Feed

Background: Conventional axillary ultrasound (US) is least reliable in lymph nodes with diffusely thickened cortex. We evaluated whether contrast-enhanced ultrasound with ultra-resolution microvascular imaging (CEUS-URM) improves discrimination of metastatic nodes in this subgroup. Materials and methods: This was a prospective single-center study of patients with histologically confirmed breast cancer; one index (most suspicious) node per patient (unit of analysis = index node). Two separately recruited consecutive cohorts were analyzed: a US-only cohort (n = 181; diffuse-thickening subgroup with histology, n = 52) and a subsequent CEUS-URM cohort (n = 42; 15 metastatic). Surgical histopathology (SLNB/ALND) was the reference standard. A CEUS-URM score was built from data-driven, Youden-optimized cut-offs (URM vessel count ≥8; DV mean density ≥ 13.05) and internally validated by bootstrap with optimism correction. As the cohorts were separate, the US-versus-CEUS comparison is cross-cohort and exploratory. Results: Within the diagnostically challenging subgroup of lymph nodes with diffusely thickened cortex, conventional axillary US alone demonstrated limited discriminatory performance for metastatic involvement (AUC 0.43). CEUS-derived quantitative parameters significantly improved diagnostic accuracy, with the best individual parameter achieving an AUC of 0.68. A simple CEUS score combining hypervascular vessel count and vascular density provided the highest diagnostic performance (AUC 0.81, 95% CI 0.68-0.92). At a low threshold, the CEUS score showed high sensitivity (93%), suitable for screening and exclusion of nodal metastases, while at a higher threshold it achieved high specificity (96%), allowing reliable confirmation of metastatic disease. Conclusions: In this exploratory study, a simple CEUS-URM score improved discrimination of diffusely thickened axillary nodes and may serve as an adjunct to conventional US. The findings are preliminary-derived and tested in the same small cohort-and require external, within-patient paired validation.