Background: Despite advances in the treatment of breast cancer, there remains a significant clinical need for improved therapeutic strategies for patients with the most aggressive types of tumors. In particular, knowledge of the specific genes and pathways driving tumor growth in these patients would allow for treatment with therapeutics directed against these molecular targets. To this end, we reasoned that through the integrated analysis of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and gene expression profiling we could understand the molecular mechanisms associated with an aggressive imaging phenotype and thus gain insight into potential therapeutic targets for these patients.
Materials and Methods: We studied 61 patients with locally advanced breast cancer, for whom DCE-MRI scans and core biopsies were available prior to the start of neoadjuvant chemotherapy. To analyse the DCE-MRI data, we used Tofts' pharmacokinetic (PK) model to quantify the rate constant kep that governs the washout of contrast agent from the tumor extravascular extracellular space. We chose to focus on the PK parameter kep since our analysis showed that it can be estimated reliably from the low temporal resolution diagnostic DCE-MRI scans that are routinely performed in the clinic. We extracted mRNA from formalin fixed paraffin embedded core biopsy samples and measured gene expression using Affymetrix U133 whole genome arrays. Following normalization and pre-processing, we used significance analysis of microarrays (SAM) to determine which genes were statistically significantly correlated with median kep.
Results: Using a local false discovery rate of 5% resulted in a total of 328 genes that were significantly positively correlated with median kep. These included copper transporter like solute carrier family31member2 (SLC31A2), cancer stem cell (CSC) related genes such as CD44, aldehyde dehydrogenase family 1 member A3 (ALDH1A3), and integrin alpha-6 (ITGA6), hypoxia regulated genes such as hypoxia inducible factor 1a (HIF1a), kinases such as eukaryotic translation initiation factor2-alpha kinases (EIF2AK2, EIF2AK1), and pyruvate dehydrogenase kinases (PDK1, PDK3).
Discussion: Our results illustrate how functional imaging modalities such as DCE-MRI can be combined with gene expression profiling to provide insight into molecular targets that may have important therapeutic implications in breast cancer. We found that locally advanced breast cancers with high vascular permeability and/or blood flow, as quantified by the washout parameter median kep, were associated with an up-regulation of genes related to CSCs and copper metabolism, the latter of which is known to play a role in angiogenesis. In addition, up-regulation of hypoxia related genes such as HIF1a, EIF2AK1&2 and PDK1&3 may promote tumor survival and spread under stress. Our results suggest that locally advanced breast cancers with high vascular permeability and/or blood flow may benefit from therapies directed at one or more of these molecular targets. Furthermore, functional imaging with DCE-MRI may be helpful as a noninvasive means of selecting and monitoring therapy against these targets.
Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P4-01-01.
This abstract was not presented at the conference.