Abstract
1376
Objectives Dysregulated cellular metabolism represents a hallmark of malignancy. Increased glycolytic rates of malignancy enable detection with the glucose analog FDG, which has found widespread clinical use. Glutamine, the most abundant amino acid in blood, may also be utilized as a nutrient source and has been implicated in the cancer phenotype. Myc-dependant glutamine addiction has been observed in several malignancies and presents a target for therapy. Inhibitors to glutaminase, the enzyme responsible for the conversion of glutamine to glutamate, represent a targeted therapeutic agent and are now being tested in early phase clinical trials. [18F](2S,4R)4-Fluoroglutamine ([18F]Fluoroglutamine), an analog of glutamine for PET imaging, has been tested in pre-clinical studies and recently studied in glioma patients. In this study, we describe preliminary results of kinetic analysis of of [18F]Fluoroglutamine in a mouse model of triple-negative breast cancer (TNBC).
Methods TNBC cells (HCC1806) exhibiting high glutaminase activity were subcutaneously inoculated in the flank of athymic nu/nu mice. After tumors reached a pre-set volume, imaging was performed on a dedicated small animal PET scanner. [18F]Fluoroglutamine (300-350 uCi) was injected into the lateral tail vein at the start of dynamic image acquisition. Images were obtained for 60 continuous minutes at 5 minutes/frame. A glutaminase inhibitor was then administered for seven days. A control mouse received a saline-based vehicle solution. Mice were then reimaged with the same imaging protocol. A mouse inoculated with receptor-positive breast cancer cells (MCF-7 cell line; low glutaminase activity with high intracellular glutamine concentration, estimated independently) was also imaged. Image analysis was performed using AMIDE data analysis software. Kinetic analysis was performed with PMOD General Kinetic Modeling Tool (PKIN). Our hypothesis, based upon cellular data, was that tumors with high glutaminolysis, which have a lower intracellular glutamine pool size, would be associated with lower [18F]Fluoroglutamine distribution volumes. Additionally, glutaminase inhibition would increase the size of the glutamine pool, reflected as an increased distribution volume for [18F]Fluoroglutamine.
Results Single-compartment models slightly underestimated glutamine counts at later time points, consistent with a small amount of label trapping, but the estimated k3 in a two-compartment model with irreversible trapping was low (best estimate of ~0.006/min). Logan plot analysis revealed linearity from which distribution volumes could be estimated. Preliminary Logan plot analysis of the two TNBC tumor-bearing mice demonstrated increased volume of distribution of glutamine post-glutaminase inhibition with individual estimates of greater than 20% and greater than 75%. No increase in volume of distribution was seen in the vehicle-treated mouse. The untreated MCF-7 tumor-bearing mouse demonstrated a particularly conspicuous tumor with a volume of distribution at least 50% larger than the TNBC untreated tumors. This result is consistent with the known reduced rate of glutaminolysis in MCF-7 compared to TNBC cell lines.
Conclusions This preliminary study demonstrated that kinetic analysis identifies increased pool size of glutamine post glutaminase inhibition in a mouse tumor model with elevated glutamines activity, providing a promising tool for measuring the impact of glutaminase-directed drugs. The results also underscore the utility of dynamic imaging and kinetic modeling in oncologic analysis with this radiotracer.