T S Mehta1, S Raza. 1. Department of Radiology, Beth Israel Deaconess Medical Center, Boston, MA 02215, USA.
Abstract
OBJECTIVE: The purpose of our study was to determine how breast cancer vascularity, as revealed by power Doppler sonography, correlates with lymph node involvement and lymphatic vascular invasion. MATERIALS AND METHODS: Breast sonograms obtained during a 2-year period were retrospectively reviewed. Patients who underwent power Doppler sonography of solid masses and had biopsy-proven carcinoma composed our study population. Power Doppler findings were categorized according to the presence or absence of vessels. Pathologic findings were reviewed for lymph node involvement and lymphatic vascular invasion. RESULTS: Of 176 patients with breast cancer, vessels were seen on power Doppler sonography in 128 (73%) and not seen in 48 (27%). Of 126 patients who underwent lymph node dissection, vessels were seen in 97 (77%) and not seen in 29 (23%). Lymph node involvement was seen in 42 (43%) of the 97 patients in whom vessels were seen and in three (10%) of the 29 in whom vessels were not seen. Of 150 patients examined for lymphatic vascular invasion, vessels were seen in 111 (74%) and not seen in 39 (26%). Lymphatic vascular invasion was seen in 47 (42%) of the 111 patients in whom vessels were seen and in five (13%) of the 39 in whom vessels were not seen. CONCLUSION: Tumor vascularity revealed by power Doppler sonography correlated strongly with detection of lymph node involvement and lymphatic vascular invasion, with sensitivities of 93% and 90%, respectively. However, the specificities were low, at 32% and 35%, respectively. More important, patients with breast cancer in whom vessels were not revealed by power Doppler sonography also were unlikely to have lymph node involvement and lymphatic vascular invasion: Negative predictive values were 90% and 87%, respectively.
OBJECTIVE: The purpose of our study was to determine how breast cancer vascularity, as revealed by power Doppler sonography, correlates with lymph node involvement and lymphatic vascular invasion. MATERIALS AND METHODS: Breast sonograms obtained during a 2-year period were retrospectively reviewed. Patients who underwent power Doppler sonography of solid masses and had biopsy-proven carcinoma composed our study population. Power Doppler findings were categorized according to the presence or absence of vessels. Pathologic findings were reviewed for lymph node involvement and lymphatic vascular invasion. RESULTS: Of 176 patients with breast cancer, vessels were seen on power Doppler sonography in 128 (73%) and not seen in 48 (27%). Of 126 patients who underwent lymph node dissection, vessels were seen in 97 (77%) and not seen in 29 (23%). Lymph node involvement was seen in 42 (43%) of the 97 patients in whom vessels were seen and in three (10%) of the 29 in whom vessels were not seen. Of 150 patients examined for lymphatic vascular invasion, vessels were seen in 111 (74%) and not seen in 39 (26%). Lymphatic vascular invasion was seen in 47 (42%) of the 111 patients in whom vessels were seen and in five (13%) of the 39 in whom vessels were not seen. CONCLUSION: Tumor vascularity revealed by power Doppler sonography correlated strongly with detection of lymph node involvement and lymphatic vascular invasion, with sensitivities of 93% and 90%, respectively. However, the specificities were low, at 32% and 35%, respectively. More important, patients with breast cancer in whom vessels were not revealed by power Doppler sonography also were unlikely to have lymph node involvement and lymphatic vascular invasion: Negative predictive values were 90% and 87%, respectively.