Oral Presentation Annual Meetings of the Endocrine Society of Australia and Society for Reproductive Biology and Australia and New Zealand Bone and Mineral Society 2016

Steroid signatures in breast cancer reveals patterns associated with Ki67 labelling index of carcinoma cells (#194)

Keely McNamara 1 , Ju-Yeon Moon 2 , Hironobu Sasano 1 , Man-Ho Choi 2
  1. Tohoku University School of Graduate Medicine, Sendai-Shi, Miyagi-Ken, Japan
  2. Molecular Recognition Research Center, , Korea Institute of Science and Technology,, Seoul, , Korea.

The highly estrogen dependant nature of breast carcinoma is reflected in the traditional histopathological classification, as well as in the modern microarray approach. Despite this estrogen centred view there has been a growing awareness of the potential significance, both biologically and therapeutically, of other steroid pathways. Therefore in this study we examined a panel of eight common and inter-related sex steroids in a series of 38 breast cancer cases drawn from the four classical tumour types (Luminal A, Luminal B, TNBC, HER2) with nine cases having matched normal breast tissue. The approach we used was using GC-MS analysis of frozen sections (≈200 microns of tissue). This approach has the advantage that is potentially allows us to study histological sections, RNA levels and steroid levels in parallel from adjacent serial frozen tissue sections within the one particular sample. In the analysis of matched normal and cancer specimens the precursor steroids dehydroepiandrostenedione (DHEA), Androstenedione and Pregnenolone  were significantly higher in the normal tissues (p<0.01), Testosterone and Estrone were unchanged between cancer and normal while Estradiol and Dihydrotestosterone (DHT) were both significantly increased in cancer specimens (p<0.03). Specific steroids were enriched in different tumour types with TNBC showing an enrichment of DHT while Luminal A cancers were enriched in DHEA and Pregnenolone. A hierarchical clustering approach, similar to that used by microarrays, revealed three distinctive groups; these tentative categories were associated with significant differences of Ki-67 labelling index (p<0.05). Of particular interest, the clusters were not a duplication of the standard tumour subtyping as there was no significant association between these two factors (p=0.5). This study demonstrates two things. Firstly, that broader steroid signatures rather than isolated steroids give us insights into breast cancer and secondly, that hierarchical clustering of steroid levels may provide meaningful information on tumour biology.