Abstract
249
Objectives Fluorine-18 fluorocholine (FCH) is a phospholipid synthesis tracer which shows promise for detecting several malignancies. However, data on its potential utility in breast tumors is still limited. We examined the feasibility of FCH PET/CT detection of breast cancer, characterized FCH uptake in corresponding normal breast tissue, and preliminarily compared tumor FCH uptake with histologic markers of cancer aggressiveness.
Methods FCH PET/CT data from 7 women (age 33-56) and 1 man (age 70) with biopsy-confirmed breast cancer was retrospectively reviewed. Whole-body scans were performed using 2.22 to 2.96 MBq/kg of FCH. Standardized uptake value (SUV) and mean tissue Hounsfield Units (HU) corresponding to breast tumor regions of interest were compared to corresponding measures from dense non-adipose regions of the contralateral breast. Relationships between tumor SUV and tumor size, histology, ER/PR/her 2 neu positivity, and mitotic index were also examined.
Results All breast tumors could be identified on the basis of increased FCH uptake. Average breast cancer SUVmax and tumor-to-background ratio (TBR) were 4.3 (range 2.2-7.0) and 3.4 (range 1.2-6.8) respectively. Lowest tumor uptake corresponded to the only lobular breast cancer in this series. Tumor SUVmax did not correlate with tumor size, although a moderate correlation with tumor density (HU) was observed (r=0.70, p < 0.05). Normal contralateral breast tissue demonstrated low SUVmax (range 1.07-1.97), with a trend for slightly higher SUV in denser breast tissue. Higher SUVmax was observed in HER2-positive vs. HER2-negative, but this did not yet attain statistical significance (mean SUVmax 6.1 vs. 3.6, p= 0.08).
Conclusions Although biopsy-confirmed breast cancers demonstrated a significant range of FCH uptake, detection was feasible due to relatively to low uptake in the surrounding breast tissue of both pre- and post- menopausal patients. To adequately evaluate relationships between FCH uptake and prognostic markers, further investigation using a larger patient sample is needed