Purpose: We examined whether pretreatment levels of angiogenesis- or hypoxia-related proteins and their changes after one cycle of first-line bevacizumab-based therapy were associated with response, PFS, or OS in patients with metastatic breast cancer.
Experimental Design: We included 181 patients enrolled in the phase II ATX trial evaluating first-line paclitaxel and bevacizumab without or with capecitabine (NTR1348). Plasma samples were analyzed for VEGF-A, soluble VEGFR2 (sVEGFR2), angiopoietin 2 (ANG2), soluble TIE2 (sTIE2), IL6, IL8, and carbonic anhydrase 9 (CA9). Baseline serum CA15-3 was documented. HR was adjusted for confounding factors. Where appropriate, an optimal cut-off value defining a high and a low group was determined with Martingale residuals.

Results: At baseline, multiple proteins were significantly associated with PFS (ANG2, IL6, IL8, CA9, CA15-3) and OS (ANG2, sTIE2, IL6, IL8, CA9, CA15-3). After one cycle, VEGF-A, ANG2, sTIE2, and IL8 significantly decreased, while sVEGFR2 and CA9 significantly increased. The relative change in sVEGFR2 (P=0.01) and IL8 (P = 0.001) was associated with response. Defining optimal cut-off, patients with a high CA9 rise (>2.9%) had better PFS (HR 0.45) and OS (HR 0.54) than those with low/no rise.
Conclusions: Multiple angiogenesis- or hypoxia-related proteins were prognostic for PFS and OS. Molecular agents targeting these proteins might be beneficial in patients with high levels. Changes in IL8 or sVEGFR2 levels at second cycle appear predictive for response. Changes in CA9 levels during bevacizumabbased therapy for prediction of PFS and OS merit further study.

doi.org/10.1158/1078-0432.CCR-15-1005, hdl.handle.net/1765/88264
Clinical Cancer Research
Erasmus MC: University Medical Center Rotterdam

Lam, S. W., Nota, N. M., Jager, A., Bos, M., Van Den Bosch, J., Van Der Velden, A. M. T., … Berends, J. (2016). Angiogenesis- and hypoxia-associated proteins as early indicators of the outcome in patients with metastatic breast cancer given first-line bevacizumab-based therapy. Clinical Cancer Research, 22(7), 1611–1620. doi:10.1158/1078-0432.CCR-15-1005