Contrast-Enhanced Ultrasound of Brain Perfusion in Cardiopulmonary Resuscitation

Abstract

To evaluate the feasibility and potential utility of contrast-enhanced ultrasound for real-time imaging of whole-brain perfusion during cardiopulmonary resuscitation, cardiac arrest was induced in eight seven-week old 10-kg piglets (Sus scrofa domesticus). Contrast-enhanced ultrasound was performed through a parietal cranial window in the coronal plane visualizing the thalami during hemodynamic-directed cardiopulmonary resuscitation. Whole-brain mean and maximum pixel intensities in each slice during resuscitation were calculated. Piglets were monitored for 24 hours post-arrest. Seven piglets achieved return of spontaneous circulation and six survived to 24 hours. Of the six surviving piglets, two piglets demonstrated greater intra-CPR brain enhancement at maximum 73.2% and 42.1% and mean 36.7% and 31.9% enhancement above background, respectively, compared to maximum 5.8%, 22.9%, 6.0%, and 26.6% and mean 5.1%, 8.9%, 2.9%, and 6.6% above background, respectively, in the other four. Intra-CPR average MAPs were similar between all six surviving piglets. One piglet achieved return of spontaneous circulation but expired 10 minutes later with enhancement maximum 45.2% and mean 18.9% enhancement above background. The final piglet did not achieve ROSC and exhibited minimal enhancement at maximum 2.8% and mean 0.9% enhancement above background. Contrast-enhanced ultrasound can detect brain perfusion during cardiopulmonary resuscitation, identifying a spectrum of cerebral blood flow responses in the brain despite similar systemic hemodynamics. This novel application can form the basis for future large animal model studies and eventually human clinical studies to further explore the neurologic implications of cerebral blood flow responses during resuscitation and stimulate novel strategies for optimizing brain perfusion restoration.

Introduction

Maintaining brain perfusion is critical during cardiopulmonary resuscitation (CPR) in neonates and infants. American Heart Association (AHA) CPR compression rate and depth guidelines are standardized by age group without consideration of hemodynamic parameters or measures of end-organ perfusion (1). Hemodynamic-directed CPR uses systemic, such as mean arterial pressure (MAP) or systolic blood pressure, and organ-specific, such as coronary perfusion pressure (CPP), measures of perfusion to guide the depth of compressions and the administration of vasopressors and inotropes during resuscitation. This approach to CPR reduces mortality and neurologic morbidity from cardiac arrest (2–7).

While hemodynamic-directed CPR has clear benefits over standard AHA CPR guidelines, measurement of brain perfusion during CPR may allow further optimization of resuscitation efforts and additional improvements in neurologic morbidity after cardiac arrest. Various modalities and tools for evaluating brain perfusion, including microspheres, near-infrared spectroscopy, transcranial Doppler, and cerebral perfusion pressure calculations, have been proposed to guide and individualize CPR (8). However, the limited area of brain tissue that can be evaluated by most of these approaches as well as their susceptibility to motion degradation hinder their usefulness. Although it is currently an off-label use of ultrasound contrast, contrast-enhanced ultrasound (CEUS), which employs microbubble contrast agents, can visualize microvasculature in whole-brain slices. More importantly, diagnostic interpretation is feasible even with motion degradation. A growing body of literature is demonstrating the utility of CEUS for depicting brain perfusion in neonates (9–12).

We employed a porcine model of cardiac arrest and resuscitation to evaluate the feasibility and potential value of CEUS during CPR. Our goal was to identify trends in brain enhancement and mean arterial pressure (MAP) during CPR as well as return of spontaneous circulation (ROSC) and survival to 24 hours post-ROSC as short-term outcomes. We hypothesized that systemic hemodynamic response suggested by MAP during hemodynamic-directed CPR would not predict brain perfusion restoration detected by CEUS in our piglet model.

Materials and Methods

Subjects

The study was performed under a protocol approved by the Institutional Animal Care and Use Committee (IACUC) of the Children’s Hospital of Philadelphia. Seven-week old piglets (Sus scrofa domesticus; n=8; weight 10.3 kg ± 0.83) were acquired from Meck Swine, LLC.

Contrast-Enhanced Ultrasound

A right parietal cranial window measuring 3 × 5 cm was created in each piglet to allow for CEUS of a whole-brain slice imaging the maximum transverse diameter of the bilateral thalami. Scans were performed on a Siemens Acuson Sequoia (Siemens Healthineers AG, Erlangen, Germany) using a 9EC4 (frequency range 2.9 – 8.1 MHz, curved array) transducer. The same operator performed all ultrasound imaging with consistent imaging parameters across all studies. An oblique coronal plane maximizing the visualized area of the bilateral thalami was contained within the imaging plane. Sulfur hexafluoride microbubbles (Lumason, Bracco, Milan, Italy) was prepared as per the product label and administered as a non-dilute steady infusion at 0.6 ml/min using an infusion pump. Imaging depth was set to 5 cm, with the focal zone at the deepest allowable depth (between 4 and 5 cm), contrast gain was set to 0 dB, dynamic range to 65 dB and the mechanical index (MI) was 0.08. Cine clips were continuously acquired in 1 to 2 minute blocks for the entire duration of CPR using burst-replenishment technique.

Anesthesia, Induction of Cardiac Arrest, and Resuscitation

Piglets were induced with 20 mg/kg intramuscular ketamine, intubated, and maintained on mechanical ventilation and inhaled isoflurane (0.3–1.0%). Arterial and venous femoral line catheters were placed in the experimental piglets. Millar pressure transducers were introduced into the lumen of the femoral line catheters to capture arterial blood pressure and right atrial pressure, which were used during CPR to monitor hemodynamic changes and enable vasopressor delivery during compressions. A Swan-Ganz catheter was floated into the pulmonary artery to identify changes in cardiac output post-arrest and allow for the delivery microbubbles directly into the central cardiovascular system.

After baseline CEUS of the whole-brain slice, each experimental piglet’s endotracheal tube was clamped and the animal was disconnected from the ventilator for 7 minutes. At 7 minutes, a current was delivered across the apex of the heart to induce ventricular fibrillation (VF). After confirming VF, compressions began with a goal rate of 100/min. The animal was ventilated at 10 breaths/min with 100% O₂ for the duration of CPR. Compression depth was directed a target systolic blood pressure of 90 mmHg was used. A maximum depth of compression was calculated using 1/3 anterior-posterior chest measurement. Vasopressors were delivered during CPR based on the CPP calculated by subtracting the right atrial pressure from the arterial line pressure. If the CPP fell below 20 mmHg for 3 consecutive compressions, a vasopressor was delivered. After 2 minutes of compressions, the animal was eligible for the first dose of epinephrine (200 mcg). If after one minute of further compressions the animal still required a vasopressor, a second dose of epinephrine was delivered (200 mcg). One minute after the second dose of epinephrine, if needed, the animal would receive 4 units of vasopressin. The animal would then not be eligible for vasopressors for two minutes following the dose of vasopressin. After 10 minutes of compressions, a 50 J defibrillator shock was delivered aiming for return of spontaneous circulation (ROSC). If ROSC was not achieved, the animal was euthanized using intravenous potassium chloride. CEUS was captured continuously during the resuscitation.

If ROSC was achieved, the animal entered the post-ROSC care pathway for 24 hours. A minimum MAP goal of 45 mmHg was maintained using an infusion of epinephrine (0.01 mcg/kg/min – 0.5 mcg/kg/min) or with the administration of intravenous fluids (0.9% Normal Saline, 10 mL/kg) based on right atrial pressure levels. Ventilation was adjusted to maintain an end-tidal CO₂ (EtCO₂) of 38–42 mmHg and oxygen saturation (SpO₂) of 94%–100%. After four hours of post-ROSC management, all incisions were closed and dressed, the animal was weaned from the ventilator, extubated, and recovered from anesthesia. At 24 hours, the piglets were euthanized with intravenous potassium chloride.

Analysis

Image analysis of CEUS clips was performed offline using ImageJ (13). The CEUS cine clips were imported into ImageJ in the Digital Imaging and Communications in Medicine (DICOM) format to maintain image quality. The visualized brain was segmented, and the maximum and mean intensity within the whole brain was evaluated for the duration of the cine clip. In one piglet that did not achieve ROSC, intra-CPR CEUS was collected for only 4 minutes.

Results

Taking into consideration achievement of ROSC, survival to 24 hours, and brain enhancement, four major patterns were identified in our cohort of eight piglets. Two piglets demonstrated whole-brain enhancement during CPR, achieved ROSC, and survived to 24 hours (Figures 1 and 2). In these piglets, whole-brain enhancement reached maximum 73.2% and 42.1% and mean 36.7% and 31.9% above background, respectively. Average MAPs during CPR were 33.7 (range 17.5–42.9) and 35.0 (9.9–96.8) mmHg, respectively.

Figure 1.

Intra-CPR brain perfusion in a surviving pig. Whole brain CEUS performed at baseline and during asphyxia, CPR, and return of spontaneous circulation (ROSC). Expected near absent flow during asphyxia and progressive increase in cerebral blood flow noted during CPR, followed by reperfusion at ROSC in a surviving pig.

Figure 2.

Time-intensity curves showing mean voxel intensity on brain CEUS over time relative to mean arterial pressure (MAP), intracranial pressure (ICP), and vasopressor administrations during cardiopulmonary resuscitation (CPR) for two piglets that achieved return of spontaneous circulation (ROSC), survived for 24 hours, and demonstrated periodic increases in brain enhancement during CPR.

Four piglets achieved ROSC and survived 24 hours but exhibited minimal brain enhancement during CPR (Figure 3). These piglets’ brains attained maximum 5.8%, 22.9%, 6.0%, and 26.6% and mean 5.1%, 8.9%, 2.9%, and 6.6% enhancement above background, respectively. Average MAPs during CPR were similar to the group with greater intra-CPR brain enhancement at 37.9 (19.8–49.2), 36.2 (15.5–48.5), 28.7 (11.6–45.1), and 37.2 (15.5–51.8) mmHg, respectively.

Figure 3.

Time-intensity curves showing mean voxel intensity on brain CEUS over time relative to mean arterial pressure (MAP), intracranial pressure (ICP), and vasopressor administrations during cardiopulmonary resuscitation (CPR) for four piglets that achieved return of spontaneous circulation (ROSC) and survived for 24 hours, but demonstrated only minimal brain enhancement during CPR.

One piglet exhibiting intermediate levels of whole-brain enhancement during CPR achieved ROSC but expired 10 minutes later (Figure 4). Brain enhancement was maximum 45.2% and mean 18.9% enhancement above background. Average MAP for this piglet was higher than for the other piglets at 49.5 mmHg with a broader range of MAPs during CPR (3.1–144.0 mmHg).

Figure 4.

Time-intensity curve showing mean voxel intensity on brain CEUS over time relative to mean arterial pressure (MAP), intracranial pressure (ICP), and vasopressor administrations during cardiopulmonary resuscitation (CPR) for the piglet that achieved return of spontaneous circulation (ROSC) but did survive for 24 hours. Only minimal brain enhancement was seen during CPR.

The final piglet did not achieve ROSC. During CPR, the piglet exhibited minimal enhancement at maximum 2.8% and mean 0.9% enhancement above background (Figures 5 and 6). Average MAP during CPR was similar to the surviving piglets at 28.4 (11.1–42.2) mmHg.

Figure 5.

Time-intensity curve showing mean voxel intensity on brain CEUS over time relative to mean arterial pressure (MAP), intracranial pressure (ICP), and vasopressor administrations during cardiopulmonary resuscitation (CPR) for a piglet did not achieve return of spontaneous circulation (ROSC). None to minimal brain enhancement was seen during CPR.

Figure 6.

Intra-CPR brain perfusion in a non-surviving pig. Whole brain CEUS as in Figure 1, except the progressive restoration of cerebral blood flow was not observed in this pig that did not achieve ROSC.

The datasets generated during and/or analyzed during the current study are not publicly available, but are available from the corresponding author on reasonable request.

Discussion

While the main goal of CPR is to preserve the heart and brain, brain injury after CPR is the most common cause of death after cardiac arrest (14–15). Current CPR guidelines from the AHA are standardized with minimal focus on individualization of treatment and with no direct indicators of brain integrity. Here, we describe a pilot study demonstrating the feasibility of contrast-enhanced ultrasound to monitor whole-brain perfusion during CPR, which could allow for the evaluation of modified CPR techniques to optimize brain perfusion.

In our cohort of piglets, divergent responses on CEUS relative to hemodynamic responses and short-term outcomes highlight the potential value of CEUS during CPR. One piglet had intermediate levels of brain enhancement and achieved ROSC, but only survived 10 minutes post-ROSC. Another piglet had minimal enhancement and did not achieve ROSC. Six total piglets achieved ROSC and survived 24 hours after cardiac arrest and CPR, but on CEUS, two of those six piglets had greater intra-CPR brain enhancement than the other surviving four. Despite these disparate sonographic findings, all piglets had similar average MAPs during CPR. These findings cast doubt on the correlation between systemic hemodynamic parameters and brain reperfusion and underline whole-brain perfusion imaging’s potential role in resuscitation.

The reason for these disparate perfusion responses is unclear. One study found similarly divergent responses in cerebral blood flow during CPR after a period of cardiac arrest, a finding the authors primarily attributed to variability in intravascular thrombosis, the “no-reflow” phenomenon (16). In the no-reflow phenomenon, microvascular perfusion does not immediately resume in some brain regions after restoration of large-vessel flow or hypoxia, possibly because of hypotension, increased blood viscosity, microemboli, edematous perivascular cells, or increased intracranial pressure (17–21). No-reflow typically results in heterogeneous microvascular reperfusion, but the initial study describing the phenomenon, which used rabbits as its model of ischemia, had multiple subjects with up to 95% involvement of brain mass (17). The rabbits with 95% involvement had been exposed to 15 minutes of brain hypoperfusion compared to 7 minutes of asphyxia in our study, but the previous study demonstrating divergent cerebral blood flow during CPR after cardiac arrest only used 6 minutes of asphyxia in dogs (16), suggesting possible inter-species differences in responses to hypoxia.

Because of their widespread availability, hemodynamic value-driven protocols have been a common avenue of investigation for optimizing cerebral blood flow during CPR (2–7). Studies using a minimum coronary perfusion pressure as a target for guiding depth of chest compressions have shown a significant improvement in cerebral perfusion pressure over standard CPR (2–7). Invasive intracranial pressure and central arterial pressure monitoring has been used to identify techniques, such as “head up” CPR, for improving cerebral perfusion pressure during resuscitation (4). Similarly, MAP has been proposed as a convenient value for optimizing brain perfusion (22). However, such techniques provide limited information regarding perfusion in the brain, and hemodynamic values are not always reflective of cerebral blood flow because of cerebral autoregulation and its potential dysfunction.

The lack of a bedside tool for detecting brain perfusion is a known impediment to evaluating and optimizing cerebral blood flow during CPR (8). Microspheres, laser Doppler, autoradiography, transcranial Doppler ultrasound, near-infrared spectroscopy, and other techniques have been considered and investigated for this purpose, but many of them, including microspheres and autoradiography, are invasive. The non-invasive options, such as transcranial Doppler ultrasound, laser Doppler, and near-infrared spectroscopy, evaluate a small sample of the brain and its diagnostic quality markedly degraded by motion (i.e. chest compressions). Furthermore, transcranial Doppler evaluates only large-vessel and high flow, which do not consistently correlate with flow in the cerebral microvasculature (23). In contrast, CEUS enables assessment of whole brain perfusion even during active chest compressions in cardiopulmonary resuscitation. A manual probe placement would be desired due to the marked motion degradation and the need to adjust probe contact real-time.

There were limitations to our study. First, we had a limited sample size, but our study represents a proof-of-concept of the feasibility of CEUS during CPR. Additionally, intravenous and inhaled anesthetics can alter cerebral physiology, and our results may therefore not be fully generalizable to cases of out-of-hospital cardiac arrest. Furthermore, we did not observe the pigs beyond 24 hours, limiting our ability to test longer-term outcomes.

Despite these limitations, this study demonstrates that CEUS can detect brain perfusion during CPR in cardiac arrest. Furthermore, our findings suggest a spectrum of cerebral blood flow restoration responses during CPR and ROSC. This can form the basis for larger animal model studies and eventually human clinical studies to examine the potential prognostic implications of our current findings and guide future innovations in resuscitation science. While our model emulated an open fontanelle with a cranial window for imaging, future research may extend to children and adults with the temporal bone as an acoustic window.