Abstract Background Kinetic analysis of FDG PET and DCE-MRI can identify patterns of breast tumor metabolism and perfusion that predict pathologic response, relapse, and survival in patients (pts) receiving neoadjuvant chemotherapy (NC). We are enrolling pts with LABC or IBC on a phase II trial of neoadjuvant sunitinib and metronomic chemotherapy. The addition of sunitinib, a tyrosine kinase inhibitor of VEGFR1-3, PDGFR, c-KIT, to NC is hypothesized to increase rate of pathologic complete response (pCR). Assessment of FDG PET measures of glucose metabolism (Ki), glucose delivery (K1) which approximates blood flow, and MRI measures of blood flow and vascularity (peak enhancement (PE), signal enhancement ratio (SER), and volume) during NC offers the opportunity to evaluate the in vivo pharmacodynamics of sunitinib. Methods: Pts with HER2−negative LABC or IBC participated in a companion imaging trial with [18F]-FDG PET and DCE-MRI before NC (T0), after a 1 wk run-in of sunitinib 25 mg po daily (T1), after 12 wks of paclitaxel 80 mg/m2 IV Qwk and sunitinib 25 mg po daily (T2), and prior to breast surgery (T3) after 15 wks of doxorubicin 24 mg/m2 IV Qwk, cyclophosphamide 60 mg/m2 po daily with G-CSF 5 mcg/kg SC days 2–6 each wk. FDG metabolic rate (Ki), glucose delivery (K1), and MR indices (PE, SER, volume) were assessed. Imaging parameters were compared for groups defined by NC pathologic complete response (pCR) vs. non-pCR using a two-sample t-test. Results: The imaging trial included 14 pts. Median age was 50 years (43-79). All had HER2−negative LABC (n=13, 93%) or IBC (n=1, 7%). Most tumors were ductal (n=12, 86%) and high grade (n=9, 64%). Seven (50%) tumors were ER negative. pCR was observed in 4/14 (29%) pts in this cohort. Changes in Ki, K1, and MRI volume were observed between baseline (T0) and the sunitinib run-in (T1). For example, 8/14 (57%) had a decrease in K1 of >20%, and 3 (21%) had an increase of ≥20%. These 1 week changes did not predict subsequent response to NC. However, declines in Ki and K1 between baseline (T0) and following sunitinib and paclitaxel (T2) did predict pCR. The average change in glucose metabolism (Ki) was a 95% decline with pCR and a 68% decline otherwise (p= 0.007). The average T0-T2 K1 change was 83% decline for pts with pCR and 47% decline otherwise (p= 0.029). In contrast to our previous studies in LABC pts treated with NC where decline in K1 was predictive of response, decline in Ki appears to be the more robust predictor of response in this cohort. Of 11 pts with PET scans at T2 and T3, 5 showed marked increase (>20%) in Ki and 6 showed marked increase in K1 after withdrawal of sunitinib. Conclusion: Changes in breast tumor glucose metabolism (Ki), glucose delivery (K1), and blood flow (MR PE, SER, volume) can be detected after 1 wk of sunitinib, but are not predictive of response to NC. In the setting of anti-vascular therapy, measures of tumor glucose metabolism (Ki) are predictive and perhaps, more predictive of outcome than measures of glucose delivery (K1) which may be altered by sunitinib. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P2-09-09.