OBJECTIVE: To compare the diagnostic ability of ultrasonography (US) and MR imaging for discriminating squamous cell carcinoma (SCC) nodes with extranodal spread (ENS) in the neck. METHODS: US and MR imaging was retrospectively evaluated for differentiating ENS-positive (n=28) from ENS-negative (n=26) SCC nodes (>10mm short-axis diameter) in 50 patients with head and neck SCCs. We assessed nodal size on US and MR images; irregular nodal margin on US; and vanishing nodal border, flare, and shaggy nodal margin signs on T1-, fat-suppressed T2-, and contrast-enhanced T1-weighted MR images, respectively. US and MR images were analyzed by 3 radiologists in consensus and the results were compared between ENS-positive and ENS-negative SCC nodes. RESULTS: The nodal sizes of ENS-positive nodes (21±9 mm) were significantly larger than those of ENS-negative SCC nodes (14±4 mm) (p<0.001). Irregular nodal margins were more frequently observed in ENS-positive SCC nodes (75%) than in ENS-negative SCC nodes (12%). The vanishing nodal margin, flare, and shaggy nodal margin signs were more frequently observed in ENS-positive SCC nodes (93%, 89%, and 82%, respectively) than in ENS-negative nodes (46%, 19%, and 19%, respectively). A combination of size (≥22 mm) and imaging criteria (irregular margin or flare sign) best discriminated ENS-positive SCC nodes with 82% sensitivity, 89% specificity, and 85% accuracy for US and 89% sensitivity, 81% specificity, and 85% accuracy for MR imaging. CONCLUSION: US discriminated ENS-positive from ENS-negative SCC nodes with comparable accuracy and higher specificity than MR imaging.
OBJECTIVE: To compare the diagnostic ability of ultrasonography (US) and MR imaging for discriminating squamous cell carcinoma (SCC) nodes with extranodal spread (ENS) in the neck. METHODS: US and MR imaging was retrospectively evaluated for differentiating ENS-positive (n=28) from ENS-negative (n=26) SCC nodes (>10mm short-axis diameter) in 50 patients with head and neck SCCs. We assessed nodal size on US and MR images; irregular nodal margin on US; and vanishing nodal border, flare, and shaggy nodal margin signs on T1-, fat-suppressed T2-, and contrast-enhanced T1-weighted MR images, respectively. US and MR images were analyzed by 3 radiologists in consensus and the results were compared between ENS-positive and ENS-negative SCC nodes. RESULTS: The nodal sizes of ENS-positive nodes (21±9 mm) were significantly larger than those of ENS-negative SCC nodes (14±4 mm) (p<0.001). Irregular nodal margins were more frequently observed in ENS-positive SCC nodes (75%) than in ENS-negative SCC nodes (12%). The vanishing nodal margin, flare, and shaggy nodal margin signs were more frequently observed in ENS-positive SCC nodes (93%, 89%, and 82%, respectively) than in ENS-negative nodes (46%, 19%, and 19%, respectively). A combination of size (≥22 mm) and imaging criteria (irregular margin or flare sign) best discriminated ENS-positive SCC nodes with 82% sensitivity, 89% specificity, and 85% accuracy for US and 89% sensitivity, 81% specificity, and 85% accuracy for MR imaging. CONCLUSION: US discriminated ENS-positive from ENS-negative SCC nodes with comparable accuracy and higher specificity than MR imaging.