About 8000 Swedish women are diagnosed with breast cancer each year, and around 1500 die from the disease. Triple-negative breast cancer (TNBC) constitutes 7-14% of all breast cancer and is characterized by an aggressive phenotype with poor prognosis and no targeted therapy available. The aim of this thesis was to evaluate new potential prognostic biomarkers in breast cancer, and with special focus on TNBC.
In paper I and II the importance of the four receptors cKIT, VEGFR2, PDGFRα and PDGFRβ, and of ligand PDGF-CC was explored in a prospectively gathered cohort of primary breast cancer patients. In both papers, tissue micro arrays and immunohistochemistry was used to evaluate protein expression, and in paper I, we also investigated gene copy number using fluorescence in situ hybridization. In paper I, we found that high tumor cell protein expression, but not elevated gene copy number, of cKIT, VEGFR2 and PDGFRα was associated to TNBC. 74% of TNBC tumors displayed high expression of at least one of these three receptors compared to 30 % of non-TNBC. In paper II, we showed that high expression of the PDGF receptors α and β, and ligand PDGF-CC correlated to several prognostic patient and tumor characteristics related to tumor inherent biological aggressiveness (e.g. hormone receptor negativity and higher tumor grade). Neither of the receptors investigated in paper I or II were associated to survival in TNBC but interestingly, in the whole cohort we found that patients with high expression of ligand PDGF-CC in the primary tumor had increased risk of 5-year distant-recurrence.
In papers III and IV, we investigated circulating tumor cell (CTC) count and morphologic CTC characteristics as prognostic markers in patients with newly diagnosed metastatic breast cancer (MBC) scheduled for 1st line systemic therapy. The CellSearch system was used for CTC isolation and characterization. In paper III, only patients with baseline (BL) CTC count ≥5 before initiation of therapy were included (N=52). We found that presence of apoptotic CTCs and CTC-clusters during treatment (but not at BL) was associated with a significantly worse prognosis. We also found that at BL, TNBC and HER2+ patients had CTC-clusters present more frequently than hormone receptor positive patients. In paper IV, 156 patients were included (irrespective of BL CTC count), and we showed that CTC count ≥5, and presence of CTC-clusters were prognostic for PFS and OS at BL and during the first 6 months of systemic therapy following diagnosis of MBC. Also, changes in CTC count during therapy significantly correlated to response evaluation and survival. Finally, both factors independently added value at all time points to a prognostic model based on clinicopathological variables.
In conclusion, paper I and II present support for the involvement of cKIT, VEGFR2, PDGFRα and PDGF-CC in TNBC. These receptors are not prognostic markers in TNBC, but they are upregulated and further studies are encouraged to elucidate their values as predictive markers and possible drug targets in TNBC. Paper III and IV show the clinical value of CTC count and CTC-cluster detection before and during 1st line systemic therapy for prognosis and treatment monitoring in patients with newly diagnosed MBC. Our results highlight the importance of serial monitoring of these variables as the prognostic value of both CTC count and CTC-cluster detection increased over time.