Strain elastography for prediction of breast cancer tumor grades.

OBJECTIVES

The purpose of this study was to determine whether the elasticity imaging/B-mode ratio on strain elastography can predict breast cancer tumor grades.

METHODS

A retrospective review of patients with breast lesions who underwent strain elastography and had a diagnosis of breast cancer by image-guided or surgical biopsy was performed. The axis of the maximum elastographic dimension was compared to the B-mode dimension to form an elasticity imaging/B-mode ratio. Lesions were categorized according to their pathologic type, including atypical ductal hyperplasia (ADH), mucinous or colloid cancer, ductal carcinoma in situ (DCIS), grade I invasive ductal carcinoma (IDC), grade II IDC, grade III IDC, invasive lobular carcinoma (ILC), and lymphoma. The mean elasticity imaging/B-mode ratio of each tumor type was calculated. The elasticity imaging/B-mode ratio of the tumor was compared to the tumor type by Kruskal-Wallis and Tukey-Kramer tests (lymphoma and ADH excluded because of small numbers).

RESULTS

Tumor grades included lymphoma (n = 3), ADH (n = 2), mucinous cancer (n = 11), DCIS (n = 19), IDC (grades I-III; n = 200), and ILC (n = 31). The mean elasticity imaging/B-mode ratio varied with increasing tumor grade. Tumor grades could not have been selected at random from one population (P < .0001, χ(2) test). Invasive lobular carcinoma and grade III IDC were statistically different from mucinous or colloid cancer, DCIS, and grade I and II IDC.

CONCLUSIONS

The elasticity imaging/B-mode ratio on strain elastography is related to the tumor grade.