ABSTRACT
HYPOTHESIS: Positron emission tomography can be useful in predicting response of esophageal cancer after preoperative chemo-radiation therapy (CRT). We evaluated the use of integrated computed tomography (CT)-PET among patients with esophageal cancer being considered for resection after CRT. METHODS: Three reviewers blinded to clinical and pathologic staging retrospectively reviewed the CT-PET scans of patients with esophageal cancer after preoperative CRT who underwent esophagectomy. [F]-fluoro-2-deoxy-D-glucose uptake for residual malignancy was determined by visual analysis and semi-quantitatively when standardized uptake value (SUV) was > or =4. RESULTS: Forty-two patients underwent esophageal resection. Using visual analysis, CT-PET had a sensitivity of 47% and specificity of 58% in detecting residual malignancy. Using semi-quantitative analysis, 19 patients had a SUV > or =4 in the region of the primary esophageal tumor and were interpreted as having residual malignancy (sensitivity 43%, specificity 50%). Of these 19, six had complete pathologic response to CRT. These false-positive results, due to therapy-induced ulceration detected at endoscopy, limit the use of CT-PET alone in detecting residual malignancy. Similarly, sensitivity (25%) and specificity (73%) of endoscopy/biopsy in detecting residual malignancy were poor. However, the accuracy of CT-PET in detecting residual malignancy was improved when combined with endoscopic findings. In the absence of ulceration at endoscopy, 8 of 8 patients with SUV > or =4 after chemo-radiation had residual malignancy at surgery. CONCLUSIONS: CRT-induced ulceration results in false-positive results on CT-PET and precludes accurate detection of residual esophageal tumor. However, CT-PET in combination with endoscopy is useful in identifying patients with a high risk of residual tumor post-CRT.
