Whole-body FDG positron emission tomographic imaging for staging esophageal cancer comparison with computed tomography.

PURPOSE: The aim of the authors in this study was to critically evaluate the role of whole-body positron emission tomographic (PET) imaging with fluorine-18 fluorodeoxyglucose (FDG) in staging esophageal cancer, and further to compare this method with conventional imaging with computed tomography (CT). MATERIALS AND METHODS: The authors performed independent, blinded retrospective evaluations of FDG PET images obtained in 47 patients referred for the initial staging of esophageal cancer before minimally invasive surgical staging. Twenty PET studies from patients with nonesophageal thoracic cancers were randomly selected for inclusion in the PET readings. In a subset of 37 of 47 cases, the PET findings were compared with independent readings of CT studies acquired within the same 6-week interval. The utility of the imaging findings was evaluated using a high-sensitivity interpretation (i.e., assigning equivocal findings as positive) and a low-sensitivity interpretation (i.e., assigning equivocal findings as negative). RESULTS: PET was less sensitive (41% in high-sensitivity mode, 35% in low-sensitivity mode) than CT (63% to 87%) for diagnosing tumor involvement in locoregional lymph nodes, which was identified by surgical assessment in 72% of patients. Notable, however, was the greater specificity of PET-determined nodal sites (to approximately 90%) compared with CT (14% to 43%). In detecting histologically proved distant metastases (n = 10), PET performed considerably better when applied in the high-sensitivity mode, with a sensitivity rate of approximately 70% and a specificity rate of more than 90% in the total group and in the subset of patients with correlative CT data. In the low-sensitivity mode, CT identified only two of seven metastatic sites, whereas the high-sensitivity mode resulted in an unacceptably high rate of false-positive readings (positive predictive value, 29%). PET correctly identified one additional site of metastasis that was not detected by CT. CONCLUSIONS: The relatively low sensitivity of PET for identifying locoregional lesions precludes its replacement of conventional CT staging. However, the primary advantage of PET imaging is its superior specificity for tumor detection and improved diagnostic value for distant metastatic sites, features that may substantially affect patient management decisions. In conclusion, PET imaging is useful in the initial staging of esophageal cancer and provides additional and complementary information to that obtained by CT imaging.

A study of injected dose for brain mapping on the ECAT HR+: activation maps for a parametric verbal working memory task.

The success of the 15O-water PET technique to localize statistically significant changes in regional cerebral blood flow is dependent on factors such as the activity level injected and the magnitude of the flow change. Undetectable changes may occur if insufficient activity is injected leading to high levels of statistical noise or the task performed results in only small changes in blood flow. To explore the relationship between injected activity and statistical significance, we performed a series of studies with the ECAT EXACT HR+, a high resolution PET tomograph. A parametric verbal working memory task (the N-back task) was selected to examine the relationship between regional cerebral blood flow and working memory load across a range of injected doses of 15O-water. At each activity level the volunteers were required to perform four different levels of the N-back task, a task in which a letter displayed on a monitor is matched with the letter displayed N letters previously. With increasing N, this task places increased load on working memory. For this study, 5, 10, and 15 mCi of 15O-water were injected into nine normal volunteers. The complete sequence of four tasks (N = 0, 1, 2, and 3) at three activity levels was repeated twice, for a total of 24 injections of 15O-water. We show that the peak count rate performance for the HR+ is approached at injected activity levels of 15O-water around 15 mCi. For this particular choice of N-back task, robust activation maps can nevertheless be obtained with as little as 5 mCi injected dose.