gms | German Medical Science

Aim: Non-invasive positron emission tomography (PET) imaging using suitable biomarkers is considered as an elegant and sensitive technique predicting tumor response to chemotherapy in preclinical development. The clinically well-established PET tracers [18F]FDG and [18F]FLT are such biomarkers for metabolic tumor profiling regarding glucose metabolism and cell proliferation, respectively. In this study, [18F]FDG and [18F]FLT were tested and compared for their suitability as PET biomarkers in a preclinical setting.

Methods: Eight different solid tumor models were used: B16/BL6 metastatic melanoma grown orthotopically in syngeneic C57/BL mice, human HT29 colon adenocarcinoma tumors, NCI-H596 and -H460 lung tumors, NWT21 and RIF-1 fibrosarcomas, SKOV-3 ovarian tumors and KB31 epidermoid tumors xenografted s.c. into Harlan athymic nude mice. For PET scanning, 15-25MBq of [18F]FDG and 10-20MBq of [18F]FLT were injected in theanaesthised animal. Raw data acquisition was initiated 30min after injection of the radiotracer using the dedicated small PET tomograph quad-HIDAC. Images were reconstructed in a single 30min time frame and evaluated by visual inspection and ROI analysis. The activity concentration in the tumor ROIs was normalized to the injected dose per body weight and expressed as SUV.

Results: Tumor uptake of [18F]FDG showed a large variability ranging from 0.18 to 1.64 in all tumor models analyzed (Table 1 [Tab. 1]). Highest tracer accumulation and best tumor visualization was found for Rif-1 (Fig. 1A) and B16. Moderate [18F]FDG uptake was identified for example in SKOV-3. [18F]FLT PET scanning also revealed high inter-tumoral variability with SUVs from 0.31 to 1.67. Again, Rif-1 (Figure 1 [Fig. 1]) and NCI-H460/-596 showed highest activity concentrations allowing clear-cut visualization of the tumors with [18F]FLT. The average intra-tumoral variability within a tumor model was 22% for [18F]FDG and 26% for [18F]FLT. Furthermore, no correlation was found for [18F]FDG and [18F]FLT uptake for the eight tumor models analyzed (r2=-0.9326).

Conclusions: Both tracers allowed high-resolution visualization of certain tumor models, thus permitting their use for metabolic profiling of tumor response to drug treatment in preclinical tests. However, significant variations in tracer uptake among various tumor models were observed and careful selection of an appropriate tumor model is therefore required beforehand.