Research Article: In vivo positron emission tomographic blood pool imaging in an immunodeficient mouse model using 18F-fluorodeoxyglucose labeled human erythrocytes

Date Published: January 25, 2019

Publisher: Public Library of Science

Author(s): Jung W. Choi, Mikalai Budzevich, Shaowei Wang, Kenneth Gage, Veronica Estrella, Robert J. Gillies, Juri G. Gelovani.


99m-Technetium-labeled (99mTc) erythrocyte imaging with planar scintigraphy is widely used for evaluating both patients with occult gastrointestinal bleeding and patients at risk for chemotherapy-induced cardiotoxicity. While a number of alternative radionuclide-based blood pool imaging agents have been proposed, none have yet to achieve widespread clinical use. Here, we present both in vitro and small animal in vivo imaging evidence that the high physiological expression of the glucose transporter GLUT1 on human erythrocytes allows uptake of the widely available radiotracer 2-deoxy-2-(18F)fluoro-D-glucose (FDG), at a rate and magnitude sufficient for clinical blood pool positron emission tomographic (PET) imaging. This imaging technique is likely to be amenable to rapid clinical translation, as it can be achieved using a simple FDG labeling protocol, requires a relatively small volume of phlebotomized blood, and can be completed within a relatively short time period. As modern PET scanners typically have much greater count detection sensitivities than that of commonly used clinical gamma scintigraphic cameras, FDG-labeled human erythrocyte PET imaging may not only have significant advantages over 99mTc-labeled erythrocyte imaging, but may have other novel blood pool imaging applications.

Partial Text

Clinical blood pool imaging is commonly performed in nuclear medicine departments using autologous human erythrocytes labeled with the radiotracer 99m-Technetium (99mTc) pertechnetate and gamma scintigraphic imaging. The current prevalent clinical indications for 99mTc-labeled erythrocyte imaging are the detection of drug-induced cardiomyopathy in cancer patients undergoing potentially cardiotoxic chemotherapy, and anatomic localization of sites of occult lower intestinal bleeding in patients [1–3]. While planar imaging of 99mTc-labeled erythrocytes is the most commonly utilized nuclear medicine blood pool imaging test, ECG-gated single-photon emission computed tomographic (SPECT) imaging of 99mTc-labeled erythrocytes has also been used for radionuclide ventriculography, demonstrating comparable results to planar blood pool imaging [4–6]. In addition, blood pool imaging with magnetic resonance imaging can be performed using long-circulating gadolinium-based magnetic resonance contrast agents such as Gadofosveset and Ferumoxytol, and have been described in the evaluation of peripheral vascular disease and other specific vascular pathologies, including deep venous thrombosis, thoracic venous outlet syndrome and pelvic congestion syndrome [7].

Unless otherwise specified, all chemicals were cell culture grade and obtained from Sigma Aldrich, St. Louis MO. Packed human erythrocytes collected in standard anticoagulant citrate dextrose (ACD) solution (either ≤ 24 hours post phlebotomy or ~ 5 day post phlebotomy) were obtained from Zen-Bio, Inc. 370–740 Megabecquerel (MBq) (1 ml) United States Pharmacopeia (USP) grade 2-deoxy-2-(18F)fluoro-D-glucose (FDG) were obtained from Cardinal Health. Vendor supplied human erythrocytes were centrifuged 1000g 10 minutes, and the remaining anticoagulant and buffy coat residual were gently aspirated. Erythrocytes were then gently washed in a 4X volume of filter sterilized “1X EDTA” solution (140 mM NaCl, 4 mM KCl, 2.5 mM Ethylenediaminetetraacetic acid dipotassium salt dihydrate (K2EDTA dihydrate)), centrifuged at 1000 x g for 10 minutes, and the wash was manually aspirated. 150 μl 1X EDTA solution was then added to the 250 μl washed erythrocytes and 100 μl (37–74 MBq) FDG to a final volume of 500 μl. For experiments using 500 μl packed erythrocytes, 400 μl 1X ETDA solution and 100 μl (37–74 MBq) FDG were added to a final volume of 1000 μl. Samples were then gently agitated on a platform rotator at either room temperature (~25° C) for 2 hours or at 37° C for 30 minutes to 2hours. Samples were then centrifuged at 1000 x g for 10 minutes and the supernatant was carefully aspirated. Erythrocytes were then washed and centrifuged twice with 4X volumes of 1X EDTA solution. For experiments characterizing residual unincorporated FDG, a 3rd wash/centrifugation step was performed. Final washed FDG-labeled erythrocytes were resuspended in 1X volume of 1X EDTA solution. Aliquots from all samples and washes were counted with the Atomlab 500 dose calibrator [Biodex Medical Systems, Inc].

We examined FDG uptake efficiency of human erythrocytes either 1 day or 5 days after phlebotomy. The percent of FDG uptake by human erythrocytes collected ≤ 24 hours prior to FDG labeling was significantly higher than that of erythrocytes collected ≈ 5 days prior to FDG labeling (Fig 1A). The mean % total FDG incorporation of 250 μl of 1 day old erythrocytes (120 minute incubation at 37° C with 37–74 MBq FDG) was 58.2% ± 0.3% (N = 5), compared to 3.4% ± 0.2% FDG incorporation of 5-day- old erythrocytes (mean ± standard error; Sample number = 6). Unpaired t test P value < 0.0001; R2 = 0.997. Currently, 99mTc-labeled compounds remain the dominant radionuclide-specific clinical blood pool imaging agents. In particular, 99mTc-labeled erythrocytes are commonly used to non-invasively search for sites of occult lower intestinal bleeding and to measure cardiac contractility (left ventricular ejection fraction) in patients at risk for chemotherapy-induced cardiotoxicity. Given the inherent advantages of PET over gamma camera imaging, development of 18-fluorine-based blood pool imaging agents for these and other clinical applications remains an area of active research. We show that human erythrocytes can rapidly incorporate sufficient amounts of FDG to obtain in vivo images of the mouse vasculature using microPET/CT. We believe this imaging technique can be readily translated to the clinical setting, given the simplicity, high labeling efficiency, and relatively short turn-around time of the labeling protocol, as well as the use of a simple incubation/wash solution composed of a few physiologic salts and a subtherapeutic dose of a clinically used compound (EDTA). As modern medical PET/CT scanners generally possess count detection sensitivities much higher than that of medical gamma camera scintigraphy, FDG-erythrocyte PET imaging may have significant advantages over 99mTc-labeled erythrocyte imaging, when used for the detection of either occult intestinal bleeding or chemotherapy-related cardiotoxicity. We speculate that this technique may have inherent advantages over the cerebrovascular perfusion SPECT imaging agents HMPAO and ECD for similar reasons. We also speculate that the relatively long radioactive half-life of 18F-FDG may allow for real-time detection of subtle changes in blood flow in the brain as well as other organs under varying conditions/stimuli that may not be discernable with clinical magnetic resonance imaging (MRI) platforms, due to the significantly higher sensitivity of PET over MRI [45].   Source:


Leave a Reply

Your email address will not be published.