Summary
Percutaneous transendocardial injections of therapeutic agents into the myocardium may not always be effective. We used an animal model for assessing the efficacy of the injections using linoleic acid as a testing agent. Efficacious delivery into the myocardium of a beating heart was indicated by rapidly developed local myocardial necrosis and wall motion abnormalities using echocardiography. We employed this experimental model to test our innovative technology, an acoustically active injection catheter. The Doppler ultrasound-guided acoustically active injection catheter effectively delivers the substance to the myocardium but needs further technical improvements to minimize an unwanted systemic distribution of the agent.
Keywords: experimental model, acoustic catheter, image guidance, ultrasound, color flow Doppler
Introduction
Efficacy is a prerequisite for the success of novel substances used in cardiac regenerative and gene therapies. Efficacy means that an administered substance achieves its desired effect in a controlled setting.1 In a clinical setting, the desired effects include significant improvement of cardiac function and reduction of symptoms. These effects are influenced by both the therapeutic substance and delivery technique.2 We focus here on the latter.
A transendocardial route has been used to administer therapeutic substances into the cardiac muscle using an injection catheter placed within the beating left ventricle.3,4 Accurate targeting of the myocardial lesion and reliable therapeutic substance delivery are the key elements for achieving transendocardial injections’ efficacy. To this end, we have developed a functional prototype of an acoustically active catheter for transendocardial injections in preclinical studies.5,6 The catheter acoustically interacts with a commercially available echocardiography system operating in a color Doppler mode. The interaction produces an instantaneous color marker for spatial guidance of a catheter tip and a needle tip extended from the catheter into the cardiac muscle.7 We found that the acoustically active catheter could deliver a testing agent into the cardiac muscle with high spatial accuracy.8 As a next step, we aimed to test the efficacy of our innovative technology; specifically, we tested whether the injection can actually achieve targeted wall motion abnormality and whether the substance leaks systemically instead of being efficiently delivered into the myocardium.
We used linoleic acid as a testing agent to evaluate the acoustically-guided transendocardial injections’ efficacy in beating hearts in anesthetized pigs. Linoleic acid is a cardiotoxic substance that induces, within minutes, morphological and functional myocardial changes in a targeted area as long as the substance is locally delivered. This preclinical model is intended for novel device development to ensure the transendocardial injection catheter delivers the agent into the cardiac muscle effectively and safely. Conventionally, the successful injection is evaluated by using the color stain deposit by pathological observations in pre-clinical models, which cannot be assessed during the procedure or immediately after while the animal is alive. In comparison, our animal model is beneficial in the ability of real-time evaluation of successful injection, using myocardial texture changes or wall motion abnormalities detected by echocardiography (a desirable effect showing localized linoleic acid toxicity) or systemic hemodynamic reaction (an undesirable effect induced by leaking of the testing substance).
Materials and Methods
General Methods
Animal Procedures
The animal studies were approved by the Mayo Clinic Institutional Animal Care and Use Committee. We used 18 male domestic pigs, weighing 58 to 81 kg (mean ± SD, 67 ± 7kg). The experimental procedures are described in the previous article.6 Each animal was initially anesthetized with intramuscular tiletamine and zolazepam (Telazol), xylazine, and glycopyrrolate. Each animal was then intubated and connected to a mechanical ventilator (Narkomed 6000, Draeger, Telford, PA, USA). Maintenance of anesthesia was achieved via inhalation of 1-3% isoflurane. Intravenous fentanyl was used for analgesia as needed. The Seldinger technique9 was used to access the femoral artery and vein to monitor central arterial pressure by a high-fidelity catheter (Millar Instruments, Houston, TX) and administer drugs and fluids, respectively. The median sternotomy was performed to open the chest, and the heart was placed on a pericardial cradle.
Echocardiographic Scans of the Left Ventricle
The Vivid 7 ultrasound system and the M4S phased-array transducer were used, including the interposed attenuative pad, also for scans of the left ventricle. The transducer was operating at 1.7/3.4 MHz (fundamental/harmonic) frequency. We obtained at least 3 cardiac cycles of left ventricular motion in the standard apical 4-, 3-, and 2-chamber views and the basal, mid, and apical short-axis views, all at baseline and in the incremental stages of the study. Scans were obtained with >50 frames/second to provide proper temporal resolution for myocardial strain analysis.
Sonomicrometry of Local Myocardial Motion
Sonomicrometry is a technique that uses the ultrasound time-of-flight principle for measuring an instantaneous distance between two piezoelectric crystals. In this study, we subepicardially implanted two pairs of 2-mm round crystals: one pair into the mid-anterior myocardium (testing region) and the other pair into the mid-posterior wall of the left ventricle (reference region). The crystals were connected to a digital sonomicrometry system (TRX Series 16, Sonometrics, London, Canada). The system locally monitored myocardial contraction and relaxation by generating traces of the distance between two crystals in the testing and reference regions.
Study Protocol
Echocardiographic scans and sonomicrometry traces were analyzed in each instrumented animal before the linoleic acid injection and then at 15 minutes and 60 minutes after the injection. Additional traces were analyzed if a noticeable change in local motion was observed on the sonomicrometry monitor. At the end of the experiment, the animal was euthanized and the heart was dissected for autopsy and histological analysis. Left ventricular global and regional function analyses were performed offline from the data recorded by echocardiography and sonomicrometry.
Intramyocardial Injection of Linoleic Acid
Linoleic acid (L1376, SIGMA, Saint Louis, MO, USA) is an unsaturated fatty acid, which triggers necrotic cell death and inflammatory reactions.10
Preparative Study Using External Epicardial Injection
As a preparative study to evaluate the effect of linoleic acid and its dosage, we performed epicardial injections of linoleic acid under direct visualization in 9 animals. We used 20-gauge hypodermic needle (2.5 cm needle length, BD PrecisionGlide Needle), and injected 1.2-2.0 mL (0.5 mL x 4 points in 2 animals, 0.3 mL x 4 points in 2 animals, 0.6 mL x 2 points in 4 animals) into the anterior wall.
Main Study with a Transendocardial Injection Catheter
Acoustically Active Injection Catheter (AAIC)
We customized an 8.8F Vado® Bidirectional Steerable Sheath manufactured by Abbot (Kalila Medical, Campbell, CA, USA). The AAIC design is in Figure 1 and follows a similar methodology detailed previously.7 Briefly, the Vado sheath’s tip was fitted with a ring-shaped piezoelectric crystal (Sonometrics, London, ON Canada; 3 mm outer diameter, 1 mm inner diameter; Figure 1a). The crystal was soldered to a 32-gauge cable, passed through the Vado sheath’s inner lumen. The cable was exteriorized through a small opening in the Vado sheath handle and connected to a signal generator (Agilent 33500 B, Agilent Technologies, Loveland, CO, USA). The generator was set to produce a square-wave signal with a frequency ranging from 90 to 110 kHz at amplitudes from 0.5 to 5.0 Volts peak to peak. This signal was fed into the navigation crystal to produce a color marker for the identification and guidance of the sheath tip by Doppler ultrasound imaging.
Figure 1: Acoustically active injection catheter prototype.
(a) A commercially available steerable sheath is customized with a navigation ring-shaped piezoelectric crystal at its tip (insert; x). An injection needle is attached to an internal catheter and extended through the opening in the navigation crystal. The needle is fitted at its tip with a small crystal (insert; y). A small portion of the 20-gauge needle wall close to the tip is filed off. A 0.5-mm piezoelectric crystal with a soldered connection wire is embedded within the opening in the needle wall with epoxy glue. The outside portion of the crystal is carefully shaved off by sandpaper to assure a smooth needle surface. The connection wire, 40 gauge (hair-thin), is placed through the lumen of the needle and the catheter and exteriorized for connecting to a signal generator. Intraluminal wires (of a ring-shaped crystal and a small crystal) and external connectors (arrowhead) deliver electrical signals from a waveform generator to both crystals at the tip of the sheath and the needle. (b) The acoustically active catheter is bent inside the left ventricle, and an elongated red Doppler marker crosses the catheter at its tip. (c) An elongated blue Doppler marker crosses the needle at the level of its tip, thus instantaneously identifying the depth of transendocardial injection.
We equipped the lumen of the Vado sheath with an internal injection catheter. The internal catheter was built in-house by attaching a 20-gauge hypodermic needle to a distal end of a 5-F angiographic catheter (Cordis Europa, Roden, Netherlands) to produce an 80-cm long straight injection catheter. The proximal end of the internal catheter extended through an integrated entry within the Vado sheath handle. By gradually pulling or pushing on the exteriorized end of the internal catheter, the needle was retracted or exposed, respectively, through the navigation ring-shaped crystal. The distal end of the needle was equipped with a miniature (0.5 mm) piezoelectric crystal (Figure 1a) and connected with a 40-gauge cable, placed within the lumen of the internal injection catheter, with the signal generator. This crystal served for the guidance of the needle into a visually approximated mid-layer of the myocardial wall.
The mechanism of the color marker was explained in our previous publication.11 Briefly, new signals are made by a crystal’s interaction with an incoming Doppler signal which impinges on the crystal. Among the new signals, there is a Doppler shift signal that is received by the ultrasound probe and interpreted as local motion, which is displayed as an instantaneous color marker. If the crystal at the needle and catheter tips are stimulated by separately controlled generators, different Doppler shift frequencies can be generated by the vibrating crystals and result in different color markers.
AAIC Insertion and Navigation by Color Doppler Ultrasound
We inserted a 14-F arterial hemostasis introducer sheath (Fast-Cath™, St. Jude Medical, Plymouth, MN, USA) from the right carotid artery with the Seldinger method in 8 animals and with surgical arterial cutdown in 1 animal. AAIC was inserted through the introducer sheath, and the crystal at its tip was used for navigation with ultrasound color Doppler imaging (Vivid 7, GE Healthcare, Milwaukee, WI, USA) using an M4S phased-array transducer operating at a 2-MHz Doppler frequency. We used the transducer epicardially with an interposed rubber pad to simulate human chest attenuation in our open-chest pig model.12 AAIC was navigated from the carotid artery to the aortic arch and then via the aortic valve into the left ventricle using ultrasound imaging guidance.
Inside the left ventricle, we navigated the AAIC tip toward the endocardium of the anterior wall and inserted the needle transendocardially into the cardiac muscle. The injection depth within the cardiac muscle was adjusted using our color Doppler navigation method of the acoustically active needle (Figure 1). Once the AAIC was positioned between the two subepicardially sutured crystals in the anterior wall, linoleic acid was injected through the internal catheter and the needle. The volume of linoleic acid was 1.2–1.5 mL into the anterior wall, which was found effective in the preparative study.
Offline Analysis of the Left Ventricular Function
Echocardiographic scans were digitally transferred to an office workstation with EchoPAC software BT11 (GE Healthcare, Milwaukee, WI, USA). Left ventricular ejection fraction (EF) was measured by using the modified Simpson method. For strain analyses, the endocardium in the standard apical 4-, 3-, and 2-chamber views and the mid-level short-axis view were traced manually, whereas the epicardial outline was generated automatically. However, the width between the two outlines was adjusted interactively to include the entire myocardial wall thickness in diastole and systole. Wall motion was tracked in 6 standard segments in each view throughout the cardiac cycle automatically. The appropriately tracked segments were accepted for analysis based on a visual inspection. A global longitudinal peak strain was calculated by the EchoPAC software in each standard view, and the average of those 3 apical views was obtained as the global longitudinal strain (GLS). As an evaluation of the local wall motion abnormalities, the regional longitudinal strain was assessed with regional strain curves of each segment in the apical views. The regional circumferential and radial strains were analyzed in the short-axis view.
Sonomicrometry data obtained at the desired study time points were processed with proprietary software (Sonosoft, Sonometrics, London, Canada) and the instantaneous mutual distances between crystals, i.e., local ventricular motion, were analyzed.
Cardiac dissection and histological samples of the myocardium
The heart was cut from the apex with 0.5 cm thickness and myocardial tissue changes created by linoleic acid injection were observed on gross sections. The region of interest was sliced and fixed with 10% neutral buffered formalin for at least 72 hours. The heart slices were embedded in paraffin, and sections were obtained and stained with hematoxylin and eosin.
Statistical Analyses
Friedman’s test, the nonparametric equivalent to the one-way repeated measures ANOVA, was performed separately for EF and GLS to test whether there was a significant difference between the mean measurements at baseline, 15 minutes, and 60 minutes. The nonparametric Wilcoxon signed-rank test was used for the three post hoc pairwise t-tests, comparing baseline with 15 minutes, baseline with 60 minutes, and 15 minutes with 60 minutes. Analyses were conducted in R, version 3.6.1.13
Results
Effects of Linoleic Acid and Dosage
In the preparative study, after the injection of linoleic acid under direct visualization, localized increased echogenicity was observed at 15 minutes after injection through the observation period in all 9 animals (Table 1). This high echogenicity area matched the necrotic region of the gross section (Figure 2a and 2d). The regional wall motion abnormality was detected by sonomicrometry recordings, which corresponded to the positive strain of the mid-anterior wall (Figure 2b and 2c). These effects were observed consistently by injecting linoleic acid. One animal (#Pre2, Table 1) had minor bleeding by coronary artery injury at the time of epicardial injection, but otherwise, no major complications occurred. Therefore, we moved to the next step to test the efficacy of AAIC injection using a minimum of 1.2 mL linoleic acid.
Table 1.
Injected Dosage of Linoleic Acid and Outcomes
| Animal Number | Linoleic Acid Dose (mL) | High Echogenicity | Necrotic Change on Gross Sections | Complications |
|---|---|---|---|---|
| Preparative Study with Epicardial Injection | ||||
|
| ||||
| Pre1 | 2.0 | + | + | No |
| Pre2 | 1.3* | + | + | Coronary artery injury |
| Pre3 | 1.2 | + | + | No |
| Pre4 | 1.2 | + | + | No |
| Pre5 | 1.2 | + | + | No |
| Pre6 | 1.2 | + | + | No |
| Pre7 | 1.2 | + | + | No |
| Pre8 | 1.2 | + | + | No |
| Pre9 | 1.2 | + | + | No |
|
| ||||
| AAIC Study with Transendocardial Injection | ||||
|
| ||||
| 1 | 0.7 | + | + | Hypotension |
| 2 | 1.2 | + | + | Left ventricular thrombus |
| 3 | 1.5** | + | + | Heart failure, myocardial penetration, death |
| 4 | 1.2 | + | + | No |
| 5 | 1.2 | + | + | Hypotension |
| 6 | 1.2 | NA | + | Hypotension, arrhythmias, death |
| 7 | 1.5 | + | + | No |
| 8 | 1.5 | + | + | Hypotension |
| 9 | 1.5 | + | + | Left ventricular thrombus |
We stopped injection because of coronary artery injury.
Estimated injected volume because of the needle assembly failure. NA: not applicable due to death soon after the injection.
Figure 2: Effects of linoleic acid (epicardial injection).
(a) High echogenicity area (arrowheads) is observed in the linoleic acid injected area (at 15 minutes). (b) Sonomicrometry recordings show dyskinetic wall motion (arrowheads) of the injected area at 15 minutes. The red lines in sonomicrometry recordings are the time points of opening and closing of the aortic valve. (c) The regional longitudinal strain of the mid-anterior wall (blue line) shows positive strain at 15 minutes (arrowheads). (d) A section of the mid papillary muscle level shows a necrotic area surrounded by hemorrhage (arrowheads), which is compatible with the echocardiography image (a).
Transendocardial Injection of Linoleic Acid Using AAIC
In 8 of 9 AAIC study animals, linoleic acid was injected at the first attempt. In one animal (#3), the needle assembly failed during the procedure, and therefore, we exchanged the catheter for the second injection attempt. In another animal (#1), we stopped the injection before the entire linoleic acid dose was administered because of hypotension, which was managed with a phenylephrine drip, after which the study was completed. Phenylephrine drip was used as needed in other animals to treat hypotension. Animals #3 and #6 presented with progressive heart failure and unmanageable ventricular arrhythmias, respectively, and died before completing the study. Consequently, we obtained baseline data from 9 animals, data at the 15-minute interval from 8 animals, and the 60-minute observations from 7 animals in the AAIC study.
An increased myocardial echogenicity of the injected area and abnormal regional strains detected by echocardiography were observed in 8 animals except the one, which did not have echocardiography recordings due to its death soon after the injection. Regional necrotic change on gross sections was observed in all 9 animals (Table 1).
As assessed by EF (Figure 3a) and GLS (Figure 3b), chronological left ventricular functional changes show systolic functional deterioration or acute decompensation leading to death. There was a significant difference in GLS when we compare baseline, 15 minutes, and 60 minutes (p=0.0058) by Friedman’s test, but no significant difference in EF (p=0.223). A significant decrease in the magnitude of the mean GLS was observed between baseline and 60 minutes (p=0.016) compared to the corrected significance level of 0.05/3=0.0167.
Figure 3: Left ventricular functional changes measured by EF and GLS in AAIC study.
(a) EF showed a trend of decrease after the linoleic acid injection, but there was no statistical difference in mean EF. (b) Mean GLS significantly decreased in magnitude from baseline (−12.6 ± 2.3%) to 60 minutes (−9.2 ± 2.4%; p=0.016).
Examples of regional strains in the linoleic acid-injected area, the corresponding regional wall motion abnormality recorded by sonomicrometry, and representative slices of the dissected hearts from animals #3 and #7 are presented in Figures 4 and 5, respectively. Figure 4a shows minimal discoloration on its surface, whereas Figure 5a demonstrates distinct discoloration due to the transmural extent of necrosis induced by linoleic acid injection. In the animal, where the injection produced necrosis localized between the sonomicrometry crystals, abnormal wall motion patterns were analyzed, as shown in Figures 4b. Figure 4b shows dyskinetic motion at 15 minutes after the injection, which is reflected by positive strain values in the strain curves and the apex area of the bull’s eye plot (Figure 4c and 4d, respectively). In animal #7 (Figure 5), the positioning of the catheter was difficult and, consequently, the injected area was lateral to the anticipated target, as shown in Figure 5a. Inserted sonomicrometry crystals thus did not encompass the lesion. Four-chamber longitudinal strains and short-axis circumferential strains showed hypo-contractility in the lateral left ventricular wall (Figure 5b).
Figure 4: Effects of linoleic acid injection with AAIC (animal #3).
(a) Crystals #1 and #2 are inserted under the epicardium to monitor the regional wall motion by sonomicrometry. Crystal #5 is sutured as a target for the injection. A slight color change on the surface is observed (arrowheads). (b) Sonomicrometry recordings show dyskinetic wall motion (arrowheads) 15 minutes after the injection. Hypotension and decreased wall motion in the control area progressed to a cardiogenic shock. ECG shows ST elevation. The red lines in sonomicrometry recordings are the time points of the aortic valve opening and closing.
(c and d) A longitudinal strain is decreased in magnitude, particularly in the mid anteroseptal region (blue line), and a positive strain (i.e., dyskinetic wall motion) is observed in the apex (pink line in image c and blue area in image d). (e) A mid papillary muscle level section shows a white necrotic area (arrowheads). We dissected the heart after 15 minutes of observation.
Figure 5: Effects of linoleic acid injection with AAIC (animal #7).
(a) A gray-white color change is observed on the surface of the lateral area (arrowheads). (b) A longitudinal strain is decreased in magnitude within the mid-level lateral (blue line) regions of the four-chamber view, and a circumferential strain is decreased in the lateral regions (sky-blue and green lines) of the short-axis view at the papillary muscle level at 15 through 60 minutes after the injection. (c) A high echogenicity area (arrowheads) is observed in the linoleic acid injected area (at 15 minutes). (d) A section of the mid papillary muscle level shows a white necrotic area surrounding hemorrhage in the lateral wall (arrowheads), which corresponds to the echocardiography image (c).
The heart of animal #3 was dissected after 15 minutes; its cross-section showed white necrosis in the anterior wall (Figure 4e). The heart of animal #7 was dissected after 60 minutes and showed a necrotic area with hemorrhage (Figure 5d), which matched the local echogenicity increase in the B-mode echocardiography image (Figures 5c).
Histological Observations
Figure 6a is the linoleic acid-injected myocardium tissue stained by hematoxylin and eosin from the same animal as in Figure 4 (i.e., the animal that did not survive beyond the 15-minute interval of linoleic acid injection). Cardiomyocytes are separated with interstitial edema, and contraction bands, pyknotic nuclei, and cytolysis are observed. Figure 6b is from the same animal as in Figure 5, which completed the 60-minute observation. The hemorrhage observed in the center on the gross cut surface is compatible with a dilated intra-myocardial vessel filled with fibrin (Figure 6b, left), and extravasations of erythrocytes are observed in the surrounding necrotic area with severe myocardial degeneration (Figure 6b, right).
Figure 6: Histological observation of the dissected heart.
(a) Myocardium at 15 minutes after linoleic acid injection. Left: The cardiac myofibers have lost differential staining and normal architecture, with margination and loss of nuclei, hyalinization, and loss of cross striations and contraction bands (necrosis). Adjacent cardiomyocytes undergo degeneration with the margination of nuclei, increase in clear space, variable retention of cross striation, and contraction bands. Right: Cardiac myofibers are preserved with centralized nuclei, cross-striations, but with separation of the interstitium by clear space (asterisks), edema, locally increased amounts of cells, and contraction bands (squares). (b) Myocardium at 60 minutes after linoleic acid injection. Left: There is a dilated intramyocardial vessel (arrowheads) filled with fibrin. Right: The surrounding area of the vessel presents severe myocardial degeneration and necrosis with vacuolated sarcoplasm (arrows) and widening myocardial spaces (asterisks) with mild extravasation of erythrocytes.
Discussion
This article provides insights into the efficacy of an intramyocardial agent delivery by transendocardial injections. We employed a preclinical setting in open-chest instrumented pigs, a prototype of an acoustically active injection catheter, and a novel spatial guidance method by a color Doppler marker to navigate the catheter tip and injection needle. Linoleic acid was used as the efficacy-testing agent, which initiated toxic tissue damage in the injection site. The expected effects included visible changes in tissue characteristics by echocardiography, segmental wall motion abnormality, and localized necrosis, as predetermined in the preparative study. On the other hand, we experienced complications presenting as unfavorable effects of linoleic acid delivery with AAIC, as further described in the Discussion section.
Percutaneous transendocardial delivery has been considered a highly promising method in cardiac gene therapy, providing better transduction efficiency than coronary artery delivery and less invasiveness against open-chest delivery.4 Several transendocardial injection catheters have been tested in preclinical and clinical trials. However, about 1–3 % risk of adverse events of myocardial perforation, which possibly leads to cardiac tamponade or death in clinical trials, and presence of injury on the epicardial surface in pre-clinical trials, prompt the importance of continuous research and development of effective and safe delivery methods.14–18 Therefore, in this technology innovation study, we tested an AAIC prototype, which could visualize the injection needle tip’s insertion depth and spatial position within the beating myocardium using routinely available Doppler echocardiography.
Evaluation of the Linoleic Acid Effects – Efficacy of the Delivery Method
We did not have significant systemic complications in the preparative study (epicardial injection under direct visualization), while we had 2 premature animal deaths in the AAIC study. Although AAIC enabled successful substance delivery, we observed systemic events related to the procedure, such as rapid hemodynamic deterioration immediately after the linoleic acid injection and global left ventricular systolic dysfunction (Table 1, Complications). We explain these events as consequences of leakage of linoleic acid into the left ventricular blood flow or systemic circulatory distribution of linoleic acid through myocardial vasculature.
In most cardiac regenerative therapy or gene therapy cases, the target is a highly localized area (e.g., an ischemic border zone). Moreover, an injection needle tip needs to be positioned well within the ventricular wall for procedural safety and effective delivery. On the other hand, the current therapeutic agents typically do not achieve immediate treatment response (i.e., wall motion recovery) after the injection. Therefore, it is hard to assess whether the delivery was effective or not at the time of the procedure. A color dye can be used in a preclinical setting as a marker of agent retention within the myocardium. However, the dye’s obvious disadvantage is that its myocardial distribution can be evaluated only after the heart dissection. Mixing the agent with microbubble contrast could reveal its intramyocardial distribution or intraventricular leak immediately after injection by switching from Doppler imaging to contrast echocardiography with a low mechanical index. However, the contrast microbubbles last only for minutes, whereas spreading of the agent would continue in a beating heart.
For that reason, instead of attempting an immediate recovery of an ischemic region, we reversed the scenario and induced a predictable myocardial injury by transendocardial injection of linoleic acid in anesthetized pigs. Linoleic acid is one of the n-6 polyunsaturated fatty acids and is considered an essential fatty acid in the body. It has been reported that linoleic acid mediated by lipolysis inhibited mitochondrial complexes I and V, and increased inflammatory mediators in acute pancreatitis lipotoxicity models,10,19 which means linoleic acid acutely induces direct toxic cellular damage while triggering inflammatory responses. We observed substantial hemorrhage and myocardial cell damage at 60 minutes after the injection in our experiments. Visual tissue color changes started around 15 minutes. We dissected 2 animals at a 15-minute time point, which showed white tissue color changes in gross anatomy. Histologically, myocyte damage and necrosis characterized by contraction bands, cytoplasmic vacuolation, and pyknotic nuclei were observed in all animals in the AAIC study. At 60 minutes, hemorrhage and dilated vessels filled with fibrin were observed on top of myocardial degeneration and necrosis. Though we only have 2 histological observations at 15 minutes, vasculature damage and hemorrhage may have followed the initial toxic tissue damage caused by the linoleic acid injection.
In our current experimental setting, linoleic acid intramyocardial delivery reproducibly caused visible changes in tissue characteristics and regional wall motion abnormalities. These changes were observable visually in the open-chest study, typically within 15 minutes of injection and verified by gross sections and histology, as well as offline echocardiographic and sonomicrometric measurements. In particular, we used either longitudinal strains, circumferential strains, or radial strains for the assessment of left ventricular wall motion. We found a segmentally decreased magnitude of the strains, delayed time to the peak strains, or positive strains in either of the three types of strain measurements obtained in all animals following linoleic acid injection. The changes in longitudinal strain patterns were compatible with the changes observed by sonomicrometry. As shown in Figures 2, 4, and 5, the location of abnormal segmental strains was dependent on the affected area (linoleic acid injected area).
In our chronological observations, the regional wall motion deteriorated most at 15 minutes after the linoleic acid injection with a compensatory hyperkinetic motion of the remote myocardial area. Global systolic ventricular function decreased till 60 minutes in 7 animals; however, rapid deterioration leading to unmanageable heart failure and premature death occurred in 2 animals. We found a statistically significant decrease in the magnitude of the GLS but not in EF. EF is an established systolic functional measurement, which calculates the ratio of stroke volume to left ventricular diastolic volume, while GLS measures the deformation of the myocardium. When the area remote to the lesion is hyperactive, as in single vessel myocardial infarction or our experimental model of local myocardial injury, EF can underestimate the impact of the damage. Therefore, GLS is considered more sensitive to detect systolic dysfunction.20,21
A recent report has proposed the association between mortality of COVID-19 and free unsaturated fatty acids whose lipotoxicity may trigger endothelial injury, shock, and multisystem organ failure.22 We speculate that a systemic toxic effect of linoleic acid could have caused the global systolic dysfunction and, especially, the one case of unmanageable acute heart failure. However, other confounding factors, such as decreased venous return in the open-chest setting, positive pressure ventilation, and anesthesia, may have also negatively influenced global cardiac function.
Safety of the AAIC System
We used only ultrasound guidance in this study and navigated the AAIC by the instantaneous color markers of its tip and the retractable needle tip. We guided the AAIC from initial transcutaneous insertion into the right carotid artery, through its advancement via aorta and placement into the left ventricle, to transendocardial injections of the left ventricular wall. There was no vascular or valvular injury. We experienced one myocardial penetration with the needle (animal #3). However, this complication was not caused by a problem in guidance. Instead, the injection assembly in our prototype detached, and the needle was inadvertently pushed far into the myocardial wall. In the rest of the animals, thanks to the color marker guidance of the needle tip, we could avoid myocardial penetration by making sure the needle crystal was within the myocardial wall. Visualization of both needle tip and guiding catheter tip can provide safer and more precise transendocardial delivery.7
In 2 animals, a thrombus formed at the left ventricular apex despite systemic heparinization. We speculate that thrombus formation was provoked by linoleic acid, which could accelerate a coagulation cascade triggered by necrotic tissue and endothelial damage.23
Although carefully installed and smoothed, the crystal embedded in the needle tip could cause additional injury to the myocardium during the injection. We speculate that a commercial-quality production and the use of thinner and short-beveled acoustically active needles would further reduce the chances of injury associated with the presence of the crystal.
Limitations
This study, which uses an injectable agent in a cardiotoxic concentration, cannot be translated into a human setting. However, the efficacy-testing method was practical and educational for improving the transendocardial agent delivery system and skills in the presented experimental preclinical setting.
In this experiment, we used two-dimensional scans whose standard views for the analysis of longitudinal strains could not always include the linoleic acid-induced lesion center. A three-dimensional echocardiography system could have captured any lesion for strain analyses with one beat to evaluate wall motion abnormality and will be considered in future studies.
Sonomicrometry was used to monitor local myocardial motion during the experiment. However, the method’s utility was limited in our setting because the distribution of the necrosis was highly variable and, thus, the implanted crystals could not reliably represent the location of the affected region.
Lastly, visually observed increased echogenicity and necrotic changes on gross sections were practical assessments, but not quantitative. Additional quantitative assessments would be required to compare the efficacy of different delivery systems.
Conclusions
We have introduced an experimental animal model for testing the efficacy of transendocardial injections. Intramyocardial injection of linoleic acid rapidly induces localized myocardial necrosis and dysfunction if the delivery is efficacious. The model was used to test our AAIC prototype navigated by real-time color markers generated in Doppler ultrasound scans. The transendocardial delivery with AAIC was successfully done, generating regional texture change and wall motion abnormality detected by echocardiography during the procedure. Further developments in design and portability are required for clinical applications of AAIC.
Acknowledgments
This study was funded by the National Institutes of Health Grant R01 EB019947 and the Mayo Clinic Center for Biomedical Discovery Grant FP00110922. The Advance the Practice Research Award from the Mayo Clinic Center for Clinical and Translational Science supported improvements in the AAIC catheter design.
References
- 1.Singal AG, Higgins PD, Waljee AK. A primer on effectiveness and efficacy trials. Clin Transl Gastroenterol Jan 2 2014;5:e45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Henry TD, Traverse JH. Location, location, location! Is it important with cell delivery? Catheter Cardiovasc Interv May 2013;81(6):978–9. [DOI] [PubMed] [Google Scholar]
- 3.Sherman W, Martens TP, Viles-Gonzalez JF, Siminiak T. Catheter-based delivery of cells to the heart. Nat Clin Pract Cardiovasc Med Mar 2006;3 Suppl 1:S57–64. [DOI] [PubMed] [Google Scholar]
- 4.Ishikawa K, Tilemann L, Fish K, Hajjar RJ. Gene delivery methods in cardiac gene therapy. J Gene Med Oct 2011;13(10):566–72. [DOI] [PubMed] [Google Scholar]
- 5.McMahon EM, Jiamsripong P, Katayama M, Chaliki HP, Fatemi M, Belohlavek M. Accurate guidance of a catheter by ultrasound imaging and identification of a catheter tip by pulsed-wave Doppler. Pacing Clin Electrophysiol Jan 2012;35(1):44–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Belohlavek M, Katayama M, Zarbatany D, et al. Acoustically active injection catheter guided by ultrasound: navigation tests in acutely ischemic porcine hearts. Ultrasound Med Biol Jul 2014;40(7):1650–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Katayama M, Kumar V, Zarbatany D, et al. Acoustic Navigation of Intramyocardial Injection Needle Catheter Using Color Doppler Echocardiography. IEEE International Ultrasonics Symposium, IUS. Kobe, Japan: IEEE; 2018. p. 1–3. [Google Scholar]
- 8.Katayama M, Zarbatany D, Cha SS, Fatemi M, Belohlavek M. Acoustically Active Catheter for Intracardiac Navigation by Color Doppler Ultrasonography. Ultrasound Med Biol Sep 2017;43(9):1888–1896. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Seldinger SI. Catheter replacement of the needle in percutaneous arteriography; a new technique. Acta radiol May 1953;39(5):368–76. [DOI] [PubMed] [Google Scholar]
- 10.Navina S, Acharya C, DeLany JP, et al. Lipotoxicity causes multisystem organ failure and exacerbates acute pancreatitis in obesity. Sci Transl Med Nov 2 2011;3(107):107ra110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Belohlavek M, Katayama M, Vaitkus VV, et al. A Real-time Color Doppler Marker for Echocardiographic Guidance of an Acoustically Active Extracorporeal Membrane Oxygenation Cannula. J Ultrasound Med Jul 2019;38(7):1875–1885. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kinnick RR, Katayama M, Belohlavek M. An Interposed Pad in Open-Chest Echocardiographic Porcine Scans for Mimicking Ultrasound Signal Attenuation in a Human Chest. J Ultrasound Med Feb 2018;37(2):501–509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.R: A language and environment for statistical computing. https://www.R-project.org/. [Google Scholar]
- 14.Losordo DW, Henry TD, Davidson C, et al. Intramyocardial, autologous CD34+ cell therapy for refractory angina. Circ Res Aug 5 2011;109(4):428–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kastrup J, Jorgensen E, Ruck A, et al. Direct intramyocardial plasmid vascular endothelial growth factor-A165 gene therapy in patients with stable severe angina pectoris A randomized double-blind placebo-controlled study: the Euroinject One trial. J Am Coll Cardiol Apr 5 2005;45(7):982–8. [DOI] [PubMed] [Google Scholar]
- 16.Povsic TJ, Henry TD, Traverse JH, et al. The RENEW Trial: Efficacy and Safety of Intramyocardial Autologous CD34(+) Cell Administration in Patients With Refractory Angina. JACC Cardiovasc Interv Aug 8 2016;9(15):1576–85. [DOI] [PubMed] [Google Scholar]
- 17.Kumar A, Haralampus CA, Hughes M, et al. Assessment of safety, accuracy, and human CD34+ cell retention after intramyocardial injections with a helical needle catheter in a porcine model. Catheter Cardiovasc Interv May 2013;81(6):970–7. [DOI] [PubMed] [Google Scholar]
- 18.Penn MS, Mendelsohn FO, Schaer GL, et al. An open-label dose escalation study to evaluate the safety of administration of nonviral stromal cell-derived factor-1 plasmid to treat symptomatic ischemic heart failure. Circ Res Mar 1 2013;112(5):816–25. [DOI] [PubMed] [Google Scholar]
- 19.Durgampudi C, Noel P, Patel K, et al. Acute lipotoxicity regulates severity of biliary acute pancreatitis without affecting its initiation. Am J Pathol Jun 2014;184(6):1773–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Potter E, Marwick TH. Assessment of Left Ventricular Function by Echocardiography: The Case for Routinely Adding Global Longitudinal Strain to Ejection Fraction. JACC Cardiovasc Imaging Feb 2018;11(2 Pt 1):260–274. [DOI] [PubMed] [Google Scholar]
- 21.Klaeboe LG, Edvardsen T. Echocardiographic assessment of left ventricular systolic function. J Echocardiogr Mar 2019;17(1):10–16. [DOI] [PubMed] [Google Scholar]
- 22.El-Kurdi B, Khatua B, Rood C, et al. Mortality from Covid-19 Increases with Unsaturated Fat, and May Be Reduced by Early Calcium and Albumin Supplementation. Gastroenterology May 27 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Grover SP, Mackman N. Tissue Factor: An Essential Mediator of Hemostasis and Trigger of Thrombosis. Arterioscler Thromb Vasc Biol Apr 2018;38(4):709–725. [DOI] [PubMed] [Google Scholar]