P2-09-15: Functional Measurements of Tumor Response during Neoadjuvant Chemotherapy Infusion and Early during Treatment Using Diffuse Optical Spectroscopic Imaging.

Abstract

Abstract Background: Non-invasive markers of neoadjuvant chemotherapy response early during treatment would provide physicians a valuable tool to make evidence-based changes to treatment strategies. In a retrospective study of 23 patients presented at SABCS 2011 we demonstrated that Diffuse Optical Spectroscopic Imaging (DOSI) can discriminate non-responding from responding subjects on the first day after the start therapy based on the presence or absence of an oxyhemoglobin flare. Here we present results of an ongoing prospective study designed to confirm the predictive nature and biological origins of oxyhemoglobin flare and to determine if similar functional changes occur at even earlier timepoints such as during infusion. Methods: DOSI was used to measure hemodynamic and metabolic information from tumors and surrounding normal tissue in patients prior to neoadjuvant chemotherapy, during chemotherapy infusion, and daily for the first week of treatment. DOSI uses temporally modulated near-infrared light to determine absolute tissue concentrations of oxyhemoglobin, deoxyhemoglobin, water and lipid content and requires no exogenous contrast agent. Measurements are made using a simple handheld probe placed on the skin. Blood samples were also collected at baseline and daily for seven days after the first infusion and tested for a panel inflammatory cytokines. Patients received paclitaxel + carboplatin + bevacizumab. Overall response to therapy was determined by the decrease in anatomic tumor size. Results: To date three subjects have been measured, two of which have completed neoadjuvant chemotherapy and undergone surgery. One subject was a pathologic complete responder (pCR) and the other a non-responder (NR). In both the pCR subject and the subject still undergoing therapy there were significant functional changes measured in the tumor during infusion. A decrease in oxyhemoglobin and oxygen saturation occurred after paclitaxel infusion (saturation changes: −4.8% and −9.7% for each subject respectively), followed by an increase in these quantities after carboplatin infusion (saturation: 3.8% and 5.9% respectively). In the NR subject both oxyhemoglobin and oxygen saturation had only small fluctuations during infusion (saturation: 1.1%). Oxyhemoglobin flare was also observed 24 hours after infusion for each of the patients. Plasma levels of several inflammatory cytokines including G-CSF, MIP-1α, and C-reactive protein increased 24 hours after infusion for the first two subjects and subsequently decreased at 48 hours. Discussion: We have confirmed the presence of oxyhemoglobin flare on day one after infusion in this small prospective patient cohort. Increased plasma levels of inflammatory cytokines were correlated in time with the presence of oxyhemoglobin flare suggesting a possible link between optical measurements and inflammatory processes induced by chemotherapy. Additionally, we report for the first time significant changes in tumoral oxyhemoglobin and oxygen saturation during chemotherapy infusion. These changes may be of prognostic significance. DOSI allows functional measurements of tumor response at timepoints unachievable with current functional medical imaging modalities. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P2-09-15.