Assessment of coronary microcirculation alterations in a porcine model of no-reflow using ultrasound localization microscopy: a proof of concept study.

BACKGROUND: Coronary microvascular obstruction also known as no-reflow phenomenon is a major issue during myocardial infarction that bears important prognostic implications. Alterations of the microvascular network remains however challenging to assess as there is no imaging modality in the clinics that can image directly the coronary microvascular vessels. Ultrasound Localization Microscopy (ULM) imaging was recently introduced to map microvascular flows at high spatial resolution (∼10 µm). In this study, we developed an approach to image alterations of the microvascular coronary flow in ex vivo perfused swine hearts. METHODS: A porcine model of myocardial ischemia-reperfusion was used to obtain microvascular coronary alterations and no-reflow. Four female hearts with myocardial infarction in addition to 6 controls were explanted and placed immediately in a dedicated preservation and perfusion box manufactured for ultrasound imaging. Microbubbles (MB) were injected into the vasculature to perform Ultrasound Localization Microscopy (ULM) imaging and a linear ultrasound probe mounted on a motorized device was used to scan the heart on multiple slices. The coronary microvascular anatomy and flow velocity was reconstructed using dedicated ULM algorithms and analyzed quantitatively. FINDINGS: We were able to image the coronary microcirculation of ex vivo swine hearts at a resolution of tens of microns and measure flow velocities ranging from 10 mm/s in arterioles up to more than 200 mm/s in epicardial arteries. Under different aortic perfusion pressures, we measured in large arteries of a subset of control hearts an increase of flow velocity from 31 ± 11 mm/s at 87 mmHg to 47 ± 17 mm/s at 132 mmHg (N = 3 hearts, P < 0.05). This increase was compared with a control measurement with a flowmeter in the aorta. We also compared 6 control hearts to 4 hearts in which no-reflow was induced by the occlusion and reperfusion of a coronary artery. Using average MB velocity and average density of MB per unit of surface as two ULM quantitative markers of perfusion, we were able to detect areas of coronary no-reflow in good agreement with a control anatomical pathology analysis of the cardiac tissue. In the no-reflow zone, we measured an average perfusion of 204 ± 305 MB/mm2 compared to 3182 ± 1302 MB/mm2 in the surrounding re-perfused area. INTERPRETATION: We demonstrated this approach can directly image and quantify coronary microvascular obstruction and no-reflow on large mammal perfused hearts. This is a first step for noninvasive, quantitative and affordable assessment of the coronary microcirculation function and particularly coronary microvascular anatomy in the infarcted heart. This approach has the potential to be extended to other clinical situations characterized by microvascular dysfunction. FUNDING: This study was supported by the French National Research Agency (ANR) under ANR-21-CE19-0002 grant agreement.

Assessment of Takayasu's arteritis activity by ultrasound localization microscopy.

BACKGROUND: Ultrasound localization microscopy (ULM) based on ultrafast ultrasound imaging of circulating microbubbles (MB) can image microvascular blood flows in vivo up to the micron scale. Takayasu arteritis (TA) has an increased vascularisation of the thickened arterial wall when active. We aimed to perform vasa vasorum ULM of the carotid wall and demonstrate that ULM can provide imaging markers to assess the TA activity. METHODS: Patients with TA were consecutively included with assessment of activity by the National Institute of Health criteria: 5 had active TA (median age 35.8 [24.5-46.0] years) and 11 had quiescent TA (37.2 [31.7-47.3] years). ULM was performed using a 6.4 MHz probe and a dedicated imaging sequence (plane waves with 8 angles, frame rate 500 Hz), coupled with the intravenous injection of MB. Individual MB were localised at a subwavelength scale then tracked, allowing the reconstruction of the vasa vasorum flow anatomy and velocity. FINDINGS: ULM allowed to show microvessels and to measure their flow velocity within the arterial wall. The number of MB detected per second in the wall was 121 [80-146] in active cases vs. 10 [6-15] in quiescent cases (p = 0.0005), with a mean velocity of 40.5 [39.0-42.9] mm.s-1 in active cases. INTERPRETATION: ULM allows visualisation of microvessels within the thickened carotid wall in TA, with significantly greater MB density in active cases. ULM provides a precise visualisation in vivo of the vasa vasorum and gives access to the arterial wall vascularisation quantification. FUNDING: French Society of Cardiology. ART (Technological Research Accelerator) biomedical ultrasound program of INSERM, France.

Quantitative stiffness assessment of cardiac grafts using ultrasound in a porcine model: A tissue biomarker for heart transplantation.

BACKGROUND: Heart transplantation is the definitive treatment for many cardiovascular diseases. However, no ideal approach is established to evaluate heart grafts and it mostly relies on qualitative interpretation of surgeon based on the organ aspect including anatomy, color and manual palpation. In this study we propose to assess quantitatively the Shear Wave Velocity (SWV) using ultrasound as a biomarker of cardiac viability on a porcine model. METHODS: The SWV was assessed quantitatively using a clinical ultrasound elastography device (Aixplorer, Supersonics Imagine, France) linked to a robotic motorized arm (UR3, Universal Robots, Denmark) and the elastic anisotropy was obtained using a custom ultrasound research system. SWV was evaluated as function of time in two porcine heart model during 20h at controlled temperature (4°C). One control group (N = 8) with the heart removed and arrested by cold cardioplegia and immerged in a preservation solution. One ischemic group (N = 6) with the organ harvested after 30 min of in situ warm ischemia, to mimic a donation after cardiac death. Hearts graft were revived at two preservation times, at 4 h (N = 11) and 20 h (N = 10) and the parameters of the cardiac function evaluated. FINDINGS: On control hearts, SWV remained unchanged during the 4h of preservation. SWV increased significantly between 4 and 20h. For the ischemic group, SWV was found higher after 4h (3.04 +/- 0.69 vs 1.69+/-0.19 m/s, p = 0.007) and 20h (4.77+/-1.22 m/s vs 3.40+/-0.75 m/s, p = 0.034) of preservation with significant differences. A good correlation between SWV and cardiac function index was found (r2=0.88) and manual palpation score (r2=0.81). INTERPRETATION: Myocardial stiffness increase was quantified as a function of preservation time and harvesting conditions. The correlation between SWV and cardiac function index suggests that SWV could be used as a marker of graft viability. This technique may be transposed to clinical transplantation for assessing the graft viability during transplantation process. FUNDING: FRM PME20170637799, Agence Biomédecine AOR Greffe 2017, ANR-18-CE18-0015.

Coronary Flow Assessment Using 3-Dimensional Ultrafast Ultrasound Localization Microscopy.

BACKGROUND: Direct assessment of the coronary microcirculation has long been hampered by the limited spatial and temporal resolutions of cardiac imaging modalities. OBJECTIVES: The purpose of this study was to demonstrate 3-dimensional (3D) coronary ultrasound localization microscopy (CorULM) of the whole heart beyond the acoustic diffraction limit (<20 µm resolution) at ultrafast frame rate (>1000 images/s). METHODS: CorULM was performed in isolated beating rat hearts (N = 6) with ultrasound contrast agents (Sonovue, Bracco), using an ultrasonic matrix transducer connected to a high channel-count ultrafast electronics. We assessed the 3D coronary microvascular anatomy, flow velocity, and flow rate of beating hearts under normal conditions, during vasodilator adenosine infusion, and during coronary occlusion. The coronary vasculature was compared with micro-computed tomography performed on the fixed heart. In vivo transthoracic CorULM was eventually assessed on anaesthetized rats (N = 3). RESULTS: CorULM enables the 3D visualization of the coronary vasculature in beating hearts at a scale down to microvascular structures (<20 µm resolution). Absolute flow velocity estimates range from 10 mm/s in tiny arterioles up to more than 300 mm/s in large arteries. Fitting to a power law, the flow rate-radius relationship provides an exponent of 2.61 (r2 = 0.96; P < 0.001), which is consistent with theoretical predictions and experimental validations of scaling laws in vascular trees. A 2-fold increase of the microvascular coronary flow rate is found in response to adenosine, which is in good agreement with the overall perfusion flow rate measured in the aorta (control measurement) that increased from 8.80 ± 1.03 mL/min to 16.54 ± 2.35 mL/min (P < 0.001). The feasibility of CorULM was demonstrated in vivo for N = 3 rats. CONCLUSIONS: CorULM provides unprecedented insights into the anatomy and function of coronary arteries at the microvasculature level in beating hearts. This new technology is highly translational and has the potential to become a major tool for the clinical investigation of the coronary microcirculation.