Background: EUS-guided FNA (EUS-FNA) is the most accurate method for lymph-node staging of esophageal carcinoma; however, it may not be necessary when EUS features are present that strongly suggest a benign or a malignant origin.
Aims: (1) To identify a combination of EUS criteria that have a sufficient sensitivity and specificity to preclude the need for EUS-FNA and (2) to assess the cost savings derived from a selective EUS-FNA approach.
Methods: A total of 144 patients with esophageal carcinoma were prospectively evaluated with EUS. Accuracy of standard (hypoechoic, smooth border, round, or width > 5 mm) and modified (4 standard plus EUS identified celiac lymph nodes, >5 lymph nodes, or EUS T3/4 tumor) criteria were compared (receiver operating characteristic curves). Resource utilization of two diagnostic strategies, routine (all patients with lymph nodes) and selective EUS-FNA (FNA only in those patients in whom the number of EUS malignant criteria provides a sensitivity and a specificity <100%), were compared.
Results: Modified EUS criteria for lymph-node staging were more accurate than standard criteria (area under the curve 0.88 vs. 0.78, respectively). No criterion alone was predictive of malignancy; sensitivity and specificity reached 100% when a cutoff value of >1 and >6 modified criteria were used, respectively. The EUS-FNA selective approach may avoid performing FNA in 61 patients (42%).
Conclusions: Modified EUS lymph-node criteria are more accurate than standard criteria. A selective EUS-FNA approach reduced the cost by avoiding EUS-FNA in 42% of patients with esophageal carcinoma. These results require confirmation in future studies.