Serum proteomic prediction of progression-free survival in HER2-negative metastatic breast cancer patients receiving docetaxel as first-line treatment

Abstract

1010 Background: Docetaxel has become a standard in management of metastatic breast cancer (MBC) patients (pts). However, there is no validated clinical or biological marker allowing to predict which pts are most likely to benefit from it. We evaluated the ability of Surface enhanced laser desorption/ionization-time of flight mass spectrometry (SELDI-TOF MS) to identify serum proteins that could predict for docetaxel activity. Methods: We prospectively collected pretreatment serum of 81 HER2- MBC pts receiving docetaxel as first-line treatment. Serum samples were denatured, and then incubated in triplicates with three different ProteinChip arrays (Ciphergen): H50, IMAC30-Cu and CM10. Arrays were analysed using a PBSIIc ProteinChip reader (Ciphergen). Results: Pt characteristics were the following: median age, 53 ys (34 -71); Hormonal Receptivity (HR) (yes/no/na, %) 76/15/9; Disease-free interval (DFI) (≤24 months/>24 months, %) 42/58; visceral disease (yes/no, %) 35/65; metastatic sites (<3/≥3,%) 85/15; Median follow-up, 19.8 months (95%CI, 16.1–21.9); Median Progression-free survival (PFS), 12.9 months (95%CI, 8.8–18). Overall, 521 protein peaks were resolved. Univariate Cox proportional hazards regression identified 20 proteins significantly associated to PFS (P < 0.01). A stepwise procedure using the Akaike information criterion was applied to build a Cox model with 7 protein peaks allowing to calculate a risk index for each pt. Using the risk index median value as cut-off, high-risk and low-risk populations with dramatic differences in PFS were identified (2-year PFS of 0% and 54.6%, p = 1.45 10–8). Validation methods included leave-one-out cross validation and iterative resampling. In a multivariate Cox regression including conventional prognostic factors (HR, DFI, visceral metastasis) and serum protein risk index, the latter retained the strongest independent prognostic significance for PFS. Identification of proteins with differential expression according to clinical outcome is ongoing. Conclusions: Serum proteomic profiling may help select MBC Pts who will benefit from treatment with docetaxel. Validation on an independent dataset is planned. No significant financial relationships to disclose.