Detection of Apoptotic Cells in a Rabbit Model with Atherosclerosis-Like Lesions Using the Positron Emission Tomography Radiotracer [ 18 F]ML-10

Abstract

[¹⁸F]ML-10 (2-(5-fluoro-pentyl)-2-methylmalonic acid) is a positron emission tomography (PET) radiotracer that accumulates in cells presenting apoptosis-specific membrane alterations. The aim of this study was to test whether [¹⁸F]ML-10 allows for the detection of apoptotic cells located in atherosclerotic plaques in rabbits. Atherosclerotic plaques were induced in the aortas of five rabbits, and five additional rabbits were used as controls. Activity in the aortas was quantified in vivo and ex vivo. The localization of [¹⁸F]ML-10 to the aortic wall was identified by autoradiography. Average target to background ratios measured in vivo by PET were higher in the aortas of atherosclerotic rabbits compared with those of control rabbits (2.00 ± 0.52 vs 1.22 ± 0.30; p < .05). Differences in [¹⁸F]ML-10 uptake between atherosclerotic and control aortas were confirmed ex vivo by PET and gamma counting (23.9 ± 11.2 vs 1.1 ± 2.4 counts/pixel; p <.05; 3.6 ± 2.0 vs 0.05 ± 0.05 % of injected activity/g; p < .05, respectively). Strong correlation was observed between the accumulation of [¹⁸F]ML-10 in aortic segments as detected by autoradiography and the number of apoptotic cells on corresponding histologic sections (r² = .75; p < .05). In this study, we found that atherosclerotic plaques rich in apoptotic cells can be detected with [¹⁸F]ML-10 and PET.

DURING THE EARLY STAGES of the apoptotic process, the cell membrane undergoes specific changes collectively designated as the apoptotic membrane imprint, including acidification of the external membrane leaflet due to the exposure of phosphatidylserine, permanent membrane depolarization, irreversible loss of intracellular pH control, and activation of the phospholipid scrambling mechanism. Throughout this process, the membrane integrity is preserved. A set of novel probes with an amphipathic structure, that is, containing both specific hydrophobic and charged moieties, has been designed to detect cells presenting with an apoptotic membrane imprint. The hydrophobic moiety of these probes provides a membrane anchor, whereas the charged moiety prohibits the crossing of the probe through the highly hydrophobic membrane core of nonapoptotic cells. In apoptotic cells, activation of the scrambling process in membranes substantially reduces this energetic barrier and allows these probes to cross from the outer to the inner membrane leaflet and to accumulate within the cytoplasm. ML-10 (2-(5-fluoro-pentyl)-2-methylmalonic acid), a low-molecular-weight (206 Da) molecule belonging to this family of amphiphatic probes, has been radiolabeled with [¹⁸F]fluoride for positron emission tomography (PET) imaging. ¹ In cell culture, accumulation of tritiated ML-10 was up to 10-fold higher in apoptotic cells than in nonapoptotic cells and was associated with the binding of annexin V to cell membranes. In addition, no significant accumulation of tritiated ML-10 was detected in necrotic cells. Hence, [¹⁸F]ML-10 represents a promising radiotracer for the in vivo detection of apoptotic cells by PET.

In ischemic stroke patients, plaque rupture is detected in approximately 60% of those presenting with symptomatic carotid stenosis by histologic analysis. ² Cellular apoptosis has been strongly implicated in the process of plaque rupture. ³ Local secretion by inflammatory cells of important amounts of cytokines and other cytotoxic mediators (proteases, reactive oxygen species) induces the apoptosis of smooth muscle cells. The decreased synthesis of extracellular components by the smooth muscle cells causes progressive thinning and weakening of the fibrous cap, ultimately leading to plaque rupture. In addition, macrophage cell death releases significant amounts of tissue factor within the plaque, which in the case of plaque rupture will favor thrombus formation. ⁴ The number of apoptotic cells in atherosclerotic plaque might therefore represent both a marker for plaque rupture risk and an indirect assessment of the thrombotic potential of the plaque. Evaluation of apoptotic cell burden in atherosclerotic plaques using noninvasive imaging techniques therefore seems to be an interesting strategy for the identification of high-risk plaques. ⁵

In this study, our aim was to evaluate whether the presence of apoptotic cells in a rabbit model of atherosclerotic plaques could be detected with [¹⁸F]ML-10 and PET imaging.

Methods

Animal Model

Aortic atherosclerotic plaques were induced in male New Zealand White rabbits (n = 5; mean age 4 months; mean weight = 3.5 ± 0.2 kg) by a combination of 4 months of a high-cholesterol diet (0.3% cholesterol-enriched diet) and double balloon injuries of the aorta 2 and 6 weeks after starting the high-cholesterol diet. Aortic injury was performed from the aortic arch to the iliac bifurcation with a 4 French Fogarty embolectomy catheter introduced through the femoral artery as previously described.^6,7 All procedures were performed under general anesthesia by an intramuscular injection of ketamine (20 mg/kg; Fort Dodge Animal Health, Fort Dodge, IA) and xylazine (10 mg/kg; Bayer Corp., Shawnee Mission, KS). New Zealand White rabbits not subjected to the surgical procedures above and fed normal chow diets were used as controls (n = 5). The Bichat Animal Care and Use Committee approved all animal experiments.

Synthesis of [¹⁸F]ML-10

[¹⁸F]ML-10 (generously provided by Aposense Ltd, Petach-Tikva, Israel; and IBA Molecular, Gif sur Yvette, France) was synthesized from the precursor ML-10 mesylate. For radiolabeling, ML-10 mesylate in anhydrous acetronitrile was added by a nucleophilic substitution reaction to [¹⁸F]fluoride delivered from a cyclotron for a 15-minute reaction at 90°C as previously described. ⁸ The radiochemical purity was greater than 99%, and the specific activity was greater than 40.7 GBq/μmol.

In Vivo PET Imaging

Image Acquisition

After anesthesia with an intramuscular injection of ketamine (20 mg/kg), xylazine (10 mg/kg), and acepromazine (0.5 mg/kg; Fort Dodge Animal Health, IA) and placement of a 22-gauge catheter in the marginal ear vein, rabbits were injected with 15 MBq/kg of [¹⁸F]ML-10. The blood radioactive half-life after injection of [¹⁸F]ML-10 was measured at 23 minutes. PET imaging of rabbits started 150 minutes after intravenous injection using a combined PET–64-slice computed tomography (CT) scanner (GE Discovery 690 PET-CT, GE Healthcare, Waukesha, WI). PET acquisitions were performed in three-dimensional (3D) mode and covered the area from the neck to the aortic bifurcation with two contiguous imaging steps of 10 minutes. PET images were reconstructed with 3D attenuation-weighted ordered subsets expectation maximization (AW-OSEM), including correction for dead time, time of flight, geometry, scatter, and attenuation, using low-dose CT acquisition performed at the beginning of the imaging session. The PET reconstructed field of view was 25.6 × 25.6 cm² in the axial plan, with a 256 × 256 data matrix, giving an in-plane voxel size of 1 × 1 mm². PET slice thickness was 3.27 mm, giving a voxel volume of 3.27 mm³. In this configuration, the in-plane spatial resolution of the PET system is 5 mm.

At the end of the PET imaging, CT angiography (CTA) of the aorta was performed for each rabbit with 64-slice CT during the injection of 10 mL of a solution composed of 5 mL of iopamidol 300 (Bracco Imaging, Princeton, NJ) and 5 mL of saline. Acquisition parameters for the CTA were as follows: tube voltage, 100 kV; current intensity, 200 mA; rotation time, 0.4 seconds; detector collimation, 64 × 0.6 mm; and table feed, 39.4 mm per rotation. Axial 0.6 mm thick slices were reconstructed on the scanner with a 30f medium smooth kernel. The data were acquired with a field of view of 36 × 36 cm². Images were reconstructed with a field of view of 15 × 15 cm² with a 512 × 512 data matrix, giving a pixel size of 0.3 × 0.3 mm².

Image Analysis

Analysis of in vivo PET images was performed using the 4.4 Advantage Workstation (General Electric Medical Systems, Waukesha, WI). The abdominal aorta was localized on axial sections using images of CTA obtained at the end of the imaging session. Two-dimensional circular regions of interest (ROI) including the abdominal aortic wall were placed on areas corresponding to the aortas on axial PET sections. Maximal standardized uptake values (SUVs) adjusted for rabbit weight were measured in the ROI on 10 adjacent axial sections. The mean blood SUV was measured in a ROI placed in the right atria of each rabbit. Target to background ratios (TBRs) were calculated for each axial section by dividing the maximal SUV measured in the abdominal aorta by the mean blood SUV. The average TBR of the aorta was calculated for each rabbit. The maximal TBR was the highest TBR measured along the abdominal aorta in each rabbit. Measurements were performed by an independent operator (F.H.) blinded to the rabbit experimental group and to the histology results. In addition, the mean SUV was measured in circular ROI placed on each tissue identified using CTA at the time of PET imaging. Tissular contrast in the rabbits at the time of imaging was expressed at the ratio between the mean tissue SUV and the mean liver SUV used as a reference. Image contrast was evaluated only in tissues present in the field of view of the PET acquisition.

Ex Vivo PET Imaging

Following PET imaging, rabbits were euthanized by an intravenous injection of 120 mg/kg of sodium pentobarbital (Sleepaway, Fort Dodge Animal Health, Fort Dodge, IA). A bolus of heparin was injected prior to euthanasia to prevent clot formation. Aortas were excised, washed with saline, and imaged using a micro-PET system (Mosaic HP, Philips Medical Systems, Eindhoven, The Netherlands) with a 20-minute imaging step in the 3D mode. A 3D Ramla algorithm was used to reconstruct the images with a 128 × 128 × 128 data matrix in a field of view of 12 × 12 × 12 cm³, giving a final voxel size of 1 × 1 × 1 mm³. In addition, micro-CT (Nano-SPECT-CT, Bioscan, Mediso Ltd., Budapest, Hungary) was performed at the end of the PET imaging to localize the aortas using the following parameters: tube voltage, 45 kV, and current intensity, 0.177 mA. CT images were reconstructed with a 300 × 300 × 300 data matrix in a field of view of 6 × 6 × 6 cm³, giving a final voxel size of 0.2 × 0.2 × 0.2 mm³. PET and CT images were merged using a dedicated workstation (Mediso Ltd., Budapest, Hungary). The maximal number of counts was measured on five adjacent ROI placed on axial sections of the aorta using the Xeleris 2 Workstation (General Electric Medical Systems, Waukesha, WI).

Quantification of Activities in the Aortic Wall with a Gamma Counter

Aortas were cut into adjacent 4 mm segments. ¹⁸F activity present in each aortic segment (n = 20 for each aorta) was then measured using an automatic gamma-counter (Perkin Elmer Wizard 1480, 3-inch well-type detector of thallium activated, sodium iodide crystal, Perkin Elmer, Waltham, MA). The counting window was 400 to 600 keV in the extended range mode (2 keV per channel), and the counting time was 120 seconds. The relevant background value was automatically subtracted from each sample measurement. Each sample activity was measured at least twice and corrected for decay from the injection time and tissue weight. The activity was expressed as a percentage of the total injected activity (IA) adjusted for rabbit and tissue weights.

Autoradiography

Aortic segments of atherosclerotic (n = 20) and control rabbits (n = 20) were obtained 4 hours after [¹⁸F]ML-10 injection. Samples were frozen in Optimal Cutting Temperature compound, cut into 20 µm thick cryosections, and placed on a BetaImager (Biospace Lab, Nesles la Vallée, France) for 4 hours. For each autoradiographic acquisition, half of the aortic sections were from atherosclerotic rabbits and the other half from control rabbits. In addition, adjacent 10 µm thick cryosections were cut for histologic analysis. Autoradiographic images were exported in a linear 16-bit monochromatic tagged image file format (TIFF) and then analyzed using ImageJ software (National Institutes of Health, Bethesda, MD). After placing circular ROI on each aortic section, the total number of counts was measured in each ROI. In addition, background activity was measured by placing the circular ROI next to the aortic section. Activities in the aortic sections were expressed as the total number of counts subtracted from adjacent background activity and adjusted to the size of the ROI.

Histology

Apoptotic cells were stained with terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL; In Situ Cell Death Detection Kit, TMR red, Roche Diagnostics, Mannheim, Germany) according to the manufacturer's protocol. Briefly, cryosections were fixed in 4% paraformaldehyde for 30 minutes and incubated with Triton X-100 for 2 minutes to permeabilize cell membranes. After rinsing, each section was incubated with the TUNEL solution at 37°C for 1 hour in the dark followed by a 5-minute incubation with 4′,6-diamidino-2-phenylindol (DAPI). The presence of TUNEL-positive signals was evaluated on each section using a fluorescence microscope (excitation wavelength 520–560 nm; detection wavelength 570–620 nm). In addition, DAPI staining of nuclei and the presence of elastin autofluorescence were evaluated in the same cryosections (excitation wavelength 385–400 nm, detection wavelength 450–465 nm, and excitation wavelength 475–490 nm, detection wavelength 505–535 nm, respectively). The total area occupied by positive TUNEL signals detected in the nuclei was measured on each section using the automated, contrast-based area analysis function of the ImageJ software. Care was taken to select only TUNEL-positive areas located in the nuclei of cells with fused TUNEL-DAPI images.

Statistical Analysis

Numeric values are expressed as mean ± standard deviation. Statistical analysis was performed using SPSS software (SPSS Inc., Chicago, IL). Values of p < .05 were considered to be significant. SUV, activities, and the intensity of TUNEL staining were compared in atherosclerotic and control rabbits using a two-tailed Student t-test. Correlations were performed using the Pearson test.

Results

In Vivo PET Imaging after Injection of [¹⁸F]ML-10

Two and a half hours after intravenous injection of [¹⁸F]ML-10, the highest activity was located in the kidneys, which reflects the renal-mediated clearance of the radiotracer (Table 1). Moderate tracer accumulation was also detected in the bone marrow and in the intestines. In addition, focal areas of [¹⁸F]ML-10 accumulation were clearly identified in the abdominal aortas of atherosclerotic rabbits by PET imaging, whereas no uptake was detected in the aortas of control rabbits (Figure 1). The average TBR measured in the abdominal aorta with PET (Figure 2) was significantly higher in atherosclerotic than in control rabbits (2.00 ± 0.52 vs 1.22 ± 0.30; p < .05). Similarly, higher maximal TBR values were measured with PET in the most diseased segment of the abdominal aorta of atherosclerotic rabbits compared to control rabbits (2.68 ± 0.90 vs 1.60 ± 0.08; p = .05).

Table 1.

Tissular Contrast Measured in Rabbits with PET 150 Minutes after Injection of [¹⁸F]ML-10

Tissue	Contrast (tissue to liver ratio)
Kidneys	15.1 ± 4.9
Bone marrow	5.5 ± 3.0
Intestines	2.4 ± 0.9
Heart	1.7 ± 0.7
Bone	1.7 ± 1.1
Spleen	1.5 ± 0.9
Blood	1.3 ± 0.2
Lungs	0.9 ± 0.2
Muscle	0.5 ± 0.2

PET = positron emission tomography.

Figure 1.

Representative examples of positron emission tomography (PET) images, computed tomography angiography (CTA) images, and merged PET-CT acquisitions obtained in an atherosclerotic rabbit and a control rabbit 150 minutes after the injection of [¹⁸F]ML-10. Note the intense uptake of [¹⁸F]ML-10 with PET (yellow and red arrows) in regions corresponding to the atherosclerotic abdominal aorta with CTA (the red arrow is placed on the coronal view at the level of the corresponding axial section). In contrast, no significant uptake of [¹⁸F]ML-10 was detected in the aorta of a control rabbit with PET. Note the intense accumulation of [¹⁸F]ML-10 in kidneys and the bladder and the moderate tracer uptake in the bone marrow of rabbits, as formerly described in rats ⁸ and humans. ²⁸ In rabbits, nonspecific accumulation of [¹⁸F]ML-10 was also present in the stomach and the intestines.

Figure 2.

Quantification of [¹⁸F]ML-10 uptake in the aortic wall. Average and maximal target to background ratios (TBRs) measured in the abdominal aorta in vivo with positron emission tomography (PET) 150 minutes after the injection of [¹⁸F]ML-10 were higher in atherosclerotic (Ath) compared to control rabbits. Significant differences in [¹⁸F]ML-10 uptake between atherosclerotic and normal aortas were further confirmed by ex vivo PET imaging and by quantification of the percentage of injected activity (IA) corrected for tissue weight (% IA/g) measured using a gamma-counter. *p < .05; ^†p = .05.

Ex Vivo Quantification of [¹⁸F]ML-10 Accumulation in the Aorta

At the end of in vivo PET imaging, aortas were imaged ex vivo using a micro-PET-CT system. A strong signal was detected in the aortic wall of atherosclerotic rabbits, whereas no significant signal could be detected in the aortas of control rabbits (Figure 3). The number of counts measured ex vivo with PET (see Figure 2) was significantly higher in the aorta of atherosclerotic versus control rabbits (24 ± 11 vs 1.0 ± 2.0 counts; p < .05). In addition, the amount of radioactivity present in the aortic wall was measured using a gamma-counter and expressed as a percentage of the injected activity corrected for tissue weight (% IA/g). Significantly higher activity was detected in the aorta of atherosclerotic versus control rabbits (3.6 ± 2.0 vs 0.05 ± 0.05 % IA/g; respectively; p < .05).

Figure 3.

Representative examples of merged PET-CT images acquired 3 hours after injection of [¹⁸F]ML-10 in aortas excised from atherosclerotic (Ath) and control rabbits. Note the intense uptake of [¹⁸F]ML-10 in atherosclerotic plaques compared to the normal aorta of a control rabbit.

Correlation between [¹⁸F]ML-10 Accumulation and the Presence of Apoptotic Cells

Average activities measured by autoradiography in aortic sections (Figure 4) were significantly higher in atherosclerotic plaques compared to the normal aortic wall (20.4 ± 10.5 vs 0.6 ± 0.5 counts/mm²; p < .05). Similarly, the mean area occupied by TUNEL-positive nuclei was significantly larger in aortic sections from atherosclerotic rabbits relative to control rabbits (105,755 ± 45,079 vs 0 ± 0 µm²; p < .05). In addition, atherosclerotic plaque sections that exhibited high-count rates, as detected by autoradiography, contained high densities of apoptotic cells in the corresponding histologic sections stained by TUNEL (Figure 5). In contrast, atherosclerotic plaques with low autoradiographic activity as well as normal aortic walls from control rabbits contained only a few apoptotic cells detected by TUNEL. A strong positive correlation (r² = .75; p < .05) was obtained between the autoradiography count number and the TUNEL-positive score in the aortic wall of adjacent histologic sections (see Figure 4).

Figure 4.

Comparison between activities detected in aortic segments by autoradiography and the presence of apoptotic cells on histology. Significantly higher activity was measured after the injection of [¹⁸F]ML-10 by autoradiography in the aortic sections from atherosclerotic compared to control rabbits. Apoptotic cells (TUNEL-positive nuclei) were detected only in the aortic walls of atherosclerotic (Ath) rabbits. A strong correlation (r² = .75; p < .05) was measured between the activities measured by autoradiography after injection of [¹⁸F]ML-10 and TUNEL-positive areas containing apoptotic cells on adjacent histologic sections. *p < .05.

Figure 5.

Representative examples of aortic sections evaluated by autoradiography and histology. Autoradiographic analyses of aortic sections acquired 4 hours after the injection of [¹⁸F]ML-10 are presented in the first row (dotted lines show the location of aortic segments). Adjacent cryosections were imaged after Masson trichrome staining to identify aortic wall morphology (second row) and using a fluorescence microscope after DAPI staining for nuclei (third row, blue), TUNEL staining for apoptotic cells (fourth row, red), and staining for elastin autofluorescence (fifth row, green). Images acquired at the different wavelengths were then merged to identify apoptotic cells presenting combined TUNEL- and DAPI-positive staining (sixth row, pink). In the first column, note the presence of intense activity on autoradiography after the injection of [¹⁸F]ML-10 in this atherosclerotic plaque, which was associated with the presence of a large number of apoptotic cells (red dots, white arrow: TUNEL-positive nuclei; pink dots: combined DAPI-TUNEL-positive areas on the merged image) on corresponding adjacent histologic sections. In the second column, a representative example of an atherosclerotic plaque with low activity on autoradiography after injection of [¹⁸F]ML-10 and only a few apoptotic cells on the corresponding adjacent histologic section. In the third column, a representative example of an aortic section of a control rabbit showing no significant activity on autoradiography and the absence of apoptotic cells on the corresponding histologic section. Black scale bar = 2 mm. White scale bar = 500 µm. *Aortic lumen. White arrowheads = internal elastic lamina.

Discussion

In this study, using a rabbit model of atherosclerosis, we demonstrated that the presence of apoptotic cells in atherosclerotic plaques can be detected in vivo by PET imaging after injection of [¹⁸F]ML-10. Significantly higher uptake of [¹⁸F]ML-10 was detected in atherosclerotic plaques of rabbits compared to normal aortic walls both in vivo and ex vivo. In addition, the intensity of [¹⁸F]ML-10 accumulation in the aortic wall quantified on autoradiographs correlated with TUNEL-positive areas measured on adjacent histologic sections.

PET Radiotracers for the Evaluation of High-Risk Plaques

Several PET radiotracers have been described for the evaluation of atherosclerotic plaques.^9–12 [¹⁸F]Fluorodeoxyglucose (FDG) uptake in the arterial wall was first noted in the aorta of patients undergoing PET imaging for cancer^13,14 and is likely attributable to accumulation within activated inflammatory cells. Indeed, strong relationships were found between the intensity of FDG accumulation in atherosclerotic plaques and the number of macrophages measured on histology in both experimental models of atherosclerosis^7,15 and in patients undergoing carotid endarterectomy. ¹⁶ In addition, patients presenting with a recent ischemic cerebral event showed a significantly higher FDG uptake in carotid plaques compared to asymptomatic patients. ¹⁷ More recently, accumulation of [¹⁸F]-sodium fluoride has also been described in high-risk atherosclerotic plaques in the carotid and coronary arteries. ¹⁸ The precise mechanism by which [¹⁸F] sodium fluoride accumulates in high-risk plaques remains to be determined. These two radiotracers are widely available but do not target specific biological activities present in high-risk atherosclerotic plaques. Analysis of their uptake in the vessel wall can therefore be hampered by the physiologically intense uptake in adjacent tissues (peripheral muscle, myocardium, brain, or blood), as in the case of FDG, or binding to other compounds present in atherosclerotic plaques, such as arterial calcifications in the case of [¹⁸F]-sodium fluoride. Evaluation of atherosclerotic plaques might be facilitated by the use of radiotracers targeting more specific biological activities present in high-risk atherosclerotic plaques.

Detection of Apoptotic Cells in Atherosclerotic Plaques with Annexin V

Several noninvasive imaging techniques have been described for the detection of apoptotic cells in atherosclerotic plaques. Annexin V is a 36 kDa protein that binds with high affinity to phosphatidylserine head groups expressed on the cell surface at early stages of apoptosis. Annexin V radiolabeled with technetium 99m was the first radiotracer able to detect apoptotic cells in animal models of atherosclerotic plaques.^19,20 In addition, in a proof-of-concept study, accumulation of radiolabeled annexin V was detected by scintigraphy in carotid atherosclerotic plaques of patients who presented with an acute ischemic stroke and was associated with the presence of apoptotic cells on corresponding carotid plaques obtained after carotid endarterectomy. ²¹ However, annexin V scintigraphy has significant limitations for the evaluation of atherosclerotic plaques. First, the relatively low spatial resolution of monophotonic gamma cameras in comparison with the thickness of the vessel wall represents a clear drawback for molecular imaging of atherosclerotic plaques. Recent development of kits for radiolabeling of annexin V for PET imaging might help overcome this limitation.^22,23 Second, annexin V was found to bind not only apoptotic cells but also necrotic cells by reaching intracellular phosphatidylserine through disrupted plasma membranes. ²⁴ Third, annexin V is characterized by a relatively slow clearance from tissues because of its relatively high molecular weight. Hence, new strategies have been developed to enable imaging of apoptotic cells using PET radiotracers. ²⁵

[¹⁸F]ML-10 for the Detection of Apoptotic Cells

[¹⁸F]ML-10 seems well suited to overcome the limitations of scintigraphy with annexin V for detection of apoptotic cells in atherosclerotic plaques. [¹⁸F]ML-10 has a low molecular weight, a fast renal clearance, and a relatively short half-life in the blood, which allows for rapid diffusion of the radiotracer within atherosclerotic plaques and limits its nonspecific retention. Intense tissular accumulation of compounds similar to [¹⁸F]ML-10 was found in several experimental models of apoptosis induced by anticancer agents in tumors, ²⁶ models of renal failure, ²⁷ and neurodegenerative diseases, ⁸ as well as in an experimental model of transient cerebral ischemia. ⁸ Accumulation of these compounds correlated strongly with the presence of apoptotic cells. In addition, [¹⁸F]ML-10 demonstrated a favorable dosimetry, biodistribution, stability, and safety profile in healthy humans ²⁸ and has already been administered for the detection of apoptotic cells in patients with glioblastoma. ²⁹ Detection of [¹⁸F]ML-10 in atherosclerotic rabbit plaques, as described in this study, supports the clinical translation of this radiotracer for atherosclerotic plaque imaging.

Vascular Wall Imaging with PET

In our study, the intensity of [¹⁸F]ML-10 accumulation measured in vivo by PET imaging in the atherosclerotic plaques of rabbits was relatively low. In contrast, the radioactivity measured in atherosclerotic plaque samples using a gamma-counter was approximately 50-fold higher than that measured in vivo by PET. These differences can be largely explained by partial volume effects, which occur when the imaged structure spatial resolution is smaller than two full widths at half maximum of the spatial resolution and/or smaller than the voxel size. Partial volume effects lead to significant underestimation of peak signals present in the vascular wall ³⁰ because focal and intense signals will be averaged in the whole volume corresponding to the voxel size of the selected spatial sampling (known as “tissue fraction effect”) and will also be spread in surrounding voxels by the spatial resolution of the imaging system (known as “spill-out”). These effects play an important role in imaging the vascular wall using nuclear medicine techniques and represent a significant limitation for the evaluation of radiotracers in animal models of atherosclerotic plaques. In addition, the dose of radiotracer injected in animals was increased to at least partially overcome the decreased signal in relation to partial volume effects. Advanced human carotid atherosclerotic plaques are three- to fourfold thicker than rabbit plaques and should be less subject to intense partial volume effects than in this experimental study. Therefore, the dose of radiotracer injected might be significantly reduced in clinical studies.

Limitations

This study has a few limitations. First, no competition study could be performed because of the receptor-independent accumulation of the radiotracer in cells. Second, the precise localization of [¹⁸F]ML-10 within the atherosclerotic plaques could not be evaluated because of the limited spatial resolution of autoradiographs and relatively low signal associated with short-life positron-emitting radiotracers. Alternatively, injection of a fluorescent molecule similar to [¹⁸F]ML-10 may have helped identify cells that accumulate these apoptosis-targeting compounds but could not be performed because the volume that needed to be injected in rabbits to obtain a detectable signal was too high to be tolerated by animals. Third, the intensity of the signal measured in this study in the atherosclerotic plaques of rabbits after injection of [¹⁸F]ML-10 was not compared head to head with the FDG signal. However, we did measure ex vivo similar concentrations of radiotracer (Bq/tissue weight) in the atherosclerotic plaques of rabbits imaged separately after injection of [¹⁸F]ML-10 or FDG (data not shown). Fourth, the results obtained in an experimental model of atherosclerosis need to be validated in complex human atherosclerotic plaques. Finally, whether the number of apoptotic cells present in atherosclerotic plaques is associated with the risk of plaque rupture and subsequent thrombosis still needs to be validated.

Conclusion

In this study, the presence of apoptotic cells in atherosclerotic plaques could be detected with [¹⁸F]ML-10 and PET in a rabbit model. Hence, [¹⁸F]ML-10 represents a promising PET radiotracer for the identification of apoptotic cells in human atherosclerotic plaques. These results are in line with previous studies where the presence of acute cerebral ischemia was also evidenced with [¹⁸F]ML-10 in an experimental model of carotid transient occlusion. [¹⁸F]ML-10-PET imaging may therefore represent an attractive approach in patients with carotid stenosis, allowing for the simultaneous evaluation of the presence of apoptotic cells in carotid atherosclerotic plaques and the detection of cerebral ischemia.