Metabolism-perfusion mismatch as assessed by PET varies with breast cancer phenotype and predicts response to neoadjuvant chemotherapy.

Abstract

Abstract Abstract #6005 Background: Kinetic analysis of FDG and water PET can identify patterns of breast cancer metabolism and perfusion in patients receiving neoadjuvant chemotherapy (NC). Previously, we found that high pre-therapy glucose tumor metabolism relative to perfusion was associated with poor tumor pathologic response, early relapse, and death in patients with locally advanced breast cancer (LABC) treated with NC. This analysis examines tumor metabolism and perfusion as a function of tumor phenotype. Material and Methods: Tumor phenotype, defined by immunohistochemistry (IHC), was determined in 51 patients undergoing NC between 1995 and 2005. Luminal tumors were defined as those expressing either estrogen receptor (ER) or progesterone receptor (PR). The triple-negative (TN) phenotype was defined as ER and PR negative without HER2 overexpression by IHC or amplification by FISH. HER2 phenotype showed HER2 overexpression or amplification but were ER/PR negative. Women with LABC underwent dynamic [18F]-FDG and [15O]-water PET scans prior to NC. The FDG metabolic rate (MRFDG) and transport (FDG K1) parameters were calculated; blood flow (BF) was estimated from the water PET scan. Response to NC was determined from surgical specimens with pathologic complete response (pCR) defined as eradication of invasive tumor in the breast vs. other. Results: Of the tumors studied, 16 (31%) were TN, 30 (59%) were luminal, and 5 (10%) were HER2. pCR was observed in 4/16 (25%) TN tumors (95% CI: 0.10-0.50) compared to only 4/30 (13%) of luminal tumors (95% CI: 0.05-0.30) and 3/5 HER2 tumors. Linear regression of the association between PET parameters and phenotype (TN vs. luminal) found that TN was associated with higher MRFDG (p=0.007) and MRFDG/BF ratio (p=0.02), but not with BF (p=0.27). Only patients with low pre-therapy MRFDG/BF ratio (<35 umol/mL) achieved a pCR. Using this value as an ad-hoc cutoff, 4/7 (57%) of TN patients with low ratios had a pCR, while 0/9 with higher ratios had pCR. In contrast, only 4/21 (19%) of the luminal patients with low ratios had a pCR (one-sided mid-p=0.04 for TN vs. luminal pCR rate for patients with MRFDG/BF < 35 umol/mL). Discussion: These results demonstrate heterogeneity in breast tumor metabolism and perfusion as assessed by PET, and suggest a clinically relevant association between PET parameters and tumor phenotypes. The high MRFDG/BF ratio that predicts poor response to NC is more common in TN tumors; whereas in luminal or HER2 tumors, high MRFDG is generally matched with higher BF. Measurement of tumor metabolism and perfusion may identify a subset of tumors which are unlikely to respond to NC. Identification of such tumors may direct therapy toward those biologic targets most likely to overcome resistance. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 6005.