A Molecular Imaging Strategy for In Vivo Monitoring of Vascular Remodeling in a Porcine Model of Femoral Artery Overdilation

Animal Surgical Preparation and Procedures

Experimental protocols were approved by the Institutional Animal Care and Use Committee of Nationwide Children’s Hospital and were in compliance with the Association for Assessment and Accreditation of Laboratory Animal Care International policies. All procedures performed in animals also conformed to the National Institutes of Health Guide for the Care and Use of Laboratory Animals. The use of radioisotopes and all PET/CT imaging procedures were approved by the Radiation Safety Committee of Nationwide Children’s Hospital.

Eight female Yorkshire pigs (23.4 ± 4.2 kg) approximately 8 weeks of age underwent percutaneous transluminal angioplasty of the right femoral artery to induce vascular injury. Animals were fed and maintained on a standard chow diet for the duration of the study. Prior to surgery, animals were fasted for a minimum of 8 h. Animals were then sedated by intramuscular administration of ketamine (15–20 mg/kg) and xylazine (1–2 mg/kg), followed by intubation and mechanical ventilation with 100% oxygen (VentiPAC, Smiths Medical International Ltd., Luton, UK). For maintenance of anesthesia during surgery, animals were intravenously administered propofol at a continuous rate of 10–30 mg/kg/hour through an ear vein catheter. Depth of anesthesia was monitored every 15 min by evaluating blink reflex and jaw tone. Blood pressure, heart rate, oxygen saturation, and body temperature were continuously monitored for the duration of the surgery (IntelliVue MP30; Philips Healthcare, Andover, MA).

To induce unilateral femoral artery injury via balloon angioplasty, the carotid artery was first accessed by percutaneous puncture, and a 5-F polyethylene introducer sheath (PINNACLE®, Terumo, Somerset, NJ, USA) was placed. 100 units of heparin per kilogram of body weight were intravenously administered prior to the insertion of wires. An 0.035” J-tip guidewire (Boston Scientific, Marlborough, MA, USA) was then positioned in the distal femoral artery under fluoroscopy guidance (OEC 9900 Elite C-Arm, GE Healthcare, Chicago, IL, USA), a 4-F angiocatheter (Mongoose, Pediavascular, Chagrin Falls, OH, USA) was inserted over the guidewire to the level of the abdominal aorta, the wire was removed, and an angiographic roadmap of the lower extremities was subsequently acquired (Fig. 1A). The guidewire was re-inserted and the angiocatheter was then exchanged for a 5-F over-the-wire balloon catheter (7 × 40 mm, Mustang Dilatation Catheter, Boston Scientific, Marlborough, MA, USA), which was inflated once for 3 min using a balloon-to-artery ratio ranging from ~ 1.25 to 1.85 to induce varying degrees of injury to the right femoral artery. The balloon was then deflated, the catheter and wire were removed, and protamine was intravenously administered (1 mg per 100 units of heparin administered) prior to removal of the carotid artery sheath to limit bleeding. Details of the femoral artery and balloon characteristics are outlined in Table 1.

¹⁸F-FDG PET/CT Imaging Protocol

Animals were fasted prior to all imaging procedures for a minimum of 8 h. Propofol was intravenously infused at 10–30 mg/kg/h through an ear vein catheter for maintenance of anesthesia levels during all imaging procedures. Depth of anesthesia, hemodynamics, and body temperature were monitored every 15 min, as described during the surgical intervention. The body temperature was maintained above 37 °C during imaging procedures. Prior to PET/CT imaging, jugular vein access was established via percutaneous puncture and placement of a 4-F polyethylene catheter (Cook Medical, Bloomington, IN), which enabled venous administration of ¹⁸F-FDG for PET imaging. Prior to each PET scan, approximately 1 ml of venous blood was sampled to measure whole-blood glucose concentration using a commercially available blood glucose monitor (AlphaTRAK 2; Zoetis Inc., NJ, USA).

¹⁸F-FDG PET/CT imaging was performed 2 weeks following angioplasty of the right femoral artery using a hybrid PET/CT system (Discovery 690, GE Healthcare, Chicago, IL) 75 min after intravenous administration of ¹⁸F-FDG (dose 306.4 ± 38.3 MBq). The PET imaging field-of-view was set from the lower abdomen to the lower legs to ensure full scan coverage of the peripheral arteries of interest. CT images were first acquired with a slice thickness of 3.27 mm, at 120 kVp, and 300 mA, for image registration, attenuation correction, and manual segmentation of the femoral arteries. PET imaging was then performed for 8 min per bed position. PET data were reconstructed at 3.27 mm slice thickness with a pixel size of 2.73 mm using the OSEM algorithm with three iterations and 32 subsets, and Gaussian post-reconstruction filter with full width at half maximum of 4 mm. Following completion of hybrid PET/CT imaging, CT angiography images were acquired with iodinated contrast to facilitate identification of the femoral arteries and aid in PET image segmentation. CT angiography images were acquired at a slice thickness of 0.625 mm, at 300 mA, and 140 kVp. Intravenous contrast injections were performed using an integrated power injector (Stellant D, Medrad Inc, Warrendale, PA) at a constant rate of 3 ml/s and total volume of 30 ml, followed by a 0.9% saline flush at 3 ml/s.

¹⁸F-FDG PET/CT Image Analysis

All PET/CT image analysis was performed using commercially available software (PMOD Technologies LLC, Fallanden, Switzerland). Co-registered PET and CT images were used to define arterial regions of interest (ROIs). CT angiograms were utilized to aid in the identification and segmentation of the femoral artery for each animal. ROIs were then translated and resampled to match PET spatial resolution (3.27 mm) and quantify ¹⁸F-FDG uptake for the injured and control arteries on fused PET/CT attenuation images. For the arterial segmentation process, circular ROIs were used for each axial slice that were ~ 2 mm larger than the outer diameter of the femoral artery to account for the inherent limitations of PET spatial resolution and potential partial volume effects. Additionally, 2 human analysts performed PET/CT image segmentation of the femoral artery to evaluate inter-rater agreement of the PET image quantification process in the pig model. Given that the right hindlimb served as the injured limb in all animals for model consistency, image analysts were not blinded as to which femoral artery represented the injured or control vessel.

¹⁸F-FDG uptake values were quantified within each arterial ROI and expressed as standardized uptake values (SUVs). To obtain the average maximum SUV (i.e., mean SUV_max) for both femoral arteries, maximum SUVs measured per ROI of each slice of the femoral artery were summed and averaged. ¹⁸F-FDG uptake for each femoral artery was ultimately evaluated and expressed as mean SUV_max.

To measure the arterial diameters of native (control) and injured femoral arteries, CT angiography images were utilized and processed with FDA-approved DICOM viewer software (Osirix, Pixmeo, Switzerland).

Vascular Tissue Collection and Analysis

Immediately following completion of PET/CT imaging at 2 weeks post-angioplasty, animals were euthanized and femoral arteries from the injured (right) and control (left) limbs were harvested, flushed, and washed with PBS to remove potential blood clots, and cut into 5 mm sections. Three 5 mm arterial samples were used to represent the injured portion of artery and three 5 mm samples from the corresponding anatomical location were used to represent the contralateral non-injured femoral artery. Samples were transferred to PBS-filled tubes to measure ¹⁸F-FDG radioactivity/uptake for each 5 mm section, which was expressed as percent injected dose per gram of tissue. All samples were weighed, counted, and corrected for radioactive decay using a gamma well counter (WIZARD2, PerkinElmer, Waltham, MA). One animal’s tissue samples were unable to undergo gamma counting due to mechanical failure of the gamma counter on the study date. Therefore, all analysis involving gamma counting included 7, not 8, datasets. Following completion of gamma counting, samples were fixed with 4% paraformaldehyde overnight and then stored in PBS at 4 °C until paraffin embedding. Embedded tissue was then sectioned at 4 μm thickness.

Hematoxylin and eosin (H&E) staining was performed on formalin-fixed paraffin-embedded arterial sections that were deparaffinized using xylene and grades of ethanol. Hematoxylin solution (Sigma Aldrich, GHS332) and Eosin Y (Sigma Aldrich, HT110332) were used according to the manufacturer’s protocol. For immunohistochemical staining (IHC) of CD64 and α-smooth muscle actin (α-SMA), embedded arterial sections were deparaffinized using xylene and grades of ethanol followed by antigen retrieval with citrate-based antigen unmasking solution (H-3300, Vector laboratories). After antigen retrieval, tissue sections were incubated with 3% H₂O₂ diluted in water at room temperature to quench endogenous peroxidase activity and blocked with 3% normal goat serum (Abcam, ab7481) in Background Sniper (Biocare Medical, LLC) for 30 minutes at room temperature to avoid nonspecific binding. Endogenous biotin was blocked using an avidin/biotin blocking kit (Abcam, ab64212) with avidin incubation for 15 min followed by biotin incubation for 15 min per the manufacturer’s protocol. Sections were then incubated with mouse anti-CD64 (1:1000, Abcam, ab140779) and mouse anti-α-Smooth Muscle Actin (anti-α-SMA) (1:1000, Abcam, ab7817) overnight at 4 °C. The next day, sections were incubated in biotinylated goat anti-mouse IgG secondary antibody for 30 min (Abcam, ab6788), followed by incubation in Vectastain Elite ABC-HRP Reagent, Peroxidase (PK-7100, Vector Laboratories) for 30 min. Sections were washed then visualized using the SignalStain DAB Substrate Kit (8059, Cell Signaling Technology).

Image quantification was performed using ImageJ (National Institutes of Health, Bethesda, MD) on images acquired at 20× magnification using 3 sections per artery (Axio Observer, Carl Zeiss, White Plains, NY). A plug-in on ImageJ function named “Local Thickness” was used to quantify vessel thickness [13]. Briefly, images of arteries, including the intima, media, and adventitia, were binarized. Then, a Euclidean distance transform algorithm was applied to create distance maps at every pixel contained by the vessels. The algorithm then computed all local thickness values as determined by the diameter of the largest circle at any point within the vessel region and provided a histogram of thickness values with a default total bin number of 256. The mean of all values was reported as the mean thickness. To quantify the total positively stained area of blood vessels, IHC images underwent color deconvolution to extract DAB-only images using ImageJ. All images were thresholded using a value determined from a positive control image. The threshold value was a mean of values determined by three different users.

One animal had complete arterial dissection in one of three samples, rendering image quantification for the section imprecise. Therefore, in one animal, two representative sections instead of three were used for gamma counting/histological evaluation as a quality-control measure to ensure consistency across all measures. H&E images for all animals were evaluated for total vessel thickness, while CD64 and α-SMA expression was reported as total positively stained area.

Statistical Analysis

A Wilcoxon signed-rank test was used to compare differences in ¹⁸F-FDG uptake between injured and control femoral arteries from the in vivo PET/CT imaging and ex vivo gamma counting measures. Pearson’s correlation was used to assess the linear relationship between in vivo PET/CT imaging and ex vivo gamma counting measures of ¹⁸F-FDG uptake. To account for repeated measures within animals (three sections per artery per animal) linear regression model with mixed effects was used to compare differences in arterial thickness, α-SMA positive area, and CD64 positive area between control and injured femoral arteries. Animal ID was included as a random effect to account for within animal clustering. The associations between arterial uptake of ¹⁸F-FDG and (1) arterial thickness, (2) α-SMA positive area, and (3) CD64 percent area were assessed using a mixed-effects linear regression model with random effect for animal. Normalized residuals from the models were examined to assess the potential influence of observations. A paired t-test was used to compare differences in CT angiography-derived measures of arterial diameter between control and injured femoral arteries. Intra-class correlation coefficient was used to evaluate the inter-rater agreement for PET image quantification. A p value < 0.05 was considered statistically significant. All statistical analyses were performed using GraphPad Prism, version 10.1.0 for windows and R statistical software (v4.5.0; R Core Team 2025).

In Vivo ¹⁸F-FDG PET/CT Imaging of Vascular Remodeling

In vivo PET/CT imaging 2 weeks after balloon angioplasty demonstrated qualitative increases in femoral artery uptake of ¹⁸F-FDG in injured versus control arteries (Fig. 1B–D). Quantitative PET image analysis, expressed as the mean SUV_max across the length of ballooned arterial segment, confirmed a significant difference for in vivo ¹⁸F-FDG uptake between injured and control femoral arteries (injured artery: 1.2 ± 0.7 SUV_max; control artery: 0.8 ± 0.4 SUV_max; p = 0.008) (Fig. 1E). Inter-rater agreement across 2 human analysts for quantification of femoral artery ¹⁸F-FDG uptake (i.e., mean SUV_max) was good to excellent with an intraclass correlation coefficient of 0.88.

In vivo CT angiography imaging demonstrated a significant difference in femoral artery diameter at 2 weeks after balloon angioplasty between injured and control femoral arteries (injured artery: 5.2 ± 0.8 mm; control artery: 4.6 ± 0.6 mm; p = 0.04) (Table 1).

Ex vivo Gamma Counting of Vascular ¹⁸F-FDG Uptake

Postmortem gamma counting of arterial samples at 2 weeks after balloon angioplasty demonstrated a significant increase in absolute measures of ¹⁸F-FDG uptake in injured versus control arteries (injured artery: 0.7 ± 0.6% injected dose/gram; control artery: 0.07 ± 0.03% injected dose/gram; p = 0.02) (Fig. 2B). Additionally, arterial uptake of ¹⁸F-FDG measured by ex vivo gamma counting was significantly and positively correlated with ¹⁸F-FDG uptake measured by in vivo PET/CT imaging (r = 0.92, p = 0.004) (Fig. 2C).

Histological Analysis of Vascular Remodeling

H&E, α-SMA, and CD64 staining of arterial sections demonstrated qualitative differences in arterial thickness and SMC proliferation between control and injured arteries 2 weeks after balloon angioplasty (Fig. 3A, B). Quantitative image analysis further confirmed significant increases in arterial thickness (injured artery: 1454.0 ± 615.3 μm; control artery: 505.8 ± 72.3 μm; β = 951.0; p < 0.0001), α-SMA positive area (injured artery: 3005049.9 ± 2457929.5 μm²; control artery: 1105610.3 ± 429915.0 μm²; β = 1909473; p = 0.0001), and CD64 percent positive area (injured artery: 1.8 ± 1.9%; control artery: 0.6 ± 1.0%; β = 1.2; p = 0.04) in injured versus control arteries (Fig. 3C–E).

Association Between ¹⁸F-FDG Uptake and Markers of Remodeling

Evaluation of the potential associations between vascular uptake of ¹⁸F-FDG in injured arteries (measured by gamma counting) and vascular remodeling characteristics demonstrated that ¹⁸F-FDG uptake was significantly and positively associated with artery thickness (β = 632.1; 95% CI 324.0, 940.2; p = 0.0003) (Fig. 4A), smooth muscle cell content (β = 3617371; 95% CI 2993449.1, 4241291.2; p < 0.0001) (Fig. 4B), and CD64 expression (β = 1.46; 95% CI 0.3, 2.7; p = 0.02) (Fig. 4C) at 2 weeks after balloon angioplasty.

Conflict of interest

The authors have no relevant financial or non-financial interests to disclose.