Rationale and objectives: To investigate the accuracy of M staging (staging of metastatic disease) in esophageal carcinoma based on a visual interpretation and based on tumor volume measurements on positron emission tomography (PET) computed tomography (CT).
Materials and methods: Fifty-nine untreated patients with gastroesophageal junction tumors were enrolled, including 36 subcardial gastric tumors (type III according to Siewert classification) and 23 adenocarcinomas of the cardia (AEG, type II Siewert). Patients were grouped in metastasis free (M0 stage, n = 34) and metastatic stages (M1 stage, n = 25). Tumor volume and mean and maximum standardized uptake value were measured on PET-CT. The accuracy of these quantitative tumor volume parameters in distinguishing metastasis-free tumors (M0 stage) from metastatic stages (M1 stage) was compared to the accuracy of a visual analysis with fused PET-CT. Furthermore, accuracy of PET-CT was compared to PET reviewed side by side with CT in a lesion-based analysis of 84 distant metastatic sites.
Results: In the visual interpretation, PET-CT (accuracy 88%, 74/84) was more accurate than PET (accuracy 78%, 66/84; P = .008) in characterizing the 84 potential metastatic sites in the 59 patients. Among the tumor parameters, the PET-CT tumor volume was the most accurate predictor of M1 stage and overall survival. With a threshold of 39 mL, PET-CT volume was able to predict M1 stage disease with a sensitivity of 96% and a specificity of 85%. The accuracy of M-staging was increased further when combining tumor volume measurements with the results from the visual analysis (combined results: sensitivity 96%, specificity 94%).
Conclusions: PET-CT was more accurate than PET (reviewed side by side with CT) in characterizing distant metastatic sites of gastroesophageal junction carcinomas. The highest accuracy for M-staging was obtained when combining the results of the visual analysis with the results from primary tumor volume measurements. Primary tumor volume was shown to be an independent prognostic factor.