Molecular subtypes of breast cancer: metabolic correlation with 18F-FDG PET/CT

Abstract

To determine whether the metabolic features of breast tumours differ among molecular subtypes. This prospective study included 168 women diagnosed with locally advanced breast cancer. PET/CT was requested in the initial staging before neoadjuvant treatment (multicentre study, FISCAM grant). All patients underwent an ¹⁸F-FDG PET/CT scan with a dual time-point acquisition. Both examinations (PET-1 and PET-2) were evaluated qualitatively and semiquantitatively with calculation of SUVmax (SUV-1 and SUV-2, respectively), and the percentage variation in the SUVs and retention indexes (RI) between PET-1 and PET-2 in the breast tumour were calculated. Biological prognostic parameters, including the steroid receptor status, HER-2 expression, proliferation rate (Ki-67) and grading, were determined from primary tumour tissue. Tumour subtypes were classified following the recommendations of the 12th International Breast Conference, by immunohistochemical surrogates as luminal A, luminal B-HER2(-), luminal B-HER2(+), HER2(+) or basal. Metabolic semiquantitative parameters and molecular subtypes were correlated. Of the 168 tumours, 151 were classified: 16 were luminal A, 53 were luminal B-HER2(-), 29 were luminal B-HER2(+), 18 were HER2(+) and 35 were basal. There were significant differences between SUV-1 and SUV-2 and the different subtypes, with higher SUVs in HER2(+) and basal tumours. No significant differences were found with respect to RI. Semiquantitative metabolic parameters showed statistically significant differences among the molecular subtypes of the tumours evaluated. Therefore, there seems to be a relationship between molecular and glycolytic phenotypes.