Frederick Wing-Fai Au1, Sandeep Ghai2, Fang-I Lu3, Hadas Moshonov2, Pavel Crystal2. 1. Joint Department of Medical Imaging, University of Toronto, Toronto General Hospital, 585 University Avenue East, NCSB 1C-571, Toronto, Ontario, Canada M5G2N2. Electronic address: frederick.au@uhn.ca. 2. Joint Department of Medical Imaging, University of Toronto, Toronto General Hospital, 585 University Avenue East, NCSB 1C-571, Toronto, Ontario, Canada M5G2N2. 3. Department of Pathology, Sunnybrook Health Sciences Center, University of Toronto, Toronto, Ontario, Canada.
Abstract
RATIONALE AND OBJECTIVES: To correlate prognostic histologic features and immunohistochemical biomarkers of breast cancer with quantitative shear wave elastography (SWE) parameters. MATERIALS AND METHODS: B-mode ultrasound (US) and SWE were performed before core biopsy on 72 cancers in 68 patients. Mean cancer size was determined from US. Histologic grade, lymph node status, lymphovascular invasion (LVI), histologic type, and immunohistochemical biomarkers (estrogen receptor, progesterone receptor, human epidermal growth factor receptor 2 [HER2]) were determined from surgical pathology reports. Correlation between these features and quantitative SWE parameters (mean elasticity [E mean], maximum elasticity [E max], and elasticity ratio [E ratio]) was made. RESULTS: There was significant correlation of mean cancer size with E mean, E max, and E ratio (correlation, 0.492, 0.500, and 0.435, respectively; all P < .001). Lymph node involvement was associated with significantly higher E max (P = .040). LVI was associated with significantly higher E mean, E max, and E ratio (P = .002, .004, and .042, respectively). There was no significant correlation of histologic grade with SWE parameters. HER2+ cancers were associated with significantly higher E ratio (P = .030). In multivariate analysis, only mean cancer size was significantly correlated with E mean and E max (P < .001). CONCLUSIONS: There was significant correlation of cancer size with SWE parameters. There was significant correlation of lymph node status and LVI with SWE, but only on univariate analysis. SWE has the potential to provide prognostic information of breast cancer in a noninvasive manner, but further study is required.
RATIONALE AND OBJECTIVES: To correlate prognostic histologic features and immunohistochemical biomarkers of breast cancer with quantitative shear wave elastography (SWE) parameters. MATERIALS AND METHODS: B-mode ultrasound (US) and SWE were performed before core biopsy on 72 cancers in 68 patients. Mean cancer size was determined from US. Histologic grade, lymph node status, lymphovascular invasion (LVI), histologic type, and immunohistochemical biomarkers (estrogen receptor, progesterone receptor, human epidermal growth factor receptor 2 [HER2]) were determined from surgical pathology reports. Correlation between these features and quantitative SWE parameters (mean elasticity [E mean], maximum elasticity [E max], and elasticity ratio [E ratio]) was made. RESULTS: There was significant correlation of mean cancer size with E mean, E max, and E ratio (correlation, 0.492, 0.500, and 0.435, respectively; all P < .001). Lymph node involvement was associated with significantly higher E max (P = .040). LVI was associated with significantly higher E mean, E max, and E ratio (P = .002, .004, and .042, respectively). There was no significant correlation of histologic grade with SWE parameters. HER2+ cancers were associated with significantly higher E ratio (P = .030). In multivariate analysis, only mean cancer size was significantly correlated with E mean and E max (P < .001). CONCLUSIONS: There was significant correlation of cancer size with SWE parameters. There was significant correlation of lymph node status and LVI with SWE, but only on univariate analysis. SWE has the potential to provide prognostic information of breast cancer in a noninvasive manner, but further study is required.