18F-FDG PET of locally invasive breast cancer and association of estrogen receptor status with standardized uptake value: microarray and immunohistochemical analysis

Abstract

PET imaging is useful for evaluating locally advanced primary breast cancer. Expression of specific molecular markers in these cancers, such as estrogen receptor (ER), progesterone receptor (PR), and HER2 status, has direct prognostic and therapeutic implications in patient management. This study aimed to determine whether a relationship exists between tumor glucose use and important molecular markers in invasive breast cancer. For our purposes, tumor glucose use is quantified by the PET-derived parameter maximum standardized uptake value (SUV).

Methods: Breast tumors from 36 patients were excised and examined histologically after PET. ER, PR, and HER2 status were determined for all lesions histopathologically. In addition, genomewide expression for a subset of 20 tumors was analyzed using the human genome U133A oligonucleotide microarray.

Results: A significant association was found between estrogen ER status and lesion SUV. ER-negative tumors (n = 17; median SUV, 8.5) demonstrated a significantly higher maximum SUV than did ER-positive tumors (n = 19; median SUV, 4.0) (P < 0.001). No significant association existed between SUV and PR status, HER2/neu status, lymph node involvement, or tumor size. Unsupervised hierarchic clustering of the 20 genetically profiled cancers segregated tumor samples into 2 primary groups of 10 patients each, largely corresponding to ER status.

Conclusion: In locally invasive primary breast cancer, ER-negative tumors display higher (18)F-FDG uptake than ER-positive tumors. Microarray analysis confirms these data and identifies genes associated with increased glucose use as measured by PET. These genes significantly overlap those of a previously validated ER-status molecular phenotype. These preliminary data support a growing body of evidence that ER-positive and ER-negative breast cancers have distinct disease-specific patterns. Further validation prospectively and with larger numbers will be required to establish a robust molecular signature for metabolic uptake and patterns of aggressive behavior in advanced breast cancer.

FIGURE 1

(A) ¹⁸F-FDG accumulation correlates weakly with tumor size. (B–D) Neither HER2 status (B), node status (C), nor PR status (D) correlates with maximum SUV. (E) Accumulation does correlate with ER status. Open circles denote tumor samples subjected to microarray analysis. Closed circles do not have gene profiles.

FIGURE 2

Twenty tumors clustered into 2 groups of 10 each on basis of genomewide expression profile. (A) Complete gene list. (B) ER and ER-related genes were main grouping discriminators.

FIGURE 3

Overlap between validated ER status gene list (generated from tumor bank of 99 women) and the 500 genes whose expression most correlated with SUV from our sample of 20 tumors (P < 0.001).