Predictors of response and follow up biomarkers for metronomic chemotherapy with cyclophosphamide and celecoxib in advanced breast cancer patients

PERROUD HA; ALASINO CM; MENACHO-MARQUEZ MA; MAINETTI LE; QUERALT F; PEZZOTTO SM; ROZADOS VR; SCHAROVSKY OG

San Antonio, TX

Simposio; 38th San Antonio Breast Cancer Symposium; 2015

American Association for Cancer Research

Lowdose metronomic chemotherapy (MC) with Cyclophosphamide (Cy) and Celecoxib(Cel) has demonstrated to be effective and welltolerated in advanced breast cancer patients (ABCP) but predictive markers of response or followup are lacking. Given the antiangiogenic properties of MC we analyzed several angiogenesisrelated biomarkers and evaluated their potential role as predictors of response or treatment followup of ABCP treated with MC. Treatment plan: Patients received Cy 50 mg p.o./day + Cel 200 mg p.o./bid. Cellular parameters: Circulating endothelial cells (CEC) and Circulating progenitor endothelial Cells (CEP) were determined by Flow Cytometry. Serologic parameters: Serum levels of vascular endothelial growth factor (VEGF), VEGFC, soluble VEGF Receptors 2and 3 (sVEGFR2, sVEGFR3) and Thrombospondin1 (TSP1) were determined byELISA. Blood samples were collected before and during treatment. Twenty patients wereenrolled. Response Rate was 5% and Clinical Benefit (CB) 55%. Most of the patients showed prolonged stable disease (SD≥24 weeks). Biomarkers were determined in all patients. Levels of CEC and CEP showed no clear trend variations during treatment. However, levels of CEC significantly increased at the time of disease progression in those patients who showed CB (P=0.014). Also baseline values of CEC and CEP showed marginally significant associations withTime To Progression. Serum VEGF concentration decreased during treatment (P=0.050) while sVEGFR2 increased (P=0.005). VEGFC, sVEGFR3 and TSP1showed nonsignificant variations. VEGF/sVEGFR2 decreased during treatment (P=0.041), whereas VEGF/TSP1, and VEGFC/sVEGFR2 ratios showed nonsignificantvariations. Baseline values of VEGF, and VEGF/sVEGFR2 0.0354 and P=0.0300,respectively) while sVEGFR2 did not. When considering the two variables together, the goodness of prediction was not improved. To confirm the value of baseline VEGF and VEGF/sVEGFR2 as predictors of response, we used the 50th percentile as a cutoff value to analyze the % of progression free survival. Patients with values lower than the 50th percentile for both biomarkers showed longer TTP (P=0.0001 and P=0.0008, respectively). The treatment had antiangiogenic effect (VEGF decrease and sVEGFR2 increase). The absence of variation in VEGFC and sVEGFR3 would indicate the lack of effect onlymphangiogenesis. Increased levels of CEC could be useful for detecting progression. If confirmed with a higher number of patients, baseline VEGF and VEGF/sVEGFR2 values could be useful as early predictors of response.