Background
Atherosclerosis is a chronic inflammatory disease of artery wall with infiltration of monocytes into subendothelial space and their differentiation into macrophages. Macrophages express pro-inflammatory and chemotactic molecules thus playing an essential role in the formation, progression, and pathogenicity of atherosclerotic plaques. The population of macrophages in atherosclerotic lesions is heterogeneous. The majority of macrophages transform into foam cells by accumulating lipids in their cytoplasm, while other macrophages may have vascular protective and repair properties. Furthermore, some macrophages may secrete enzymes that degrade the extracellular matrix and fibrous cap, possibly leading to a rupture of the atherosclerotic plaque. Significantly higher quantities of macrophage-rich plaque areas have been reported in atherosclerotic patients with acute coronary syndromes than in patients with stable angina. Since the rupture-prone plaques commonly contain high amounts of activated macrophages, imaging of the macrophages may provide a useful tool for determining plaque vulnerability [
1‐
5].
68Gallium (
68Ga,
T
1/2 68 min) is a generator-produced, positron-emitting radionuclide with a high positron yield (89%) possessing suitable physical properties for
in vivo PET imaging.
67Gallium (
67Ga) has traditionally been used to image inflammation with gamma cameras. Since
68Ga owns the same chemical characteristics as
67Ga, it is assumed that ionic
68Ga could accumulate at the sites of macrophage-rich areas in inflammatory lesions. Furthermore,
68Ga-based PET may have some added advantages compared to
67Ga-based SPECT, such as a better resolution for the detection of inflammatory plaques [
6]. The aim of this study was to determine the uptake of ionic
68Ga in the atherosclerotic plaques in mice.
Methods
Production of ionic 68Ga
68Ga radionuclide is readily available by elution from a
68Ge/
68Ga generator (Cyclotron Co., Obninsk, Russia) with 0.1 M HCl. The radioactive elution peak, monitored online with a positron-sensitive photodiode detector (Hamamatsu S5591, Hamamatsu Photonics K.K. Solid State Division, Hamatsu City, Japan), was collected and the
68Ga-chloride was neutralized with 1 M NaOH to give pH value of 7, as described earlier [
7]. The quantity of colloidal
68Ga in the final product was measured by ultrafiltration, as indicated by the activity retained on the 0.2- μm ultrafilter. The final
68Ga product contained 13% of colloidal forms of
68Ga.
Ex vivo biodistribution of 68Ga
Nine 10-11-months-old male atherosclerotic LDLR
-/-
ApoB
100/100 mice (weight 35 ± 5 g, mean ± SD) [
8,
9] and six 2.5-month-old male non-atherosclerotic C57BL/6N control mice (weight 28 ± 1 g) were
ex vivo investigated. LDLR
-/-
ApoB
100/100 mice were kept on a high-fat, Western-type diet (TD 88173, Harlan Teklad) for 3-4 months, starting at 7 months of age. C57BL/6N mice were fed with regular diet. The experiments were approved by the Laboratory Animal Care and Use Committee of the University of Turku, Finland.
68Ga (17 ± 2 MBq) was intravenously administered via a tail vein in non-anesthetized mice. Three hours after injection the mice were sacrificed under deep isoflurane anesthesia by cervical dislocation. Samples of blood and various tissues were excised, patted dry, weighed, and measured for radioactivity using an automatic gamma counter (1480 Wizard 3" Gamma Counter; EG & G Wallac, Turku, Finland) cross-calibrated with a dose calibrator (VDC-202, Veenstra Instruments, Joure, The Netherlands). The radioactivity remaining in the tail was compensated, background counts were subtracted, and the radioactivity decay was corrected to the time of injection. The concentration of the radioactivity accumulated in the various tissues over the 3-h period following 68Ga injection was expressed as a percentage of the injected activity per gram of tissue (% IA/g).
Autoradiography analysis of aortic cryosections
The distribution of 68Ga-radioactivity in the aortic tissue was studied by combining digital autoradiography and hematoxylin and eosin (HE) staining. The aorta and internal control muscle sample from the same animal were frozen and sequential longitudinal 8- and 20- μm sections were cut with a cryomicrotome at -15°C, thaw-mounted onto microscope slides. The sections were air dried for 5 min and apposed to an imaging plate (Fuji Imaging Plate BAS-TR2025, Fuji Photo Film Co., Ltd., Japan). After an exposure time of 2.5 h, the imaging plates were scanned with Fuji Analyser BAS-5000 (Fuji Tokyo, Japan; internal resolution 25 μm). The 20- μm cryosections were stained with HE and examined for morphology under a light microscope. After carefully co-registration of the autoradiographs and HE images, 68Ga-radioactivities were measured in the following regions of interest (ROIs): (1) plaque (excluding media), (2) healthy vessel wall (no lesion formation), (3) adventitia (including adjacent fat), and (4) muscle (internal control tissue). In control mice, no plaque areas were seen and ROIs were defined for healthy vessel wall, adventitia, and muscle. In each mouse, the uptakes of 68Ga-radioactivities in the regions of interest were calculated and normalized against internal control tissue (muscle). The results of digital autoradiography analysis were expressed as photostimulated luminescence per square millimeter (PSL/mm2).
Assessment of degree of inflammation in atherosclerotic plaques
Immunohistochemical staining with the mouse macrophage specific antibody (Mac-3, clone m3/84, BD Pharmingen, BD550292, USA - BD Biosciences Pharmingen, Franklin Lakes, NJ, USA) was performed using aortic longitudinal 8- μm cryosections in order to detect activated macrophages, as described earlier [
10]. Based on the Mac-3 staining, plaque areas were divided into four categories for the semiquantitative estimation of the degree of plaque inflammation; (1) no inflammation = no macrophages, (2) mild inflammation = occasional macrophages, (3) moderate inflammation = occasional and some groups of macrophages, (4) severe inflammation = abundant infiltration of macrophages. The uptake of
68Ga-radioactivity in the different categories was studied with digital autoradiography from 8- μm cryosections combined with images of Mac-3 immunostained sections. The results of digital autoradiography analysis were normalized against internal control tissue (muscle) and expressed as PSL/mm
2.
In vivo imaging
Three additional female LDLR
-/-
ApoB100/100 mice, aged 14 to 15 months and kept on a high-fat, Western-type diet for 5 to 6 months (weight 36 ± 10 g, mean ± SD) were in vivo imaged with the Inveon small animal PET/CT scanner (Siemens Medical Solutions, Knoxville, TN, USA) at 3 h after injection of 68Ga (16 ± 3 MBq). PET images were acquired for 15 min, followed by CT angiography (10 min) without moving the animal. To obtain vascular contrast, 0.2 ml of iodinated intravascular contrast agent eXIATM160XL (Binitio Biomedical Inc, Ottawa, ON, Canada) was injected. CT acquisition consisted of 270 projections acquired with exposure time of 400 ms, x-ray voltage of 80 kVp, and anode current of 400 μA for a full 360° rotation. Mice were kept fully sedated with 1.5% isoflurane during imaging. PET image was reconstructed using OSEM3D algorithm. The resulting matrix was 128 × 128 pixels with 159 transverse slices in a transaxial field of view of 12.7 cm (pixel size 0.87 × 0.87 × 0.80 mm). CT image was reconstructed using a modified Feldkamp algorithm. The resulting matrix was 256 × 256 pixels with 384 transverse slices (pixel size 0.17 × 0.17 × 0.17 mm). Co-registration of PET and CT images was done using an automatic weighted mutual information algorithm and confirmed visually on the basis of anatomical landmarks. Quantitative PET analysis was performed by drawing ROIs of the same size in the left ventricle (blood pool), aortic arch, and brachiocephalic artery as identified on the basis of the CT angiography by using the Inveon Research Workplace software (Siemens Medical Solutions, Knoxville, TN, USA). Radioactivity concentrations were corrected for injected radioactivity, and the results were expressed as percentage of injected radioactivity per gram of tissue (% IA/g) assuming 1 mL equals 1 g.
Statistical methods
All the results are expressed as mean ± SD. Student's t- test and one-way ANOVA were applied when comparing the biodistribution results of the atherosclerotic and control mice. T test with a mixed Dunnett's model was used to correct the p values of biodistribution of aortic uptake versus different tissue uptakes in the atherosclerotic and control mice groups. In the autoradiographic analysis, a mixed model with Tukey-Kramer corrected p values was applied to individual mean values of ROIs. A p value less than 0.05 was considered as statistically significant.
Discussion
Macrophages play a crucial role in the progression and development of atherosclerosis from early fatty-streak lesions to advanced atherosclerotic plaques [
11]. Ruptured plaques, in particular, commonly contain high quantities of macrophages [
3]. Therefore, the contents of macrophages may be an excellent imaging biomarker for the vulnerability of atherosclerotic plaques. The present results revealed that the
68Ga-radioactivity uptake in atherosclerotic plaques, especially in their macrophage-rich areas, was higher when compared to healthy vessel wall. This observation strengthens our conception that
68Ga-radioactivity accumulates at the site of inflammation.
68Ga radionuclide is readily available by elution from a
68Ge/
68Ga generator with 0.1 M HCl in the form of
68Ga-chloride. However, after neutralization,
68Ga-chloride is hydrolyzed
in situ or after administration
in vivo. In general, ionic
68Ga
3+ is in the form of soluble anion called gallate,
68Ga(OH)
-
4, and/or insoluble colloidal particles
68Ga(OH)
3. In the present study, the final
68Ga product contained 13% of colloidal forms of
68Ga. Uncharged gallium hydroxide or gallium nanoparticles are capable of penetrating the cell. Under the lower pH conditions, inside the cell, the material ionizes and becomes bound. Accordingly, accumulation into the liver, spleen, and bone marrow would be expected [
12,
13].
In blood,
68Ga-radioactivity can circulate also as a free
68Ga
3+ and mimic ferric ions. In spite of their physical similarities, gallium and iron behave differently
in vivo. The major difference is the inability of gallium to be reduced
in vivo wherefore gallium remains bound to iron-transport proteins and carrier molecules with the inability to interact with protoporphyrin IX to form heme. A high blood uptake of gallium is explained by its binding to iron-binding molecules; transferrin, lactoferrin, and ferritin. One way to facilitate the solubilization of
68Ga may be its chelation with citrate. For SPECT imaging,
67Ga is indeed administrated as citrate to avoid protein binding. Since citrate is only a weak chelator
in vivo,
67Ga is rapidly released, hydrolyzed, and bound to transferrin and other iron-binding proteins [
12,
14‐
16], suggesting that chelation does not improve the uptake of the tracer. We have compared uptakes of
68Ga and
68Ga-citrate in rat osteomyelitis model (unpublished data), and ascertained that there were no differences between
68Ga and
68Ga-citrate uptakes in inflammation lesions.
In comparison to
67Ga-based SPECT,
68Ga-based PET may have some advantages, such as a better resolution for the detection of inflammation and a smaller radiation exposure for the subject. However, the blood clearance of
68Ga-radioactivity was slow as demonstrated by
ex vivo measurement and by
in vivo PET/CT imaging. Unfortunately, the relatively short half-life of
68Ga does not allow delayed imaging times for better blood clearance. In an inflamed lesion,
67Ga is accumulated in the areas rich in infiltrated macrophages. Interestingly,
67Ga is not accumulated at the sites of crowded neutrophils, suggesting that the polymorphonuclear leukocytes do not play a major role in the accumulation of
67Ga in inflammatory lesions [
17].
In the present study, the plaque/wall ratio of
68Ga was 1.8 ± 0.2. As compared to 2-deoxy-2-[
18F]fluoro-
D-glucose (
18F-FDG), we have measured the plaque/wall ratio of 2.7 ± 1.1 approximately in the same aged LDLR
-/-
ApoB
100/100 mice as used in the present study [
18]. It has been shown that the clinical image qualities of
68Ga-tracers (
β + decay 97%,
Eβ + max 0.64 MeV) and
18F-tracers (
β + decay 97%,
Eβ + max 0.64 MeV) are not necessarily very much different from each other [
19]. However,
68Ga has a higher kinetic energy compared to
18F, which could decrease the imaging spatial resolution, particularly when a small plaque lesion is imaged with a high resolution small PET/CT. Based on these earlier observations,
18F-FDG seems to be more promising PET tracer for imaging of atherosclerotic plaques than
68Ga.
Clinical observations indicate that in normal circumstances, the major proportion of the gallium retained in the body is localized in the bone and liver, and that inflammatory lesions alter this biodistribution, frequently causing decreased hepatic uptake [
15]. In the present study, a difference in the uptake of
68Ga between the atherosclerotic mice and controls was found in bone and liver. The uptake of
68Ga-radioactivity in the whole aorta of atherosclerotic mice was higher than that in the control mice, but the difference was not significant. The uptake in the inflammatory plaques could partly explain the lower uptake of
68Ga in the liver of atherosclerotic mice. More likely this may be due to liver failure caused by high fat content, which can occur in older LDLR
-/-
ApoB
100/100 mice fed with a Western-type diet. Concerning the uptake of
68Ga in the bone, we assume that in the young control mice (aged 2.5 months), active bone remodeling may have contributed to the higher bone uptake as compared to the older LDLR
-/-
ApoB
100/100 mice (aged 10 to 11 months).
The accumulation of
68Ga-radioactivity in inflammatory lesions is complex, and the majority of
68Ga is probably delivered into the inflammatory lesions through the increased permeability at the site of atherosclerotic arteries. In the present study, accumulated
68Ga-radioactivity was detected in the macrophages-rich plaques suggesting that
68Ga accumulates at the site of inflammation. In addition to that, calcified areas of atherosclerotic plaques were also detected by
68Ga, which may be explained by the competitive binding of gallium ions to the Ca
2+ and Mg
2+ binding sites [
17]. Since the calcified areas can be localized with CT, this observation is not limiting the use of
68Ga to distinguish macrophages-rich plaques and calcified plaques.
To some extent, gallium may also be taken up by atherosclerotic plaques due to its binding to the circulating transferrin and avidly to the transferrin receptors at the site of atherosclerotic arteries. Transferrin receptor expression has been shown to correlate positively with macrophage infiltration in human carotid atherosclerotic lesions [
20]. Moreover, symptomatic patients with carotid atherosclerosis have significantly higher expression of transferrin receptors than asymptomatic patients [
21]. However, although the uptake of
68Ga in inflamed atherosclerotic plaques was promising as measured by autoradiography, the slow blood clearance limits the use of ionic
68Ga for
in vivo imaging of atherosclerotic plaques.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
JMUS contributed the conception and design, performed the experiments, analyzed and interpreted data and drafted the manuscript. IL performed the experiments, acquired, analyzed and interpreted data and participated in writing of the manuscript. HJS prepared the 68gallium and participated in writing of the manuscript. VJOL involved in designing the analysis and interpretation of data, and in drafting and revising the manuscript. PL involved in designing the experiments and revising the manuscript. SY-H, JK, and AR contributed the conception and design, critically contributed and revised the manuscript and enhanced its intellectual content. All authors have approved the final content of the manuscript.