| Popis: |
Introduction: Breast cancer is a leading cause of death in women of all ages in many parts of the world. The management of this disease is typically based upon its clinico-pathological features and usually involves a combination of surgery and systemic therapy. Many patients however, do not respond to these therapies as anticipated and may therefore suffer needlessly from medication side effects and delayed initiation of more effective treatment, all of which impose an enormous financial burden on healthcare systems. Efforts are underway to improve the classification, prognostication and prediction of response to treatment for breast cancer patients, in the hope of providing better, individualised care Advances in our understanding of the disease and its management are anticipated to come from investigations into the molecular pathways and gene expression underlying breast cancer development, growth and metastasis. Microarray technology, used to simultaneously analyse the patterns of expression of tens of thousands of genes from tumour biopsies, has permitted the identification of new intrinsic subtypes of breast cancer based upon their transcriptional signatures, and of prognostic and predictive markers that are beginning to show clinical utility and promise to outperform standard clinico-pathological markers. Methods: Microarray and quantitative polymerase chain reaction (qPCR) technologies were used to compare the efficacy of five multi-gene signatures for their ability to predict clinical response to three months of neoadjuvant treatment with the Aromatase inhibitor Letrozole, in a population of postmenopausal ER positive breast cancer patients. The levels of gene expression in biopsy samples acquired from each patient were measured prior to treatment and at week two of a three-month treatment regimen. Tumour response was assessed dynamically by means of three-dimensional ultrasound scanning. The predictive capacity of proliferation markers was further explored by focusing on the expression of several key genes involved in different stages of cell cycle progression, including Cyclins B1, A2, D1, CDKs 1, 2 and 4 and the NUSAP1 and Ki67 genes. In addition, changes in Ki67 after 3 months of neo-adjuvant treatment were related to long-term survival. Results: A total of 394 women with large or locally advanced oestrogen receptor positive breast tumours were enrolled into the study population in two independent datasets. A previously defined neoadjuvant predictive signature from our unit has been validated on both independent and extended datasets. In addition, two further proliferative-gene signatures were shown to have significant power to predict longer-term clinical response to Letrozole therapy, based upon their gene expression activity after two weeks of treatment. On the other hand, neither of the two stromal-gene signatures chosen for inclusion in this study was able to predict response based on their absolute expression profiles at two weeks. However, when considering the change in their gene expression profiles between baseline and two weeks, both stromal signatures became informative in one of the datasets. Only one of the proliferative signatures was predictive in this way. Analysis focusing on the genes representing different phases of the cell cycle, demonstrated that two weeks of Letrozole therapy strongly decreased the expression of Cyclins B1, A2, D1, CDK1 and NUSAP1 but not CDK 2 and 4. Significant correlations between the change in Ki67 and changes in Cyclins B1/CDK1 and Cyclin A2/CDK2 were observed, demonstrating the importance of transcriptional regulation of S and G2-M phase Cyclins/CDKs for the anti- proliferative effect of neoadjuvant Letrozole. Changes in the expression of NUSAP1 and Cyclin B1 significantly correlated with clinical response and Ki67 positivity, especially after two weeks of treatment. The following factors were found to be significantly associated with breast cancer specific survival: T-stage at diagnosis; Nodal status at the time of surgery; and Ki67 at diagnosis and at three months and the corresponding change in Ki67 expression between these time-points. Conclusions: The results presented here demonstrate that changes in the expression of multi-gene transcriptional signatures as well as single genes, can be measured in sequential tumour biopsies taken from patients during neoadjuvant treatment for breast cancer. The ability of several signatures and individual genes to predict the clinical response to neoadjuvant treatment has been compared and evaluated. The difference between the signatures composed of proliferative and stromal genes highlights the different contributions of these components to tumour development and response to treatment. |
