PET imaging of tumor neovascularization in a transgenic mouse model with a novel 64Cu-DOTA-knottin peptide

Abstract

Due to the high mortality of lung cancer, there is a critical need to develop diagnostic procedures enabling early detection of the disease while at a curable stage. Targeted molecular imaging builds on the positive attributes of positron emission tomography/computed tomography (PET/CT) to allow for a noninvasive detection and characterization of smaller lung nodules, thus increasing the chances of positive treatment outcome. In this study, we investigate the ability to characterize lung tumors that spontaneously arise in a transgenic mouse model. The tumors are first identified with small animal CT followed by characterization with the use of small animal PET with a novel 64Cu-1,4,7,10-tetra-azacylododecane-N,N',N'',N'''-tetraacetic acid (DOTA)-knottin peptide that targets integrins upregulated during angiogenesis on the tumor associated neovasculature. The imaging results obtained with the knottin peptide are compared with standard 18F-fluorodeoxyglucose (FDG) PET small animal imaging. Lung nodules as small as 3 mm in diameter were successfully identified in the transgenic mice by small animal CT, and both 64Cu-DOTA-knottin 2.5F and FDG were able to differentiate lung nodules from the surrounding tissues. Uptake and retention of the 64Cu-DOTA-knottin 2.5F tracer in the lung tumors combined with a low background in the thorax resulted in a statistically higher tumor to background (normal lung) ratio compared with FDG (6.01±0.61 versus 4.36±0.68; P<0.05). Ex vivo biodistribution showed 64Cu-DOTA-knottin 2.5F to have a fast renal clearance combined with low nonspecific accumulation in the thorax. Collectively, these results show 64Cu-DOTA-knottin 2.5F to be a promising candidate for clinical translation for earlier detection and improved characterization of lung cancer.

Figure 1. Schematic representation of experimental design

In the lung specific conditional transgenic mouse model used in the study, expression of the reverse tetracycline transactivating protein (rtTa) is driven by the lung specific Clara cell secretory protein promoter (CCSP) and the expression of the oncogenes Kras^G12D and MYC are under control of the tetracycline responsive operon (Tet-O). Absence of doxycycline prevents rtTA from binding to Tet-O thus no expression of Kras^G12D and MYC. Presence of doxycycline triggers a conformational change of Tet-O that enables rtTa binding and expression of Kras^G12D and MYC. Expression of Kras^G12D and MYC are controlled by administering of doxycyline in the drinking water. Lung tumors develop with an average latency of 36 weeks after administration of doxycycline (20). Serial small animal CT screening was used to monitor tumor development. Mice with positive CT findings underwent combined PET/CT with ¹⁸F-FDG at day 0 and ⁶⁴Cu-DOTA-knottin 2.5F on day 1. Mice were sacrificed after imaging with ⁶⁴Cu-DOTA-knottin 2.5F (endpoint), and tumor sections were flash frozen for immunofluorescence staning.

Figure 2. Ex vivo biodistribution of ⁶⁴Cu-DOTA-knottin 2.5F

The biodistribution of knottin 2.5F in healthy mice (N=5) one-hour post-injection. Expressed as %ID/g (left) and %ID/cm³ (right), error bars indicate standard error mean.

Figure 3. Small animal PET/CT

Transverse sections of the same mouse imaged with FDG (top panel) and knottin 2.5F (lower panel). Left: CT images showing the lungs (L), the heart (H), the spine (S) and a nodule (N) immediately adjacent to the heart. The grayscale is in Hounsfield units (HU). Right: Fused PET/CT images. The nodule can easily be outlined based on the knottin 2.5F whereas the high uptake of FDG in the heart makes it difficult to delineate the nodule. Note different maximum on upper and lower PET color scale bars.

Figure 4. Tumor to background ratio of knottin 2.5F and FDG

ROIs were drawn over tumors (N=10) and areas in the lungs without visible tumor mass on CT (background). For each ROI the mean uptake in %ID/g was estimated and the tumor to background ratio calculated. Bars represent mean ± standard error mean. *, p<0.05 two sided Mann-Whitney test.

Figure 5

(A) Volume renderings of the thorax of a mouse imaged with FDG (left) and knottin 2.5F (right). A nodule is present immediately adjacent to the heart and is clearly visible when knottin 2.5F is used. The intense FDG uptake in the heart makes it impossible to delineate the nodule. Movies of the volume rendering images are available online as supplementary material. (B) Immunofluorescence of tumor sections. Sections were stained for CD31 (left) and α_v-integrin (center). Merging of the two images (right) shows that the expression of α_v-integrin co-localizes with the expression of the CD31 on the vessels (arrows).