13.43% in transrectal ultrasonography and 8.79% in computed tomography. Understaging was 20.90% and 12.09%, respectively. The sensitivity was 66.67% in transrectal ultrasonography and 81.67% in computed tomography, while specificity was 64% and 74.19%, respectively. Positive predictive values were 75.68% and 85.96% and negative predictive values were 53.33% and 67.65%, respectively. In staging lymph node metastasis (N) the overall accuracy was 50.82% in transrectal ultrasonography and 58.54% in computed tomography. Overstaging was 13.11% in transrectal ultrasonography and 23.17% in computed tomography. Understaging was 36.11% and 18.29%, respectively. The sensitivity was 37.14% in transrectal ultrasonography and 60.53% in computed tomography, while specificity was 69.23% and 56.82%, respectively. Positive predictive values were 61.90% and 54.76% and negative predictive values were 45% and 62.50%, respectively. Staging of depth of tumor invasion (T) was better than staging of lymph node metastasis (N) in both methods. There was no superiority in preoperative staging by transrectal ultrasonography over computed tomography in the analyzed group of patients.