Ricerc@Sapienza

OBJECTIVES: The role of functional techniques, such as magnetic resonance spectroscopy (H-MRS), as noninvasive tools to increase breast MR imaging reliability has been widely investigated during the last 2 decades. Considering the growing interest in tumor biology and its influence on functional parameters, the aim of this study was to investigate the relationship between H-MRS parameters and breast cancer biomarkers and to evaluate whether the results of H-MRS at 3 T can correlate with established breast cancer prognostic factors in our clinical experience. MATERIALS AND METHODS: One hundred two patients with biopsy-proven breast cancer underwent 3 T breast MR imaging. Single-voxel H-MRS was performed after the T1-weighted sequence, using a PRESS water-suppressed sequence (BREASE). Data were collected from a single rectangular volume of interest that encompassed the lesion. Magnetic resonance images and spectra of 102 Breast Imaging Reporting and Data System 6 lesions were prospectively evaluated by 2 radiologists in consensus. H-MRS results were considered positive if the choline peak signal-to-noise ratio was 2 or higher. H-MRS findings were then compared with morphological features and to histological findings, such as lesion size, nuclear grade, Ki-67, hormone receptor status, and Her2 expression. RESULTS: Elevated levels of total choline were detectable in 68/102 cases (66.67%) and undetectable in 34/102 (33.33%). A statistically significant association between the presence of choline peak and higher tumor grading (P < 0.0001), greater Ki-67 value (P < 0.0001), and larger lesion size (P < 0.0001) was found. No statistically significant associations were observed between choline peak and the luminal subgroups, even if higher levels of choline were more frequent in nonluminal A lesions. CONCLUSIONS: Our study confirms that 3 T breast H-MRS can be a valid additional tool to obtain further information about breast cancer biology and to predict tumor aggressiveness, because the detection of elevated levels of total choline in the spectrum is associated with a biologically aggressive breast cancer phenotype (large dimensions, grade 3, high values of Ki-67). Our results need to be validated in standardized larger-scale studies.