Ligation of the Left Circumflex Coronary Artery with Subsequent MRI and Histopathology in Rabbits

Abstract

Provided is the surgical procedure for ligating the left circumflex coronary artery to simulate heart ischemia by using a rabbit model. Heart rate monitored by electrocardiogram was increased at 5 min after ligation (mean ± SEM, 205 ± 13 bpm) when compared with that before ligation (170 ± 12 bpm), but returned to baseline at 25 min after ligation (183 ± 11 bpm). A marked elevation in the ST segment and reduction of the QRS wave of the electrocardiogram indicated the evolving myocardial infarct. The ejection fraction derived from MRI was decreased by 20% in the infarcted heart. The extent of necrosis and fibrosis in the myocardium due to ischemia led to decreased compliance and efficiency of the left ventricle. Masson trichrome staining showed blue-stained fibrils with the appearance of loose, threadlike scar tissue dispersed transmurally in the left ventricle and extending toward the apex. This study demonstrates the feasibility of MRI analysis of myocardial infarction in a rabbit model. The myocardial architecture, including the geometry of the myofibers which determines the contractile function of the heart, is clearly demonstrated by using cardiac MRI. Understanding the 3-dimensional arrangement of the myocardial microstructure and how remodeling of the infarcted myocardium affects cardiac function in an animal model has important implications for the study of heart disease in humans.

Atherosclerotic plague in the coronary artery causes narrowing of its lumen, resulting in failure to supply adequate oxygen and nutrients to the heart. This limitation accounts for the clinical manifestations of myocardial infarction. The resulting impairment in the contraction and relaxation of the myocardium also causes the heart to function suboptimally. Such coronary heart disease is the leading cause of death in the United States. Each year, an estimated 1.2 million Americans have a first or recurrent coronary heart attack, with mortality exceeding 450,000 deaths annually.^1,19 Improved treatment has had a substantial effect on reducing the number of deaths due to coronary heart disease. However, 37% of patients who experience heart attacks during a given year die from it.

Numerous studies have been performed that involve the use of angiogenic growth factor therapies, such as fibroblast growth factor 1 and vascular endothelial growth factor.^{4,20,27,28,34} Potential cellular therapies using stem cell populations have been assessed as well.^17,26 Despite a substantial number of articles describing coronary artery disease in animal models generated from coronary artery manipulation, only a few provide detailed methodology of the procedure specifically in the rabbit model.^5-7,33 Here we describe a comprehensive protocol to expose the left circumflex artery of rabbits by means of a left thoracotomy; the artery then can be ligated to simulate irreversible left ventricular ischemia in humans. Through the use of this model, numerous studies can be generated and modified to understand and treat the ischemic heart disease. In our current study, we include electrocardiography to analyze the evolving myocardial infarction and monitor the heart rate during surgery, MRI to assess the longitudinal progression of ischemia by monitoring cardiac function, and Masson trichrome staining to differentiate cellular and tissue changes.

Materials and Methods

Study group.

Male New Zealand white rabbits (n = 13; weight, 2.5 to 3.0 kg; Western Oregon Rabbits, Philomath, OR) were quarantined for 6 d prior to surgery. The rabbits were cared for in accordance with federal and local animal welfare regulations in an AAALAC-accredited facility, and the study was approved by the University of Utah Institutional Animal Care and Use Committee and followed guidelines in the Guide for the Care and Use of Laboratory Animals.¹² Eight rabbits underwent left circumflex coronary artery ligation by means of a left thoracotomy, and 5 rabbits served as no-surgery controls. MRI and Masson trichrome staining were performed in both operated and nonoperated controls. A fentanyl patch (25 µg/h) was applied to the rabbit 12 h before surgery and 48 h afterward to relieve any pain associated with the surgical procedure. If no patch was applied or if a patch was applied the morning of surgery (that is, less than 12 h prior to surgery), we injected 0.15 mg buprenorphine (0.06 mg/kg IM) prior to sedation of the rabbit.

Rabbit sedation and preparation.

Rabbits were sedated and prepared for surgery in an animal prep room. A combination of ketamine (40 mg/kg IM) and xylazine (4 mg/kg IM) was administered in a hindquarter. Appropriate sedation was achieved in 5 to 15 min. The left lateral thorax of the rabbit was shaved. A 22-gauge angiocatheter was inserted into the dorsal ear vein and secured with tape. A needle-free valve port was connected to the angiocatheter through which we injected 100 mg cefazolin (40 mg/kg IV) and flushed the catheter with 0.5 mL sterile normal saline.

Intubation.

During intubation, additional ketamine (60 mg in 0.6 mL) and xylazine (20 mg in 0.2 mL) was administered in aliquots of 0.1 to 0.2 mL IV to the rabbit as necessary to maintain appropriate sedation. The operating table was kept warm with circulating water at 37.5 °C from a warming tank. With the operator's left hand holding the dorsal side of the rabbit's neck just below the jaw, the rabbit's head was raised and its neck slightly hyperextended. Using the left thumb to protect the upper teeth, the Miller 1 blade of a laryngoscope (Teleflex Medical, Bannockburn, IL) was inserted into the larynx (Figure 1). The handle of the blade was tilted up, and the operator's left thumb maneuvered the trachea to bring the vocal cords into view. A 3.0-mm endotracheal tube was inserted gently into the rabbit's larynx through the vocal cords. This portion of the procedure required gentle manipulation of the endotracheal tube and trachea to allow a smooth insertion. Intubation was verified by fogging of the endotracheal tube on expiration and feeling the rabbit's breath through the tube.

Figure 1.

Intubation of a rabbit, with the neck hyperextended to facilitate insertion of the Miller 1 blade of the laryngoscope and the endotracheal tube.

After placing the rabbit in the left lateral decubitus position on the operating table and connecting the respirator (SurgiVet, Waukesha, WI) to the endotracheal tube, we monitored chest movement and the lack of gastric expansion to reaffirm correct positioning of the endotracheal tube. The respirator was set at 20 breaths per minute with a tidal volume of 75 mL and oxygen flow of 2.0 L/min. Sedation was maintained by using isoflurane gas at an average setting of 1.5%. The front and hindlimbs were restrained to the operating table. Heart rate was monitored with a 10-channel custom 3-lead electrocardiography system (University of Utah, Salt Lake City, UT), with the leads attached to the left foreleg and right hindlimb and the ground to the left hindlimb. Analog waveforms were sampled digitally at 500 Hz through an analog-to-digital board (Data Acquisition Card 6062E, National Instrument, Austin, TX) and analyzed by using custom software (LabView, National Instruments, Austin, TX).¹¹ Cycle length was defined as the time interval between consecutive upstrokes of P waves in the electrocardiogram, and heart rate was calculated from the reciprocal of cycle length.

Operation.

The previously shaved portion of the left thorax was disinfected with povidone–iodine. Aseptic technique was maintained throughout the operative procedure, and 2.5× surgical loupes were used to aid in the surgery.

Skin and intercostal incision.

A 4-cm incision was made with a no.10 blade, starting approximately 4 cm posterior to the scapula and extending medially over the heart region. The skin was separated from the latissimus dorsi by using Metzenbaum scissors to dissect bluntly along the incision (Figure 2). The latissmus dorsi and serratus anterior were separated in a similar fashion to expose the ribs and intercostal muscles, at which point the rabbit was injected with the first bolus of lidocaine (0.30 mg IV) to decrease the likelihood of arrhythmia after ligation. An appropriate rib space overlying the heart was chosen by using palpation. The intercostal muscles were incised with Metzenbaum scissors, with care to remain along the superior edge of the rib to avoid the neurovascular bundle. The pleural space was entered sharply but gently during expiration to avoid injury to the lung. A small Weitlaner retractor was inserted between the ribs and lifted upward while spreading. A strip from a 5 × 5 cm² gauze pad was moistened with sterile normal saline and inserted using wet cotton-tipped applicators to push the lung toward the back of the thoracic cage to expose the heart.

Figure 2.

Left posteriolateral thoracotomy incision of a rabbit. Blunt dissection is applied to separate the skin, and latissimus dorsi and serratus anterior muscles prior to incision of the intercostal muscles.

Ligation of the left circumflex coronary artery.

The pericardium was incised carefully, starting at the apex of the heart, to avoid injury to the left atrium. After the heart was freed from the pericardium, a cotton applicator was used to manipulate the left ventricle and identify the left circumflex artery. Once the artery was identified, 6-0 polypropylene suture was used to ligate the left circumflex artery approximately 1 cm below the left atrial appendage. We administered a second dose of lidocaine (0.15 mg IV) and observed the heart for infarction. Hypoxic heart tissue turned from pink to deep purple at the sutured area, and this change in coloration extended toward the apex of the ventricle (Figure 3). This area frequently became hypokinetic.

Figure 3.

Thoracotomy of a rabbit in left lateral decubitus position. The left lung is pushed to the back of the thoracic wall with a wet gauze pad to expose the heart. The left ventricle develops ischemia as a result of ligation of the left circumflex coronary artery with nonabsorbable 6-0 prolene suture (black arrow) and shows a distinct boundary of discoloration from red to purplish color (white arrowheads). Part of the circumflex coronary artery (black arrowhead) is still visible in the anterior portion of the left ventricle.

Closing the incisions.

The retractor was removed by lifting it to raise the rib cage while releasing the retractor lock. This practice prevented puncturing of the lung. Using 3-0 polydioxanone suture, we tied 2 figure-8 stitches around the ribs. Prior to tying down the sutures, the gauze pad strip was removed from the thoracic cavity, and residual air in the pleural space was expelled by inflating the lung using a 100-mL rebreathing bag connected to the endotracheal tube. The tube was reconnected to the ventilator after the ribs were tightly closed. The final bolus of lidocaine (0.10 mg IV) was administered. A 3-0 polydioxanone suture was used to close the 2 layers of muscle individually in a running fashion, and the skin was closed in a subcuticular fashion. We turned off the isoflurane but continued monitoring the rabbit until it showed signs of spontaneous breathing. Surgery was concluded by extubating the rabbit and removing the angiocatheter from the ear.

MRI study.

At 30 d after ligation, the rabbit was sedated with the previously described ketamine–xylazine combination. The left side of the chest from the dorsal to ventral midline and superior scapula to costal margin was shaved, and sensitive-formula depilatory cream (Nair, Church and Dwight, Princeton, NJ) was applied to remove any remaining hair. A 22-gauge angiocatheter inserted into the dorsal ear vein was connected to a needle-free valve port and minibore administration set tubing filled with sterile normal saline. With the rabbit enclosed in the MRI chamber, the long tubing facilitated administration of additional aliquots of intravenous ketamine–xylazine as required during the scanning. Control rabbits underwent the same procedures for MRI study.

Scanning was performed on a 3-T instrument (Trio, Siemens, Erlangen, Germany) with 2D cine acquisition of the whole heart by using a retrospective electrocardiogram-gated, spoiled-gradient echo, k-space–segmented sequence.^9,21,24,30 The rabbit was placed in supine position and dual custom 10.2 × 11.4 cm paddle-shaped coils, each designed with 2 elements that were oriented in opposite positions, were positioned over the heart region.¹⁰ The parameters of the imaging sequence involved using retrospective cardiac gating with a multiple-average or multiple technique of 5 averages to ‘blend’ respiratory motion artifact from the sedated rabbit. Myocardial tagging was performed immediately after the R-wave trigger prior to a cine acquisition. Cine MR images were acquired along both the short- and long-axes of the heart, and in-plane resolution was acquired at the submillimeter level at 0.8 × 0.7 mm with a slice thickness of 2.5 mm. The short-axis sequence contained 14 slices and had a repetition time of 57.4 ms, echo time of 3.54 ms, field of view of 130 × 130 mm, a matrix of 192 × 154, and a flip angle of 12°. The long-axis sequence contained 3 slices each of 2-chamber (left and right ventricle) and 4-chamber views, with scan parameters the same as those for the short-axis sequence. Volumetric evaluation was performed by using standard software (ARGUS, Siemens). Contouring of left ventricular endo- and epicardial borders was performed semiautomatically, and end-diastolic and end-systolic volumes were calculated.

Histology section.

Each rabbit was sedated as described for MR imaging, injected intravenously with 0.5 mg/kg heparin, and 10 min later, euthanized with 120 mg/kg ketamine. The skin and muscle tissue were excised from the xiphoid to the neck, and the chest plate was removed by cutting along the sides of the costochondral cartilages and clavicles to expose the heart. The entire heart was isolated from the lungs and thymus, and the aorta of the heart was tied over the cannula of a 22-gauge catheter with a 0 silk suture. The heart was perfused with heparin in PBS (1:100) to flush any remaining blood, flushed with 2.56 M KCl to arrest it in diastole, and fixed in 10% formalin. The heart was sliced in coronal sections, dehydrated through an ascending ethanol series, and embedded in paraffin. Serial sections of 5 µm were cut and stained with Harris hematoxylin stain, which colored basophilic nuclei bluish purple, and counterstained with eosin Y to stain cells eosinophilic.⁸ In addition, Masson trichrome staining of fixed samples was used to differentiate between collagen and muscle fibers. Trichrome staining is applied by immersion of sections into Weigert iron hematoxylin followed by Biebrich scarlet–acid fuchsin (also known as plasma stain), phosphomolybdic–phosphotungstic acid, and aniline blue (also known as fiber stain).¹³ Images from the sections were captured digitally (DFC420 camera fitted to a DM1000 microscope, Leica, Bannockburn, IL) by using Leica Image Manager software.

Statistical analysis.

Data are presented as mean ± SEM. We used paired t tests for comparison of electrocardiogram cycle lengths before and after vessel ligation and 2-sample t-tests of MR imaging between unoperated and ligated rabbits. Significance was defined as a P value of less than 0.05.

Results

Electrocardiography.

Before surgical ligation of the left circumflex coronary artery, all rabbits displayed typical electrocardiographic waveforms, with distinct P waves, QRS complexes, and T waves (Figure 4 A). At 5 min after ligation, T wave inversion and diminished R waves were present (Figure 4 B). Prominent changes in the electrocardiogram pattern emerged as early as 25 min after ligation, when marked elevation in the ST segment and distinctive T wave peaking appeared (Figure 4C). This type of peak, which was tall and broad, is also known as a ‘hyperacute T wave.’²⁹ During surgery, we observed occasional shortening of the QRS complex but no distinctive change in the Q wave. Cycle lengths, derived from cardiac waveforms, at 25 min after ligation (332.8 ± 18.1 ms) were comparable to those prior to surgery (362.3 ± 26.9 ms) and corresponded to heart rates of 183 ± 11 and 170 ± 12 bpm, respectively. Cycle lengths at 5 min after ligation (299.5 ± 17.9 ms; 205 ± 13 bpm) were decreased significantly (P < 0.05) compared with those before ligation.

Figure 4.

Rabbit electrocardiogram (A) during preligation, (B) at 5 min after ligation, and (C) at 25 min after ligation of the left circumflex coronary artery. (A) The tracing shows a typical P wave, QRS complex, and T wave. (B) After 5 min of ligation, there is a distinct inversion (arrow) at the end of the T wave and the R wave is diminished, suggesting onset of cardiac ischemia. (C) Marked elevation of the ST segment combined with the peak in the T wave at 25 min after ligation signifies myocardial infarction. The S wave is affected also in this tracing.

MRI.

MRI analysis revealed reliably reproducible findings in the infracted hearts (Figure 5). Stroke volume, calculated from the difference between end-diastolic volume and end-systolic volume (Table 1), was decreased (P < 0.05) in infarcted myocardium (1.2 ± 0.2 mL) when compared with control hearts (2.0 ± 0.3 mL). The ejection fraction of infarcted hearts, derived from the ratio between stroke volume and end-diastolic volume, was decreased by 20% in infarcted hearts compared with control hearts (Table 1).

Figure 5.

MR images of an infarcted and normal heart at end-diastole and end-systole, captured in real time by using gating to acquire sequences at each stage of the cardiac cycle over several heart beats. Blood appears typically brighter during ventricular filling because of its contrast property and rapid flow. In contrast to the normal rabbit heart, the infarcted heart loses its definite margin of the left lateral ventricular wall (the shorter circumference between a and a′). In addition, wall motion is restricted where the infarcted zone loses its spontaneous pumping action.

Table 1.

Cardiac measures derived from MRI scanning

	End-diastolic volume (mL)	End-systolic volume (mL)	Ejection fraction (%)
Control hearts	3.9 ± 0.3	2.0 ± 0.2	49.3 ± 4.2
Infarcted hearts	3.1 ± 0.3	1.9 ± 0.3	38.2 ± 3.1

Histology sections.

The infarcted region of the myocardium displayed a pale-whitish color. The collagen fibers were characterized by lighter staining with hemotoxylin and eosin reflecting decreased density of myocytes, and heavier Masson trichrome staining indicating fibrosis (Figure 6). The collagen fibers appeared as slender, threadlike structures woven into varying degrees of density or in a loose areolar tissue. The scar tissue extended transmurally across the width of myocardium and had a thinner compact layer. The amount of collagen also dispersed toward the apex of the ventricle.

Figure 6.

(A) Coronal section of the heart showing the ventricular septum (arrow) and the left ventricle. Trichrome stain differentiates between collagen and smooth muscle. Because the cytoplasm is less dye-permeable than collagen, the ventricle retains most of the red dye whereas the decolorized infarcted area absorbs the collagen dye (aniline blue). The scar tissue extends transmurally in the left ventricle and spreads toward the apex. Magnification, ×40. (B) The boxed area in panel A, showing the blue-stained collagen as loose, threadlike structures. Magnification, ×100.

Discussion

Through precise technique and careful monitoring, our surgical procedure for ligation of the left circumflex coronary artery in rabbits yields high levels of success without mortality. Cardiac MRI was completed successfully in all unoperated and ligated rabbits. To ensure survival, we occasionally make use of an aspirator mask to resuscitate the rabbit when necessary. These situations generally arise due to potential difficulties during intubation that stem from rabbit anatomy.^5,6 Intubation of rabbits requires considerable patience and delicate manipulation to avoid any injury. If the rabbit stopped breathing during this part of the procedure, we momentarily stopped intubating and placed the aspirator mask ventilated with oxygen flow at 2.0 L/min over the nose of the animal until voluntary breathing resumed. Although the possibility of performing rabbit thoracotomy without intubation has been proposed,⁷ we determined this part of the procedure to be highly beneficial, as it prevents the near 30% mortality reported in the absence of intubation. An additional benefit to intubation is being able to maintain strict control of sedative dosage. Fatality due to intravenous anesthetic overdose arose in a previous study.⁶ Although use of water-based lubricant (for example, KY Jelly) has been suggested to facilitate the intubation of rabbits,⁵ we do not find it necessary for successful intubation.

Although we had a defibrillator and lidocaine readily available, we did not encounter any complications during surgery that required their use. We ligated the left circumflex coronary artery to produce maximal infarction of the left lateral ventricle. The left main coronary artery divides into left anterior descending artery and left circumflex artery; the former supplies the anterior wall of the heart and most of the ventricular septum, whereas the latter runs between the left atrium and left ventricle and supplies the lateral wall of the left ventricle. Almost all myocardial infarctions involve the left ventricle, which is the most muscular chamber that does the most work, and therefore is most vulnerable to a compromised blood supply.³³ Ligation of the coronary artery close to midventricle typically results in infarction of 50% of the left chamber.³⁴ Bleeding is usually minimal at the area of coronary artery occlusion and can be stopped when slight pressure from a cotton applicator is applied to the sutured area. We also used pinch pressure or hemostats to stop bleeding from the skin or muscle incision. All medication administrated intravenously should be flushed with 0.5 to 1.0 mL sterile normal saline.

The electrocardiography leads need to be secured tightly to the rabbit throughout the procedure. Any change in the lead positions during surgical procedure may alter the pattern of the waveform and consequently yield inaccurate readings. The electrocardiography system we use for animal research has only 3 leads yet still provides essential diagnostic information. The inversion of the T wave accompanied by shortening of the R wave in our EKG waveform is the initial sign of myocardial ischemia soon after ligation. Marked elevation in the ST segment combined with T wave peaking signifies transmural myocardial infarction. Occlusion of the left circumflex coronary artery creates myocardial infarction in the left lateral wall of the heart. The diminished waveform in the QRS complex after ligation suggests that the heart has undergone severe myocardial injury. Any depletion of oxygen to the ventricle generates adverse effects to the conduction system that may be reflected in the ventricular depolarization.¹⁴ The intense manipulation of the heart during occlusion of the coronary artery may increase the troponin level to elevate muscle contraction, resulting in an increased heart rate.²³

The stroke volume and ejection fraction derived from MRI are reduced in the infarcted myocardium of our rabbits. Ventricular contraction expels only a part of the blood contained in the heart chamber.²² A residual volume of 30% to 60% may still remain in the ventricle at the end of systole, the actual cardiac output and aortic stroke volume could be substantially lower. Occlusion of the circumflex coronary artery results in extensive necrosis and proliferation of fibrosis, thus decreasing myocardial contractility. As a consequence of this structural remodeling, regional and global myocardial functions are altered.^3,25 The ventricle thus becomes much stiffer, and the heart is less compliant. Therefore, interstitial myocardial fibrosis is a distinctive pathogenesis of adverse left ventricular remodeling in myocardial infarction. Since the myocardial architecture, including the geometry of myofibers, accounts for the wall motion and pump function of the heart, distribution of interstitial fibrosis during ventricular remodeling is a critical factor for prognosis of coronary heart diseases.¹⁵

The use of Masson trichrome staining to distinguish collagen from muscle is a valuable histopathologic tool that indicates fibrotic changes, particularly an increased amount of collagen. The acidic dye stains the acidophilic cytoplasm of healthy myocardial tissue. Due to its comparatively loose texture, collagen is readily permeable to almost any dye. However, the dye will easily diffuse out on subsequent treatments, enabling collagen to be differentially stained with aniline blue, leading to blue-stained collagen tissue.¹³ Apoptotic death of cardiac myocytes occurs long after ligation of the left circumflex coronary artery and becomes noticeable at 30 d after ligation.³⁴ Postinfarction remodeling of the left ventricle is characterized by hypertrophy and fibrotic changes to the heart. Many myocytes die through necrosis due to acute myocardial ischemia.² Therefore, myocardial infarction results in a decrease in contractile force as a result of loss of myocytes. Surviving myocytes are hypertrophied, probably in an attempt to compensate for the infarction-induced decline in cardiac function.

Rabbits are an excellent model for myocardial study.³³ Their heart rate, even when anesthetized, does not exceed the range of normal human heart. Rabbit hearts are sufficiently small, and transverse section can readily be viewed on standard microscope slides. In addition, rabbits are relatively inexpensive for a midsized specimen. The methods we describe here yield preliminary data that have considerable diagnostic value and show direct correlations among the procedures. Using a combination of MRI and histology, we are able to explore the helical ventricular myocardial band in rabbits to understand the 3D organization of myocardial microstructure and how its form as a muscular band relates to cardiac motion; and to provide novel insights into the structure–function relationship in the post-infarct myocardium.^18,32 Further studies in any of these individual techniques would yield more definitive information concerning this fundamental concept and would require comprehensive MR scan protocols that include electrocardiography-gated cine imaging, myocardial tagging, perfusion imaging, and delayed enhancement or viability imaging to visually assess the damage due to myocardial infarction. Other applications of this particular model of heart ischemia might include implantation of ameroid constrictors or injection of therapeutic agents, gene plasmids, or growth factors directly into the infarcted myocardium.^{16,20,31,34,35} Prior studies describing different surgical methods have proven to be adequate for their specific goals.^5-7,33 These techniques have their own advantages and limitations. When combined, they may be able to complement each other. While the preference of surgical procedure depends greatly on the availability of facilities as well as investigators’ requirements, we intend to provide a thorough methodology that is effective and easily reproduced by any researcher in this field.