Association of serum brain-specific protein fragments and brain metastases in patients with metastatic breast cancer.

e14007 Background: Brain metastases (BrMs) occur frequently in patients with breast cancer (BC) and represent one of the most difficult-to-treat events. Novel clinical tools are needed to identify BC patients at high risk of developing BrMs. Blood-based liquid biopsies represent a minimally invasive procedure that may detect surrogate markers of pathological changes occuring in the brain as a result of BrMs. Here we evaluated serum levels of a panel of brain-specific protein fragments in metastatic BC patients with BrMs, with no-BrMs and controls. Methods: Our cohort includes stage IV BC patients with BrMs (n= 27; median age 51 y.o.) and no-BrMs (n=50; median age 53.5 y.o.), and healthy controls (H) (n=23, median age 50 y.o.). Serum levels of 5 brain-specific protein fragments were measured by immunoassays and compared between groups. No-BrMs samples were collected at the time of metastatic disease diagnosis; BrMs samples were collected at BrMs diagnosis and before local treatment. Association with overall survival (OS) was evaluated by univariate Cox regression analysis. Results: All markers were significantly higher in patients with no-BrMs compared to H. Only Neurocan, Tenascin-R and GFAP were significantly higher in patients with BrMs compared to H. When comparing metastatic BC patients, significantly higher levels of Tau-A, Tau-C, Neurocan and GFAP were found in no-BrMs patients versus BrMs (Table). Association with OS was not found in BrMs patients, although a trend for high Tau-A levels to predict poor OS was detected in the no-BrMs group (HR 4.4, 95%CI 0.9-21.3, p=0.068). Conclusions: Protein fragments expected to be derived from brain remodeling consequent to BrMs were elevated in BC patients both with BrMs and no-BrMs at the time of sampling, with an overall trend for the highest levels to be found in patients with no-BrMs. This may suggest that brain remodeling processes are associated with BC and could partly explain the relatively high incidence of BrMs relapse in BC. Further studies requiring samples collected pre- and post-treatment are needed to further validate the diagnostic and prognostic value of these non-invasive biomarkers in BC BrMs setting. [Table: see text]