Does Positron Emission Tomography Add Information to Endoscopic Ultrasound Sonography and CT in the Management of Esophageal Cancer? A Prospective Single Center Study

Abstract

Purpose: EUS and CT have been standardly used for staging esophageal cancer. With the advent of PET, many centers are adding this modality to their staging armamentarium. This prospective blinded study analyzes the utility of FDG-PET as an adjunct to EUS and chest CT for the management of patients with esophageal cancer. Methods: Between Dec 2003 and Mar 2006, patients diagnosed with esophageal carcinoma underwent EUS, CT and FDG-PET as their initial evaluation. All EUS procedures were performed by one ultrasonographer (VMS) blinded to the PET results. Two thoracic surgeons (DRJ and RS), blinded to patient identifying information and demographics, were asked to decide if the patient needed surgical resection, neoadjuvant chemotherapy followed by resection, or palliation. Both surgeons were presented the staging EUS and CT results. With each case, one surgeon was unblinded in alternating fashion to the FDG-PET results. The treatment decisions of each surgeon were compared to determine if PET altered clinical management. Coefficients of agreement (k) were calculated between readers treatment decisions. Random cases were represented to readers to assess inter- and intraobserver variability. Results: 43 patients (40 male, 3 female) with mean age of 63 y/o (range 34–89) were enrolled and data was prospectively collected. 38 (88.4%) had adenocarcinoma and 5 (11.6%) had squamous cell carcinoma of the esophagus. EUS was completed in 88.4% (38) of cases while 5 (11.6%) were incomplete secondary to tight stenosis. No statistical difference was found in inter- and intraobserver variability (Kappa = 1.0). Of the 43 patients, 16 were treated with surgery, 22 with neoadjuvant chemotherapy and surgery, and 5 with palliative chemo/radiation. In 42 of 43 patients, the readers came to identical clinical management decisions independent of PET results. In the one case that their treatment decision differed, the EUS was incomplete. The agreement on treatment strategy between readers presented with PET results and those blinded to PET results was 97% (k = 0.96, 95% CI 0.88–1.00). Conclusions: This study shows that the addition of FDG-PET to EUS and CT offers little if any additional information to the initial treatment stratification of patients with esophageal cancer. However, in patients with incomplete EUS, FDG-PET may have some clinical utility.