In-vitro high-resolution ultrasonography of benign and malignant lymph nodes. A sonographic-pathologic correlation

The authors assess the value of combining high-resolution ultrasonography (HRUS) findings in a scoring scale for distinguishing malignant from reactive lymphadenopathy and explain the pathologic causes of altered nodal sonographic architecture.

Sixty-one nodes obtained from 32 consecutive patients were prospectively scanned with 7.5-MHz ultrasound probes in a waterbath. Three sonographic features--long-to-short axis ratio (L/S), hilar width, and cortical width--were graded on a 5-level scoring scale (0-4). Nodes scored > or = 3 were considered malignant and < or = 2 benign. Subsequently, all nodes were microsectioned in a plane matching the sonograms, allowing direct sonographic-pathologic correlation.

Eighty-two percent of nodes were correctly characterized using the above cut-off point (sensitivity: 87%, specificity: 74%). Eighty-three percent of nodes scored 4 were malignant and 95% scored 0 were benign. Eighty-two percent of nodes with L/S < 2, 81% with no hilus, and 70% with eccentric cortical widening were malignant, whereas 72% with L/S > or = 2, 86% with a wide hilus, and 91% with a narrow cortex were benign. Sonographic-pathologic correlation showed that tumor infiltration results in rounded nodal shape, loss of hilar echogenicity and cortical widening, whereas reactive disease tends to preserve nodal architecture.

Morphologic changes detectable using HRUS help distinguish benign from malignant lymph nodes.