Imaging a Murine Rheumatoid Arthritis Model via CD206-Targeted Fluorescent-Labeled TilmanoceptTilmanocept Binding to CD206 Using a Mouse Model

Objectives: There is an unmet clinical need for an non-invasive biomarker that can distinguish macrophage infiltration in the rheumatoid synovium. Biopsy is invasive with a low rate of patient acceptance. Tilmanocept is targeted to CD206, which is highly expressed on the surface of macrophages. The objective of this project is to determine if the CD206-targeting molecular imaging agent, fluorescent-labeled tilmanocept, can detect synovial macrophage infiltration in a murine rheumatoid arthritis (RA) model using a standard hand-held fluorescence imaging camera.

Methods: The near infra-red fluorescent dye, IRDye800CW (LI-COR Biosciences) was covalently attached to DTPA-mannosyl-dextran (tilmanocept, Navidea Biopharmaceuticals) and DTPA-galactosyl-dextran (gal-dextran, specific for hepatocytes). Both dextran conjugates had approximately 2 IRDye800CWs, 5 DTPAs, 25 mannose or galactose units per dextran, which yielded an average molecular weight of 18.5 kDa for both conjugates. Passive K/BxN serum RA model was induced injecting (IP 150 uL) of serum, This model is characterized by infiltration of inflammatory cell types including macrophages. The mice were imaged on Day 5 (the RA peak). Three experiments were performed. First, using two healthy (D^-) and two RA mice (D⁺), IRDye800CW-tilmanocept (FL-TM) was injected intra-venously using a low scaled molar dose (0.02 nmol per gram TBW). Fluorescent images (0.10 sec, Fluobeam800, Fluoptics) where obtained 15, 90, and 360 min post injection (PI). A second experiment was performed using a higher scaled molar dose of 0.1 nmol/g. At each imaging time point, regions-of-interest (ROI) were drawn around the four paws and abdomen of the healthy and RA mice. The maximum counts within each paw ROI was background-corrected fluorescence by subtraction from the abdomen ROI. A third experiment was designed to test the specificity of FL-TM for CD206. FL-TM was labeled with Tc-99m and fluorescent-gal-dextran (FL-gal-dextran) was labeled with Ga-68. Both radiopharmaceuticals (RPs) were loaded into a single syringe and co-injection intra-venously at a scaled molar dose of 0.02 nmol/g into 4 healthy and 4 RA mice. At 30 min PI, the four paws of each mice were amputated and assayed for Ga-68 (400 - 800 keV), and at 24 hr PI, assayed for Tc-99m (100 - 200 keV). Using Ga-68 and Tc-99m injection standards, the data was casted into percent-of-injection dose (%ID) for each RP. A paired Student’s-t-test was used to test for significance between %IDs of the RPs and non-paired t-test was used for differences between background-corrected count rates between healthy and RA mice.

Results: The arthritis scores for all paws of the 8 RA mice were 3 or greater (4 is max) and all paws of the healthy mice were scored at zero. Individual joints of the RA mice could be visualized in all four paws of the low- and high-dose studies. The maximum background-correct fluorescence count rates of the joint ROIs were highest at 90 mins PI, when the mean and SD of the high-dose study was 110 ± 37 kcps and the low-dose study was 32 ± 17 kcps. Count rates of healthy paws were very weak (high dose 36 ± 10 kcps; low dose 8.8 ± 3.6 kcps). At 15 and 90 min PI, there was a significant (P < 0.01) difference between D^- and D⁺ mice when a low dose was used, and a significant different at 90 mins using the high dose. The dual isotope study demonstrated CD206-specificity; P < 0.0001 for the mean difference between Tc-99m-FL-TM and Ga-68-FL-gal-dextran %IDs in the 16 paws of the 4 RA mice.

Conclusion: This murine model of RA demonstrated that fluorescent-labeled tilmanocept provides adequate contrast enhancement of inflamed synovium when imaged by a hand-held fluorescence camera. Further studies are needed to determine the correlation between fluorescence intensity and macrophage infiltration for detection of early arthritis. Research Support: local