OBJECTIVE: The objective of our study was to evaluate the diagnostic performance of helical CT for evaluating the preoperative staging of gastric cancer in the remnant stomach. MATERIALS AND METHODS: Preoperative helical CT images of 67 patients with gastric cancer in the remnant stomach were independently analyzed regarding staging of the tumor by two radiologists who were blinded to histopathologic and surgical results. The differences in their assessments were resolved by consensus including the opinion of a third radiologist. The radiologists were asked to determine the depth of tumor invasion of the gastric wall (T stage), classifying it as </= T2, T3, or T4; local lymph node involvement (N stage); and solid organ metastasis or peritoneal involvement (M stage). TNM staging on CT was correlated with the histopathologic results of the resected specimen or with the surgical findings. Interobserver agreement was assessed using weighted kappa statistics. RESULTS: The overall accuracy of T staging for reviewers 1 and 2 and for the consensus reading were 83.6%, 78.2%, and 85.4%, respectively. Five of 34 </= T2 lesions were misdiagnosed as T3. The cause of the overstaging was the misconception of the postoperative fibrotic change of the anastomotic site as perigastric tumoral infiltration. Three of 18 T4 lesions were understaged because of inadequate gastric distention (n = 1) and misinterpretation of adjacent organ involvement as partial volume averaging (n = 2). The accuracy of N staging and M staging were 81.8% and 94.0% for reviewer 1; 78.2% and 91.0% for reviewer 2; and 81.8% and 94.0% for the consensus reading, respectively. The weighted kappa values of T staging, N staging, and M staging were 0.676, 0.619, and 0.924, respectively. CONCLUSION: Contrast-enhanced helical CT can be used successfully to preoperatively evaluate the staging of remnant stomach cancer in patients who have undergone previous gastric resection.
OBJECTIVE: The objective of our study was to evaluate the diagnostic performance of helical CT for evaluating the preoperative staging of gastric cancer in the remnant stomach. MATERIALS AND METHODS: Preoperative helical CT images of 67 patients with gastric cancer in the remnant stomach were independently analyzed regarding staging of the tumor by two radiologists who were blinded to histopathologic and surgical results. The differences in their assessments were resolved by consensus including the opinion of a third radiologist. The radiologists were asked to determine the depth of tumor invasion of the gastric wall (T stage), classifying it as </= T2, T3, or T4; local lymph node involvement (N stage); and solid organ metastasis or peritoneal involvement (M stage). TNM staging on CT was correlated with the histopathologic results of the resected specimen or with the surgical findings. Interobserver agreement was assessed using weighted kappa statistics. RESULTS: The overall accuracy of T staging for reviewers 1 and 2 and for the consensus reading were 83.6%, 78.2%, and 85.4%, respectively. Five of 34 </= T2 lesions were misdiagnosed as T3. The cause of the overstaging was the misconception of the postoperative fibrotic change of the anastomotic site as perigastric tumoral infiltration. Three of 18 T4 lesions were understaged because of inadequate gastric distention (n = 1) and misinterpretation of adjacent organ involvement as partial volume averaging (n = 2). The accuracy of N staging and M staging were 81.8% and 94.0% for reviewer 1; 78.2% and 91.0% for reviewer 2; and 81.8% and 94.0% for the consensus reading, respectively. The weighted kappa values of T staging, N staging, and M staging were 0.676, 0.619, and 0.924, respectively. CONCLUSION: Contrast-enhanced helical CT can be used successfully to preoperatively evaluate the staging of remnant stomach cancer in patients who have undergone previous gastric resection.