The aim of this study was to characterize the biologic response of locally advanced breast cancer (LABC) to chemotherapy using (15)O-water-derived blood flow measurements and (18)F-FDG-derived glucose metabolism rate parameters.
Methods: Thirty-five LABC patients underwent PET with (15)O-water and (18)F-FDG before neoadjuvant chemotherapy and 2 mo after the initiation of treatment. Kinetic analysis for (15)O-water was performed using a single tissue compartment model to calculate blood flow; a 2-tissue compartment model was used to estimate (18)F-FDG rate parameters K(1), k(2), k(3), and the flux constant, K(i). Correlations and ratios between blood flow and (18)F-FDG rate parameters were calculated and compared with pathologic tumor response.
Results: Although blood flow and (18)F-FDG transport (K(1)) were correlated before chemotherapy, there was relatively poor correlation between blood flow and the phosphorylation constant (k(3)) or the overall (18)F-FDG flux (K(i)). Blood flow and (18)F-FDG flux were more closely matched after chemotherapy, with changes in k(3) accounting for the increased correlation. These findings were consistent with a decline in both the K(i)/flow and k(3)/flow ratios with therapy. The ratio of (18)F-FDG flux to transport (K(i)/K(1)) after 2 mo of chemotherapy was predictive of ultimate response.
Conclusion: The pattern of tumor glucose metabolism in LABC, as reflected by analysis of (18)F-FDG rate parameters, changes after therapy, even in patients with modest clinical responses. This may indicate a change in tumor "metabolic phenotype" in response to treatment. A low ratio of glucose metabolism (reflected by K(i)) to glucose delivery (reflected by K(1) and blood flow) after therapy is associated with a favorable response. Further work is needed to understand the tumor biology underlying these findings.