Clinical Validity of Circulating Tumor Cell (CTC) Enumeration in 841 Subjects with Metastatic Breast Cancer (MBC).
M. C. Liu,M. Mego,S. Nakamura,F. Nole,J. Pierga,M. Toi,E. Munzone,H. Yagata,M. T. Sandri,F. Bidard,H. Wang,D. F. Hayes,M. Cristofanilli
DOI: https://doi.org/10.1200/jco.2011.29.15_suppl.10592
IF: 45.3
2011-01-01
Journal of Clinical Oncology
Abstract:10592 Background: CTCs are associated with poor clinical outcomes in MBC. Multiple technologies have been evaluated for CTC capture, enumeration, and characterization, but only one (CellSearch; Veridex) has been evaluated independently by different investigators with serial assessments and longitudinal follow-up in a defined prospective clinical trial. We combined these datasets to further validate the prognostic and predictive significance of this biomarker in MBC. METHODS Six studies were identified in the peer-reviewed published literature; all used the same CTC decision point (>=5 CTC/7.5 mL blood) and clinical endpoints (PFS, OS) in patients with progressive MBC. CTC enumeration was performed with the FDA cleared immunomagnetic/immunofluorescence assay (CellSearch). Study teams contributed blinded data to create a dataset of 841 subjects with CTC results linked to clinicopathologic variables and survival. Analyses were defined prior to data collection and conducted in subsets where pertinent data were available. RESULTS >=5 CTCs predicted for worse PFS at baseline (HR 1.32; CI 1.13, 1.55) and 2-5 (HR 2.16; CI 1.67, 2.78), 6-8 (HR 2.75; CI 2.05, 3.70), 9-12 (HR 2.71; CI 2.08, 3.52), 13-20 (HR 2.64; CI 2.03, 3.42), and >=21 wks (HR 1.88; CI 1.45, 2.42) from initiating new therapy (tx). OS results were also significant. The predictive value of CTC level was not affected by type (chemotx, endocrine tx, biologic tx) or line (first, second, >=third) of tx, presence/absence of bone metastases, or presence/absence of disease limited to bone. Multivariate Cox regression analysis was performed with the baseline CTC level adjusting for clinicopathologic variables to predict PFS and OS; some factors maintained an association, but CTC level was the strongest predictor of outcomes. CONCLUSIONS This analysis substantiates the prognostic and predictive utility of CTCs assessed before and during tx for MBC. Persistently elevated CTC levels are highly associated with treatment failure irrespective of clinicopathologic variables, disease subtype, and type or line or tx. Serial CTC assessments should be used as a surrogate marker for patient outcomes to assist in selecting appropriate treatment regimens in MBC.