Comparison of 68Ga-DOTA-Siglec-9 and 18F-Fluorodeoxyribose-Siglec-9: Inflammation Imaging and Radiation Dosimetry

Sialic acid-binding immunoglobulin-like lectin 9 (Siglec-9) is a ligand of inflammation-inducible vascular adhesion protein-1 (VAP-1). We compared 68Ga-DOTA- and 18F-fluorodeoxyribose- (FDR-) labeled Siglec-9 motif peptides for PET imaging of inflammation. Methods. Firstly, we examined 68Ga-DOTA-Siglec-9 and 18F-FDR-Siglec-9 in rats with skin/muscle inflammation. We then studied 18F-FDR-Siglec-9 for the detection of inflamed atherosclerotic plaques in mice and compared it with previous 68Ga-DOTA-Siglec-9 results. Lastly, we estimated human radiation dosimetry from the rat data. Results. In rats, 68Ga-DOTA-Siglec-9 (SUV, 0.88 ± 0.087) and 18F-FDR-Siglec-9 (SUV, 0.77 ± 0.22) showed comparable (P = 0.29) imaging of inflammation. In atherosclerotic mice, 18F-FDR-Siglec-9 detected inflamed plaques with a target-to-background ratio (1.6 ± 0.078) similar to previously tested 68Ga-DOTA-Siglec-9 (P = 0.35). Human effective dose estimates for 68Ga-DOTA-Siglec-9 and 18F-FDR-Siglec-9 were 0.024 and 0.022 mSv/MBq, respectively. Conclusion. Both tracers are suitable for PET imaging of inflammation. The easier production and lower cost of 68Ga-DOTA-Siglec-9 present advantages over 18F-FDR-Siglec-9, indicating it as a primary choice for clinical studies.


Introduction
Inflammation plays role in several diseases, such as, rheumatoid arthritis, diabetes and atherosclerosis. The early detection of inflammatory foci is critical for the adequate treatment of patients, and quantitative PET imaging may provide a valuable tool for diagnosis and monitoring of the effects of therapeutic interventions. 18 F-FDG is the gold standard for PET, but not specific to inflammation. In addition, the high physiological accumulation of 18 F-FDG in heart and brain makes it difficult to detect inflammatory foci close to these organs [1].
Vascular adhesion protein-1 (VAP-1) is an endothelial adhesion molecule, which is involved in leukocyte transendothelial migration from blood into the sites of inflammation. During inflammation VAP-1 translocates from intracellular storages on the endothelial cell surface where it contributes leukocyte-endothelial adhesion. Although VAP-1 plays important role in early phases of inflammation, its luminal expression on the endothelium will remain constant 2 Contrast Media & Molecular Imaging  16 19 13 Ex vivo autoradiography (no.) ND 12 10 LDLR −/− ApoB 100/100 = low-density lipoprotein receptor-deficient mouse expressing only apolipoprotein B100; ND = not done; no. = number of investigated animals.

Animal Models.
Twenty-four hours before the PET studies, Sprague-Dawley rats (weight, 350 ± 22 g; = 16) were subcutaneously injected with turpentine oil (Sigma-Aldrich) to induce focal acute, sterile inflammation [12]. Before the injection, rats were shaved on the both forelegs. Inflamed area on the left foreleg contained both skin and muscle. The intact, contralateral side (right foreleg) was used as a control.
All animal experiments (Table 1) were approved by the national Animal Experiment Board in Finland and carried out in compliance with the EU directive.

Rat Studies.
Rats were divided into two groups with Group 1 being intravenously (i.v.) given 68 Ga-DOTA-Siglec-9 (16 ± 2.9 MBq, = 8) and Group 2 18 F-FDR-Siglec-9 (18 ± 5.1 MBq, = 8). A 60 min dynamic PET acquisition was performed on a High Resolution Research Tomograph (HRRT; Siemens Medical Solutions, Knoxville, TN, USA). The PET data were reconstructed into 5 × 60 s and 11 × 300 s frames using an ordered-subsets expectation maximization 3D algorithm (OSEM3D). Quantitative PET image analysis was performed by defining regions of interest (ROIs) within the inflamed area (on the left foreleg), control area (on the right foreleg), kidneys, lungs, heart, liver, and urinary bladder using Carimas 2.8 software (Turku PET Centre). Results were expressed as standardized uptake values (SUV) and time-activity curves. SUV was calculated as a ratio of tissue radioactivity concentration (Bq/mL) and given radioactivity dose (Bq) divided by animal's body weight.
After PET imaging, rats were sacrificed and various tissues were excised and weighed, and their radioactivity levels were measured with a gamma counter (1480 Wizard 3 , PerkinElmer, Turku, Finland). The ex vivo biodistribution results were expressed as a percentage of the injected radioactivity dose per gram of tissue (% ID/g) and target-tobackground ratio.
The inflamed area and control area tissue samples were frozen, cut into sections, and stained with hematoxylin-eosin (H&E) for morphological evaluation.
Absorbed doses of 68 Ga-DOTA-Siglec-9 and 18 F-FDR-Siglec-9 were calculated with the OLINDA/EXM version 1.0 software (organ level internal dose assessment and exponential modeling; Vanderbilt University, Nashville, TN, USA), which applies the MIRD schema (developed by the Medical Internal Radiation Dose committee of the Society of Nuclear Medicine) for radiation dose calculations in internal exposure. The software includes radionuclide information and selection of human body phantoms. The residence times derived from the rat data were integrated as the area under the time-activity curve. The residence times were converted into corresponding human values by multiplication with a factor to scale the organ and body weights: (W Body,rat /W Organ,rat ) × (W Organ,human /W Body,human ), where W Body,rat and W Body,human are the body weights of rat and human (a 70-kg male), respectively; and Organ,rat and Organ,human are the organ weights of rat and human (organ weights for a 70 kg male), respectively [14].

Mouse Studies.
To detect luminal expression of VAP-1, mice were intravenously (i.v.) injected with a monoclonal rat anti-mouse VAP-1 antibody (7-88, 1 mg/kg diluted in saline) [15] 10 min before sacrifice. Aorta samples were frozen and cut into 8 m longitudinal sections, incubated for 30 min at room temperature in the dark with a secondary goat anti-rat antibody (working dilution, 5 g/mL in phosphate-buffered saline (PBS) containing 5% normal mouse or human AB serum), conjugated to a fluorescent dye (Alexa Fluor 488; Invitrogen, Eugene, OR, USA), and rinsed twice in PBS for 5 min.
In PET studies, mice (19 atherosclerotic, 13 controls) were injected with 14 ± 4.4 MBq of 18 F-FDR-Siglec-9. Twenty-five minutes after 18 F-FDR-Siglec-9 injection, blood was drawn by cardiac puncture and the animals were killed. The thoracic aorta was excised and rinsed in saline to remove the blood. In addition, various other tissues were excised and patted dry. Samples of blood and urine were collected, and blood plasma was separated by centrifugation. All tissue samples were weighed, and their radioactivity levels were measured with a gamma counter (Triathler 3 , Hidex Oy, Turku, Finland). The results were expressed as % ID/g and target-to-background ratio.
Autoradiography was used to study the distribution of radioactivity in the aorta in more detail, as described previously [13]. After careful superimposition of the autoradiographs and H&E stained images, the count densities of 540 ROIs (185: plaques, 241: normal vessel walls, and 114: adventitia) were analyzed using Tina 2.1 software. The autoradiography results were calculated as the photostimulated luminescence per unit area (PSL/mm 2 ) normalized for injected radioactivity dose, and as ratios between the atherosclerotic plaque, normal vessel wall, and adventitia.
Quantitative PET image analysis was performed by defining ROIs in the heart left ventricle (blood pool) and aortic arch as identified on the basis of the CT angiography by using the Inveon Research Workplace software (Siemens Medical Solutions, Knoxville, TN, USA). Time frames 10-20 min after injection were used for PET quantification, as previously reported in the same mouse model using 68 Ga-DOTA-Siglec-9 [4]. The results within ROIs were expressed as SUV and target-to-background ratio (SUV max,aortic arch /SUV mean,blood ).

Statistical Analyses.
All results are expressed as mean ± SD. Paired 2-tailed Student's -tests were applied for intra-animal comparisons. Nonpaired data were compared between two groups using -tests and between multiple groups using ANOVA with Tukey's correction. A value less than 0.05 was considered statistically significant.

Discussion
VAP-1 targeted ligands are promising tools for PET imaging of inflammation. In this study, we compared the VAP-1 targeting tracers 68 Ga-DOTA-Siglec-9 and 18 F-FDR-Siglec-9 in the detection of experimental inflammation in rats and mice and also estimated the radiation burden to humans. We found that the uptake of both tracers was higher in skin/muscle inflammation than in healthy muscle, and in atherosclerotic rather than in nonatherosclerotic arterial walls. Both tracers resulted in a low radiation exposure, but the lower-cost and more straightforward radiolabeling procedures support the potential use of 68 Ga-DOTA-Siglec-9 for PET imaging of patients with inflammation. The tested tracers have a similar amino acid sequence but a differently conjugated peptide structure ( 68 Ga-DOTA versus 18 F-FDR). We hypothesized that the 18 F-labeled tracer would provide improved visualization of inflammatory foci because it has a lower positron range (0.27 mm) than 68 Ga (1.05 mm). For PET imaging, 18 F ( 1/2 = 110 min, + max = 640 keV, + = 97%) is an ideal radionuclide, providing a high spatial resolution in the resulting images. 68 Ga ( 1/2 = 68 min, + max = 1899 keV, + = 89%) is a positron-emitting  radiometal that is particularly suitable for the labeling of chelate-conjugated peptides. While production of 18 F requires a cyclotron, 68 Ga is produced with an easily accessible low-cost 68 Ge/ 68 Ga-generator [17]. Although the inflammation detection characteristics of 68 Ga-DOTA-Siglec-9 and 18 F-FDR-Siglec-9 were similar, the uptake of 18 F-FDR-Siglec-9 was higher in several nontarget tissues, including the control area. We do not have a clear explanation for the distinctive distribution patterns, particularly in the liver, pancreas, heart, and kidneys, but suspect that they were at least partly due to the sugar moiety. Similar results have been observed with 68 Ga-DOTANOC and 18 F-FDR-NOC [18]. 68 Ga-DOTA-Siglec-9 and 18 F-FDR-Siglec-9 showed comparable in vivo imaging of inflammation in the rat model. The difference in the control area between the two tracers might at least partly be explained by the higher blood pool radioactivity of 18 F-FDR-Siglec-9. Although both 68 Ga-DOTA-Siglec-9 and 18 F-FDR-Siglec-9 clearly delineated inflamed area by in vivo PET, the 18 F-FDG uptake was higher (SUV mean 2.0 ± 0.52 at 90 min after injection) as reported in our previous rat studies with turpentine-induced inflammation [16].
In general, in vivo PET imaging of such a small target as atherosclerotic lesion in mice aorta is very challenging. When size of the imaged structures is smaller than the spatial resolution of the scanner, spillover from adjacent tissues and partial volume effect may invalidate the quantification of PET data in addition to cardiac and respiratory movement artifacts. Although small-animal PET/CT image of atherosclerotic mouse showed hot spot in the lesion-rich aortic arch (Figure 4(a)), the ex vivo biodistribution showed that uptake in the whole thoracic aorta was much lower than the blood level (Figure 4(b)). Therefore, it is possible that the PET/CT imaging of atherosclerotic lesions is interfered with blood pool radioactivity. On the contrary, in the rat model, the size and location of focal skin/muscle inflammation as well as the blood radioactivity concentration were much more favorable for reliable PET imaging of inflamed area. The PET scanning protocols and quantification methods used in this study were based on our previous research to allow direct comparison of new and already existing results.

Conclusion
VAP-1 targeted 68 Ga-DOTA-Siglec-9 and 18 F-FDR-Siglec-9 peptides are potential tracers for the PET imaging of inflammation. The human radiation dose estimates indicate a low radiation exposure with either of the investigated tracers. The present study further strengthens the concept of a VAP-1-based imaging strategy for the in vivo detection of inflammation by PET.

Conflicts of Interest
Sirpa Jalkanen owns stocks in Faron Pharmaceuticals Ltd. The remaining authors have no conflicts of interest to disclose.