Prediction of Molecular Subtypes Using Superb Microvascular Imaging and Shear Wave Elastography in Invasive Breast...

To investigate the efficacy of the advanced imaging methods, superb microvascular imaging (SMI) and shear wave elastography (SWE) in predicting molecular subtypes in invasive breast carcinomas.

A total of 210 biopsy-proven breast carcinomas in 200 patients who underwent ultrasound (US) imaging with SMI and SWE were included in this study. Quantitative analyses were performed using mean elasticity (E_mean) score by SWE and vascular index (VI) by SMI. For qualitative assessment of microvascularity, first, lesions were graded according to Adler's classification in four types. Then, a new morphological model was used to classify the microvascular architecture into six patterns: type one, no signal; type two, penetrant; type three, rim-like; type four, dot-like/linear/regional; type five, wheel-like and type six, irregular signals. The correlation between these variables and molecular subtypes, nuclear grade, the Ki-67 levels and axillary status was investigated.

The average VI and E_mean values were relatively higher in non-luminal subtypes (VI, p = 0.002; E_mean, p > 0.05). The two microvascularisation models were significantly able to differentiate the molecular subtypes according to the Kruskal Wallis test (p < 0.05). Rim-like, penetrant and regional patterns were primarily observed in luminal subtypes. The dominant pattern in non-luminal subtypes was wheel-like pattern. VI, E_mean, Adler's classification and morphological vascularisation model were not significantly correlated with the nuclear grade, Ki-67 index or axillary status.

The proposed microvascular categorization model may be more valuable in predicting molecular subtypes of breast carcinomas compared to VI and E_mean and may contribute to the management of breast carcinomas as a non-invasive variable.