Introduction: The use of fine needle aspiration has improved the staging accuracy of EUS in evaluating lymph node invasion, although the technique prolongs the duration of the EUS and increases the cost. The use of selective EUS guided fine needle aspiration has been suggested as an approach to maximize the yield and minimize duration and cost The success of such an approach would rely on the ability of EUS, without aspiration, to identify benign lymph nodes. Therefore, this ability has to be re-evaluated. Hypothesis: EUS can through the presence of certain endosonographic features (small size, non-oval shape, hotnoechogeneity) identify benign lymph nodes, without aspiration. Methods: To test this hypothesis, a retrospective review of all patients who underwent EUS for cancer staging at our institution between 12/95-12/96 was performed. Those patients whose lymph nodes were evaluated by EUS but not surgically retrieved and histologically examined were excluded from the study. The EUS images of the lymph nodes that fit the inclusion criteria were reviewed. The size, shape and echotexture of each of the lymph nodes of interest were recorded. Small size (cross-sectional area < 25 mm2), non-oval shape and homoechogeneity were defined as features of benign lymph nodes. Large size (cross-sectional area > 25 mm2), oval shape, and inhomoechogeneity were defined as features of malignant lymph nodes. 2×2 tables were constructed to evaluate the sensitivity, specificity, PPV and NPV of each of the features of interest (small size, non-oval shape, homoechogeneity) using histology as the "gold standard". Results: 15 evaluable lymph nodes were identified (2 subcarinal, 6 mediastinal, 2 pen-gastric, 4 celiac mink, 1 perirectal) 3 of the lymph nodes were benign and 12 were malignant on histological evaluation. Small size correctly identified 1/3 benign and 12/12 malignant lymph nodes (sen 33%, spec 100%, PPV 100%, NPV 85%). Non -oval shape correctly identified 0/3 benign and 11/12 malignant lymph nodes (sen 0%, spec 92%, PPV 0%, NPV 79%). Homoechogeneity correctly identified 0/3 benign and 5/12 malignant lymph nodes (sen 0%, spec 42%, PPV 0%, NPV 62.5%). Conclusions: In this study, small size (cross-sectional area <25 mm2) was an endosonographic criterion able to identify benign lymph nodes, without aspiration.
ASJC Scopus subject areas
- Radiology Nuclear Medicine and imaging