Intravascular Fluorescence Molecular Imaging of Atherosclerosis

Abstract

Optical molecular imaging using near-infrared fluorescence (NIRF) light is an emerging high-resolution imaging approach to image a wide range of molecular and cellular species in vivo. Imaging using NIR wavelengths (650–900 nm) enables deeper photon penetration into tissue and reduced tissue autofluorescence, resulting in higher sensitivity to detect exogenously administered NIR fluorophores (injectable molecular imaging agents). Greater imaging depth of several centimeters is further achievable in the NIR window as blood absorption is as an order of magnitude lower than in the visible range. Furthermore, as optical imaging is routinely performed in the cardiac catheterization laboratory (e.g. optical coherence tomography), intravascular NIRF offers a promising translational approach for clinical coronary and peripheral arterial imaging. To this point, the first human intravascular NIRF imaging study recently demonstrated the ability to detect NIR autofluorescence in patients with coronary atherosclerosis. This study provides a foundation for targeted intravascular NIRF molecular imaging studies in coronary patients. In this chapter, we detail system engineering, imaging agents and translational applications of intravascular NIRF molecular imaging.

1.0. Introduction

Atherosclerosis is the dominant form of coronary artery disease (CAD) and the major cause of death worldwide. Clinically, non-invasive imaging modalities such as computed tomography (CT), magnetic resonance imaging (MRI) and ultrasound (US) are widely used for diagnosis of atherosclerosis in larger vessels, and in the case of CT, both in coronary arteries and larger arteries. Nonetheless, these techniques have limited ability to provide specific biological and molecular information underlying atherosclerosis progression and complications. Intravascular near-infrared fluorescence (NIRF) molecular imaging is a promising approach for comprehensively understanding coronary atherosclerosis on cellular and molecular level [1–3].

Fluorescence spectroscopy is the optical detection of molecules that absorb a particular spectrum of electromagnetic radiation (light) and then re-emit electromagnetic radiation at another lower energy level. Figure 1 illustrates a simplified Jablonski diagram [4] that demonstrates the fluorescence mechanism. The absorption of a single photon at an appropriate frequency, and this energy, causes an electron to move from the electronic ground state to a high-energy state. Over a very short period of time (i.e. 10⁻¹² s)[5], the electron loses some energy to the environment (in the form of heat) which returns the electron to the lowest singlet excited state. After some time in this state (i.e. 10⁻⁸ s)[5], the electron is then able to return back to the electronic ground state by emitting the remaining energy as another single photon at a lower frequency (longer wavelength).

FIGURE 1: Jablonski diagram of the fluorescence mechanism.

A photon is absorbed causing the molecule to move to an excited electronic state (blue line). The molecule loses some energy to heat as it moves to the lowest excited state (green). A photon is spontaneously emitted allowing the molecule to return to the ground state (red line).

The optical detection of this process, known as fluorescence spectroscopy, is conducted by exciting the fluorophore with a light source at an appropriate wavelength (ideally tuned to the peak absorption of the fluorophore) and collecting the emission light that is released at a longer wavelength. The excitation source can be a laser, light-emitting diode or a broadband filament bulb with an appropriate filter in place. The difference between the wavelength of maximum absorption and the wavelength of maximum emission is known as the Stokes’ shift. The larger the Stokes’ shift, the easier it is to resolve the emission light from the excitation light. It is common for an organic fluorophore to have a Stokes’ shift of approximately 10 – 20 nm.

Near-infrared fluorescence (NIRF) imaging detects fluorophores that absorb and emit light in the wavelength range of 650–900 nm. NIR fluorescence detection is more efficient than visible fluorescence detection due to reduced scatter [6] and lower autofluorescence [7] of common catheter substrates such as nitrocellulose, polyvinylidene difluoride (PVDF), poly(methylmethacylate) (PMMA) and polydimethylsiloxane (PDMS). NIR fluorescence is also useful for in vivo imaging due to its deeper penetration into tissue and reduced tissue autofluorescence, as compared to visible light. The deeper penetration is mostly enabled due to the relatively lower absorption and scatter of NIR light by hemoglobin (oxygenated and deoxygenated) and water. Figure 2 illustrates the extinction coefficient of deoxygenated hemoglobin and oxygenated hemoglobin in the wavelength range of 400 to 1000 nm [8]. The extinction coefficient of a material is the amount of light attenuation per mole of the material, per unit length of depth. In addition to the increased depth of penetration, the lower absorption and scatter of biological materials in the NIR also reduces the autofluorescence and total background signal, which ultimately leads to improved signal-to-noise. The advantages of NIR fluorescence for in vivo imaging of animals and humans has led to applications, including imaging of normal and disease vasculature, tissue perfusion, protease activity, hydroxyapatite and malignancy [9].

FIGURE 2:

Molar extinction coefficients of oxy- (red)/deoxy- (blue)- haemoglobin over a large spectral range

Given the favorable high sensitivity in the NIR range, NIRF imaging can further enable signal detection through blood, without the need for flushing. The first real-time intravascular catheter molecular sensing probe was described by Jaffer et al.[10], who detected in vivo inflammatory cysteine protease activity in experimental atheroma of human-sized arteries of the aorta of rabbits. The first rotational, automated 2D imaging catheter was engineered by Jaffer et al.[11] who performed intra-arterial imaging of stented rabbit aortas and coronary bare-metal stents, and demonstrated excellent nanomolar sensitivity to fluorophores resident within plaque and stent injury-induced arterial inflammation. Further research into this novel imaging approach will also employ new tagged-monoclonal autoantibodies implicated in atherosclerosis [i.e. oxidized LDL (oxLDL), acetylated LDL, matrix metalloproteinases). Indeed, Khamis et al. [12] have developed a quantitative antibody-based NIRF approach for targeting oxLDL in vivo – providing new potential avenues for exploring molecular compositions of atheroma oxidative stress longitudinally. In addition, Food and Drug Administration (FDA)-approved indocyanine green is a promising translational NIR fluorophore for plaque imaging [13], and has recently been shown to be a marker of impaired endothelial barrier function in human plaques [14].

Intravascular NIRF imaging was introduced by engineering an optical fiber-based catheter to excite NIR fluorophores to sense molecular information within the arterial wall. After initial studies of standalone NIRF systems [10][11], it became apparent that standalone NIRF imaging was limited by the lack of co-registered anatomic information. Therefore, a dual-modality NIRF catheter incorporating the high-resolution clinical structural imaging method, optical coherence tomography (OCT), was developed for acquiring co-registered structural-molecular information simultaneously. Intravascular OCT is a highly sensitive, minimally invasive imaging modality that provides high resolution (~10 μm) cross-sectional views of the tissues, enables a real-time in vivo optical “biopsy” of arterial wall to identify the morphological features of the plaque such as thin cap fibroatheroma (TCFA), macrophages, calcifications, thrombus and cholesterol crystals. Combination of NIRF with OCT (NIRF-OCT) is technically straightforward compared to the combination of NIRF with intravascular ultrasound (IVUS), as with NIRF-OCT, the two modalities share the same optical components to deliver and receive light at the catheter tip. In order to acquire co-registered NIRF and OCT simultaneously, simultaneous light delivery and reception is enabled by use of a dual-channel rotary junction and a double-clad fiber (DCF) based catheter (Figure 3) [15].

FIGURE 3:

Dual-modality intravascular NIRF-OCT imaging system and catheter (reproduced from reference[18] with permission)

Intravascular NIRF-OCT was demonstrated for in vivo characterization of inflammatory cells using OCT and quantify molecular activity using cathepsin protease-activated NIR molecular beacons. Furthermore, stent-induced injury and healing, as reflected by fibrin deposition, can be quantified in vivo and investigated using this single pullback technology to resolve different stent healing profiles [16]. The translational potential of NIRF-OCT is evident by the recent development of an intravascular NIRF-OCT system allowing reporting on NIR autofluorescence in the coronary arteries of patients, a new signature of coronary CAD that may indicate intraplaque hemorrhage and oxidative stress [17].

This chapter details the methodology of the use of intravascular NIRF molecular imaging to assess the effect of interventions on atheroma inflammation in coronary-sized arteries in vivo. We highlight a specific application of assessing paclitaxel drug-coated balloon angioplasty on plaque inflammation and progression in rabbit atherosclerosis in vivo and compare to animals treated with conventional angioplasty or sham-angioplasty. Methods described in this chapter include creation of atherosclerosis in the aorta of rabbits, in vivo intravascular imaging harnessing intravascular ultrasound and the novel serial intravascular near-infrared fluorescence-optical coherence tomography (NIRF-OCT) molecular-structural imaging of inflammatory protease activity. Furthermore, methods used to elucidate mechanisms underlying the in vivo findings are described, including histological and molecular assays of resected lesions.

2.0. Materials

2.1. NIRF-OCT Imaging System and Catheter

A prototypical intravascular NIRF-OCT system consists of an OCT console, a NIRF console, a dual-channel optical rotary junction and a catheter (Figure 3).

2.2. OCT console

The OCT console is comprised of the following set up:

1. A wavelength-swept laser cavity comprises of three components: a gain medium, wavelength scanning filter and an output coupler (Figure 4) [18].

2. A semiconductor optical amplifier (SOA) used as a gain medium, typically having a > 15 dB gain at a broad wavelength range from 1240 nm to 1360 nm.

3. A polygon filter is used as a wavelength scanning filter. The wavelength tuning repetition rate is determined by the number of mirror facets (72) and the spinning frequency (52,000 rpm) of the polygon scanning mirror.

4. A 50/50 power splitting directional fiber coupler is employed as an output coupler. A wavelength-swept laser with center wavelength at 1310 nm, a bandwidth of >100 nm and average output power up to 60 mW can be achieved. This source can provide an axial resolution below 10 μm in air and ranging depth of 4.5 mm along the axial direction.

5. The wavelength swept laser has a repetition rate of 50 kHz, capable for cross-sectional imaging with frame rate of 50 Hz (1024 A-lines) or 25 Hz (2048 A-lines).

6. Light from the wavelength swept laser source is coupled into a fiber-based interferometer. A 90/10 fiber coupler (Gould Fiber Optics, 90/10 1310 nm splitter) is used to divide the light into a sample arm (90%) towards the imaging catheter and a reference arm (10%) with a fixed mirror.

7. Backscattered light from sample/reference arms are combined by a 50/50 fiber coupler (Gould Fiber Optics, 50/50 1310 nm splitter) and interferes at the interferometer.

8. The balanced detector (New Focus, 80 MHz Balanced Receiver) has a dual optical input, and before signal amplification, subtract the signal from the second input to reduce the common-mode noise, thus eliminating the autocorrelation signal and enhancing the interference signal.

9. Polarization-diverse detection is applied for eliminating artifacts that can arise from birefringent tissues such as collagen and muscle, as well as the system and optical fibers.

10. The analog signal from the photo-receivers is digitized and captured by two channels at 200 MS/s with 14-bit resolution. The data is streamed to a 5 TB RAID hard drive array.

11. A Fourier transform is utilized to reconstruct the backscattering as a function of depth (A-line) that reveals the depth-resolved tissue microstructure.

FIGURE 4:

Wavelength-swept laser (reproduced from reference [18] with permission)

2.3. NIRF console

This NIRF console consists of the following items:

1. In the NIRF console (Figure 5) [15], a multimode fiber-coupled continuous wave NIR laser diode at 750 nm (LDX-3110–750-FC) is used to excite NIR fluorophores resident within the arterial wall.

2. A narrow band laser clean-up band-pass filter (zet750/20X) is used for suppressing ambient light.

3. A specialized catheter (2.4) transmits the NIR excitation light to the tissue via a rotary junction (2.5). Then, the collected NIRF signal from the sample is detected by a photomultiplier tube.

4. The NIRF signal is next filtered by a DC coupled low-pass filter (50 kHz).

5. The filtered NIRF signal is then amplified and acquired by a data acquisition board.

FIGURE 5:

NIRF console and rotary junction (reproduced from reference [18] with permission)

2.4. NIRF-OCT catheter

The catheter has the following set up:

1. A dual-modality NIRF-OCT imaging probe (Figure 6) consists of a double-clad fiber (DCF, FUD-4305)-based catheter with outer diameter of 0.55 mm, allowing it to fit into a commercial OCT housing sheath of 2.4 French (0.79 mm) full outer diameter.

2. The DCF has a single mode fiber core (diameter: 7.8 μm; NA: 0.130) that transmits the OCT light at center frequency 1.3 μm.

3. In order to improve light collection efficiency, the inner cladding (diameter: 125 μm; NA: 0.46) of the DCF is used for propagating the NIRF excitation and emission signal.

4. At the distal end of the DCF, a ball lens (diameter: 320 μm) is equipped to focus the light on the arterial wall.

5. The ball lens can be fabricated by an automatic CO₂ laser splicing station (LZM-100) from a coreless fiber, which is spliced to DCF and melted on the other side by laser ablation to form a spherical shape with demanded diameter.

6. The ball-lens is subsequently polished to a 38-degree angle by a polishing machine to achieve side-view imaging.

7. The polished ball-lens is next assembled into a driveshaft and inserted into a protective sheath with transparent imaging window at the distal portion.

8. The lateral resolution is defined by the optical focus of the ball-lens, and typically has a focal spot of ~30 μm obtained at a working distance of 1.5 mm.

9. The maximum power output from the imaging window is set to 20 mW.

FIGURE 6:

Dual-modality NIRF-OCT catheter. (reproduced from reference [18] with permission)

2.5. Dual-channel rotary junction

The rotary junction has the following set up:

1. A dual-channel rotary junction is used to combine and transmit the NIRF and OCT light between the rotating probe and static system. Lens-based collimators are used to collimate the light from both sides, as shown in Figure 5.

2. The NIR light (excitation 750 nm) is reflected by dichroic mirror-1 (T770LPXR), and then reflected by dichroic mirror-2 (T1000LP, R: 750 nm - 860 nm, T: 1200 nm - 1400 nm) into the dual-clad fiber (DCF) catheter, while the OCT light is transmitted and directly coupled to the core of the DCF catheter.

3. The collected NIRF emission from the catheter is reflected by the same dichroic mirror (T1000LP) and filtered by an emission filter (ET810/90 nm and then focused by a plano-convex lens (LA1805-B) into the PMT.

4. The rotary junction rotates at a constant speed (e.g., 25 Hz or 50 Hz) corresponding to 2048 or 1024 A-lines per rotation.

5. Simultaneously, three-dimensional NIRF-OCT images are acquired by helical scanning of the catheter by the rotary junction.

6. The catheter pullback rate can be adjusted from 1–10 mm/s. The tradeoff for a faster pullback rate is lower spatial resolution.

7. An alternative high-resolution (HR)-NIRF-OCT imaging design uses a single mode laser diode (FIDL-30S-750X) with a high-throughput DCF rotary junction (throughput >92%) to precisely transmit the NIRF excitation light to the core of a DCF catheter. These advances enable three times higher lateral NIRF resolution (< 30 μm) compared to first-generation NIRF-OCT imaging systems (Figure 7A) [19].

8. The HR-NIRF-OCT imaging system enables high-resolution visualization of atherosclerosis inflammation and demonstrates greater spatial detail of inflammatory cathepsin activity (Figure 7B) [19], and can enhance the effectiveness of NIRF-OCT for molecular phenotyping of atherosclerosis and stent-induced vascular injury.

FIGURE 7:

High-resolution NIRF-OCT imaging. (reproduced from reference [19] with permission)

2.6. NIRF calibration and data processing

1. To achieve quantitative NIRF information from the vessel wall, the NIRF console and catheters should be calibrated by using saline-based solutions comprised of a range of concentrations (1–1000 nM) of an NIR fluorophore (i.e. Alexa Fluor 750).

2. The intravascular NIRF signal intensity decays as a function on the distance between the imaging probe and the arterial wall. NIRF distance calibration is therefore performed with a glass tube filled with NIR fluorescent dye (AF 750) with concentration of 1 μM. A plot of NIRF signal intensity vs distance (mm) is then generated. The plot is used to calculate the linear curve for fitting the distance function use for further distance correction in NIRF-OCT imaging.

3. The NIRF-OCT pullback is performed and the co-registered OCT and NIRF datasets are acquired simultaneously. First the vessel wall on the OCT image is segmented, subsequently, NIRF data is calibrated by the distance of the segmented vessel wall to the sheath surface.

4. The calibration procedure runs automatically in MATLAB code. The distance corrected NIRF intensity profile along the pullback direction, a two-dimensional NIRF map (carpet view), co-registered NIRF over the OCT frames, and three-dimensional volume image are reconstructed and displayed.

2.7. Experimental rabbit atherosclerosis model

1. New Zealand white rabbits (weight 3–4 kg). Lipid-rich lesions developed by 1% high-cholesterol diet (1% cholesterol, 5% peanut oil, Research Diets) for total of 6 weeks. This is followed by 4 weeks of normal chow.

2. Anaesthesia: intramuscular ketamine (35 mg/kg) and xylazine (5 mg/kg) with a total of 0.01 mg/kg of buprenorphine. Inhaled anaesthesia with 1–5% (v/v) of isoflurane Novaplus and supplemental carrier oxygen at 1 L/minute.

3. Electric razor.

4. Artificial tears (over the counter pharmacy).

5. Scalpel 15 blade.

6. 28G, 30G and 32G needles.

7. Dissecting scissors, forceps, 3–0 nylon sutures for slings, 3–0 vicryl suture.

8. 4F intravascular sheath (Avanti Sheath introducer, Cordis).

9. 50% saline and 50% iopamidol contrast media (Bracco).

10. 3F Fogarty embolectomy catheter (Edwards).

11. Hard E-collar for rabbit neck.

12. Saline.

13. Dissecting board.

14. Dry ice.

15. 2-methyl butane.

16. RNAlater.

2.8. Angioplasty balloons

1. 4.0 × 40 mm drug-coated balloon angioplasty device with 3.5 ug/mm² of paclitaxel coating on balloon (IN.PACT Admiral Balloon, Medtronic).

2. 4.0 × 40 mm plain angioplasty balloon device (Medtronic).

2.9. Intravascular ultrasound imaging

1. Intravascular ultrasound (IVUS) clinical console [iLab2, Polaris2 Software Imaging System (Boston Scientific)].

2. OsiriX imaging platform and Fiji ImageJ v2.0.0.

3. USB drive.

2.10. NIRF-OCT imaging

1. ProSense750 VM110 (PerkinElmer, Waltham, MA).

2. Butterfly needle 25G.

3. NIRF-OCT imaging console.

2.11. Molecular analysis of tissue

1. RNAase-free water.

2. RNeasy kit (Qiagen): Contains buffers RLT, RPE and RW1, RNAeasy spin column.

3. 1.5 mL microcentrifuge tubes.

4. Microcentrifuge.

5. 70% ethanol.

6. Nanodrop Spectrophotometer.

7. Rotar-stator homogeniser.

8. Reagents for cDNA synthesis: 10x reverse transcriptase, 10 nM dNTPs, 10x random hexamers, Multiscribe, water.

9. Thermocycler or heating block at different temperatures.

10. Basic Local Alignment Search Tool (BLAST).

11. Cathepsin B primer: F, TTCTTGCGACTCTTGGGACTTC; R, TGACGAGGATGACAGGGAACTA.

12. Cathepsin L primer: F, AGGGTCAGTGTGGTTCTTGTTG; R, TGAGATAAGCCTCCCAGTTTTC.

13. Cathepsin S primer: F, TGTTCACACTTTGCCCTATGAC; R, AGGGGCTCCATAAGGAAATAAA.

14. GAPDH primer: F, GGGGCTGGCATTGCCCTCAA; R, GGCTGGTGGTCCAGGGGTCT.

15. 96 well plate.

16. SYBR Green real-time PCR reagents.

17. QuantStudio 3 Real-Time PCR System or equivalent.

2.12. Histology

1. Optimal cutting temperature embedding medium.

3.0. Methods

3.1. Experimental atherosclerosis lesion generation in rabbits

1. Induce macrophage- and lipid-rich atherosclerosis in the aortas of male or female New Zealand white rabbits (weight 3–4 kg) by local balloon endothelial injury and high-cholesterol diet as detailed in steps below.

2. Feed that rabbits the atherogenic diet and subject the infrarenal abdominal aorta to balloon injury (e.g. with standard angioplasty balloon or a 3F Fogarty balloon which is capable of greater endothelial denudation and injury) (see Note 1).

3. Induce anesthesia with appropriate agent(s) (see 3 in section 2.7) and continue with inhalational anesthetic.

4. In preparation for intravascular NIRF-OCT via the femoral artery, shave and clean the overlying rabbit skin in a sterile manner using the electric razor.

5. Lubricate the eyes with artificial tears.

6. Make a small incision (approximately 1.5 cm) in the oblique groin over the natural lie of the common femoral artery.

7. Dissect and slung the artery with an absorbable suture and ligate the distal end.

8. Make a small arteriotomy and introduce a 4 French (F) sheath into the vessel.

9. At this point, blood can be aspirated and stored for later analysis (see Note 2).

10. Take an angiogram using 50% saline and 50% iopamidol contrast media.

11. Advance a 3F Fogarty arterial embolectomy catheter through a 4F sheath into the infrarenal aorta, approximately 15 mm below the lowest renal artery. Fill the balloon with a 0.9% saline/contrast 50:50 mix to a volume of 0.3 mL. Injure the aorta by three sequential manual pullbacks of a nominally inflated 3F Fogarty under x-ray angiographic guidance over a distance of approximately 60 mm.

12. Ligate the femoral artery and close the muscle and skin closure with 3–0 vicryl suture (see Note 3).

13. Recover the rabbits and place in a neck E-collar for one week to prevent any manual disruption of the wound.

14. Continue feeding the rabbits a 1% high-cholesterol diet until week six, followed by a normal chow diet for the remainder of the study (typically four additional weeks; Figure 8).

FIGURE 8: Experimental Study Design.

All 25 rabbits received 1.0% high-cholesterol diet (HCD) 2 weeks before aorta balloon injury, and for 4 weeks thereafter, followed by normal chow for the final 4 weeks. At 4 weeks after balloon injury, rabbits were intravenously injected with ProSense VM110 (400 nmol/kg). Rabbits underwent in vivo multimodal survival near-infrared fluorescence–optical coherence tomography (NIRF-OCT), intravascular ultrasound (IVUS), and x-ray angiography 24 h later. Then animals randomly underwent drug-coated balloon percutaneous transluminal angioplasty (DCB-PTA) (n = 10), PTA (n = 10), or sham-PTA (n = 5) therapy. Four weeks later at week 10, multimodal imaging NIRF-OCT and IVUS imaging were repeated, followed by sacrificed and ex vivo fluorescence imaging, as well as RNA and histopathological analysis. FRI = fluorescence reflectance imaging; IHC = immunohistochemistry; qPCR= quantitative polymerase chain reaction. Reproduced by permission from reference [22]

3.2. Randomization of angioplasty therapy

1. After survival intravascular NIRF-OCT and IVUS imaging (section 3.3) at baseline 6-weeks after balloon-injury, randomize rabbits to one of three treatments (see Note 4): paclitaxel drug-coated balloon percutaneous transluminal angioplasty (DCB-PTA, n=10, 3.5 ug/mm² of drug coated concentration with urea excipient) or conventional PTA (n=10) or sham PTA (n=5, placement of a balloon without inflation), followed by imaging and sacrifice 4 weeks later at week 10 (see Note 5). As explained in the note, this protocol is specific to angioplasty interventions, but the same protocol can be used when examining other types of intervention (e.g. pharmacotherapeutics, stents).

3.3. Intravascular ultrasound (IVUS) imaging

1. Acquire IVUS images of the abdominal aorta with a 40 megaHertz (MHz) clinical catheter by automated 0.5 mm/sec pullback.

2. Commence imaging from the abdominal aorta at the level of the lowest renal artery until into the common iliac vessel.

3. Perform IVUS at serial timepoints 6 weeks and 10 weeks. IVUS should be performed both pre- and post-angioplasty, and the data recorded and stored.

4. Manually co-register the IVUS datasets at 6 and 10 weeks to each other using anatomical landmarks including side branches as fiducial markers and the known pullback rates of the imaging systems, and then analyze further by manual segmentation [20], in accordance to expert consensus IVUS recommendations [21].

5. Measure the external elastic membrane (EEM) and vessel lumen every 0.4 mm from axial IVUS images for each rabbit at 6 and 10 weeks, across the area that has undergone angioplasty (total 40 mm length, 200 images per animal [100 images per timepoint]). These measurements permit calculation of the atheroma cross-sectional area (CSA) and plaque burden (PB) for each IVUS slice. All cross-sectional slices obtained can be analyzed and included in the image analysis results. On a per slice basis every 0.4 mm, atheroma progression is quantified as the change in IVUS PB (ΔPB) between 6 and 10 weeks.

6. IVUS formulae include:

   $${\text{Atheroma}}_{\text{CSA}}\left(\text{m}{\text{m}}^2\right)=\text{EE}{\text{M}}_{\text{CSA}}-\text{Lume}{\text{n}}_{\text{CSA}.}$$ 5.1.

   $$\text{PB}\left(\text{plaque}\text{burden}\right)\left(\text{\%}\right)=\left(\text{Atherom}{\text{a}}_{\text{CSA}}÷\text{EE}{\text{M}}_{\text{CSA}}\right)\times 100\text{\%}$$ 5.2.

   $$\text{ΔPB}\left(\text{plaque}\text{burden}\right)\left(\text{\%}\right)=\text{P}{\text{B}}_{\text{follow-up}}-\text{P}{\text{B}}_{\text{baseline}.}$$ 5.3.

   $$\text{PAV}\left(\text{percent}\text{atheroma}\text{volume}\right)\left(\text{\%}\right)=\left(\Sigma {\text{Atheroma}}_{\text{CSA}}÷\Sigma {\text{EEM}}_{\text{CSA}}\right)\times 100\text{\%}.$$ 5.4.

   $$\text{Percentage}\text{Maximum}\text{Stenosis}\left(\text{\%}\right)=\left(1\text{-}\left(\text{MLA}/\text{reference}\text{luminal}\text{area}\right)\right)\times 100\text{\%}.$$ 5.5.

3.3. Intravascular NIRF-OCT imaging of plaque inflammation and structure

1. To image plaque inflammation, twenty-four hours before NIRF-OCT imaging sessions, intravenously inject the rabbits inflammatory protease NIRF imaging agent ProSense750 VM110 (a quenched sensor engineered to generate fluorescence following protease activation by cathepsins B, L, or S) [15]

2. Perform automated reconstruction of dual-modal NIRF-OCT pullbacks [17]. Generate quantitative NIRF by using co-registered OCT images to correct the NIRF signal according to the distance between the catheter and the lumen surface, and express as both the NIRF concentration (nanomolar; nM) averaged over all the slices comprising the plaque and the average NIRF concentration per OCT slice at 0.4 mm intervals [15]. Calculate the change in NIRF concentration (ΔNIRF, in nM) on a per slice basis as the difference between the NIRF concentration at 6 and 10 weeks. All IVUS and NIRF image analyses should be performed in a blinded fashion in terms of group assignment of each animal (see Note 6).

3.4. Randomization of angioplasty therapy: experimental setup

1. At the baseline first imaging timepoint of six weeks (Figure 7) [22], to image arterial inflammation (cathepsin protease activity) in vivo, inject rabbits with ProSense VM110 24 hours prior to imaging (400 nmol/kg via ear vein intravenous injection)[11, 15].

2. The next day, carry out the same surgical preparation and anesthetic protocol as at week two (Section 3.1; steps 3–10) on the rabbits, and secure the femoral and carotid artery access using open surgical cutdown.

3. Next using NIRF-OCT, IVUS and x-ray angiographic guidance, inflate a single paclitaxel-coated angioplasty balloon of size 4.0×40 mm or a single plain angioplasty balloon 4.0×40 mm to a maximal pressure of 8 atmospheres (atm) for a period of 120 seconds.

4. For the control sham-PTA group, place a plain PTA balloon with the same 4.0×40 mm dimension in the aorta and inflate to 0 atm for a period of 120 seconds. The average aortic diameter in the rabbit is 3.25 mm correlating with a balloon overdistention ratio of 1.2:1 [23].

5. Only the approximate distal 40 mm of the Fogarty-balloon injured area is subject to angioplasty. Take care not to disrupt the drug-coated balloon once out of packaging (no direct handling) and to work as swiftly as possible. Follow the IFU for the use of both balloons; pre-dilate the planned region of angioplasty with a plain balloon before drug-coated balloon use.

6. Calculate the area subject to angioplasty using a combination of the angiogram, IVUS and OCT to ensure the co-registration is accurate. This involves a combination of distance from renal arteries, side branches and fiducial markers.

3.5. Sacrifice and end of study: experimental setup

1. One day prior to final 10-week imaging timepoint, inject rabbits again with ProSense VM110 24 hours prior to imaging (400 nmol/kg via ear vein intravenous injection).

2. The next day (Figure 8), perform the same surgical preparation and anesthetic protocol as in week 2 (Section 3.1; steps 3–10) with femoral and contralateral carotid access secured with surgical cut down.

3.6. Molecular analysis: aortic tissue

1. Following sacrifice, perfuse rabbit aortas with cold saline and harvest.

2. Measure the explanted aorta out to its in vivo size and pin to measuring board. Create sections every 0.5 mm, with selected tissue from 3 regions stored in RNAlater for subsequent RNA analysis. The 3 regions include : 1) normal aorta (no balloon injury), 2) PTA or DCB-PTA segment and 3) Fogarty balloon-injured region without additional PTA (see Note 7).

3. Snap-freeze the remaining tissue using dry ice and 2-methylbutane, placed in Optimal Cutting Temperature compound medium and store in −80°C for later sectioning and histopathology as desired.

3.7. Molecular analysis: RNA extraction and cDNA synthesis

1. Extract RNA by homogenization of the tissue sample in question using a rotor-stator homogenizer. Use the RNeasy mini-kit to extract RNA from the tissue.

2. Cut the RNAlater-stabilized tissue into small pieces and place in 600μL of RLT buffer provided in the kit. Homogenize until the solution is uniformly homogenous.

3. Centrifuge the solution at 12000× g for 3 minutes.

4. Transfer the resultant supernatant into a new microcentrifuge tubes, add 600 μL of 70% ethanol and pipette multiple times to mix the solution.

5. Add 700 μL of this solution to the RNeasy spin column and centrifuge for 15 seconds at 8000× g.

6. Discard the flow through, add 700 μL of buffer RW1 provided in the kit to the spin column and centrifuge for 15 seconds at 8000 ×g.

7. Discard the flow through, add 500 μL of buffer RPE to the column and centrifuge for 15 seconds at 8000× g.

8. Repeat step 7 but centrifuge the column for 2 minutes at 8000× g.

9. Change the collection tube and centrifuge further in a microcentrifuge at maximum speed for 1 minute to dry the membrane.

10. Place the spin column in a 1.5 mL collection microcentrifuge tube, add 30 μL of RNase-free water directly to the center of the column and centrifuge for 1 minute at 8000× g.

11. Place samples immediately on ice and determine RNA concentration (ng/μL) using the Nanodrop Spectrophotometer.

12. Synthesize complementary DNA (cDNA) from 1 μg of total RNA using a reaction mix with a final volume of 20 μL containing 2μL of 10x reverse transcriptase buffer, 0.8 μL of 10 nM dNTP, 2 μL of 10x random hexamers, 1 μL of Multiscribe and 14.2 μL of RNAase-free water and mix.

13. Perform reaction for 10 minutes at 25°C, 120 minutes at 37°C, 5 minutes at 85°C and hold at 4 °C to produce cDNA.

14. Dilute the cDNA to 25 ng/μL and store at −20 °C until further use.

15. Assess cathepsin B, cathepsin L and cathepsin S expression using SYBR Green real-time PCR using primer sequences designed using appropriate software (e.g. Primer Express™, ver. 3.0.1, Thermo Fisher Scientific). Verify gene specificity of all primer sequences by BLAST (Basic Local Alignment Search Tool) searches.

16. Perform q-RT-PCR using the QuantStudio 3 Real-Time PCR System (ThermoFisher Scientific) in a 96-well plate format using following cycle parameters: initial denaturation at 95°C for 4 minutes followed by 40-cycles of denaturation at 95°C for 10 seconds, annealing at 60°C for 30 seconds and elongation at 72°C for 5 to 30 seconds.

17. Analyze data using the QuantStudio 3 Sequence Detection System 2.2.1 software (v20040907–2, 2004, ThermoFisher Scientific) with the detection threshold set manually at 0.05 for all the assays. Normalize all transcripts to GAPDH housekeeping gene) and determine relative gene expression by using the comparative Ct method (2^-ΔΔCt).

FIGURE 9: Experimental Study Design.

Paclitaxel drug-coated balloons (DCBs) reduce restenosis, but their overall safety has recently raised concerns. The work highlighted in this chapter hypothesized that DCBs could lessen inflammation and reduce plaque progression. Using 25 rabbits with cholesterol feeding- and balloon injury-induced lesions, DCB-percutaneous transluminal angioplasty (PTA), plain PTA, or sham-PTA (balloon inserted without inflation) was investigated using serial intravascular near-infrared fluorescence-optical coherence tomography and serial intravascular ultrasound. In these experiments, DCB-PTA reduced inflammation and plaque burden in non-obstructive lesions compared with PTA or sham-PTA. These findings indicated the potential for DCBs to serve safely as regional anti-atherosclerosis therapy (as evidenced in Figure 5). Reproduced by permission from reference [22]