INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

THE RAT IS ONE of the most versatile and frequently utilized animal models in cardiovascular research. Models of myocardial hypertrophy, myocardial infarction, and congestive heart failure have been successfully developed in rats for structural, functional, and metabolic studies (4). Necropsy is the gold standard in determining the morphometric alterations in the disease state [e.g., changes in left ventricular (LV) mass]. However, there is a need for noninvasive or minimally invasive techniques in small animals that would enable serial determinations of cardiac structure and contractile function to follow disease progression and/or response to therapeutic interventions.

Echocardiography has been one of the most widely and successfully employed noninvasive techniques to provide quantitative measurements of ventricular structure and function in humans and large animals. In the rat, the heart is generally ~1 g in weight with a 2-mm LV wall thickness and a rapid heart rate of 300-400 beats/min (11). Conventional transthoracic echocardiography (TTE) transducers often do not provide optimal spatial resolution for precise morphometric measurements. Although high-frequency TTE transducers (7.5 MHz) have been used to monitor the morphometric and functional changes with experimental hypertrophy or myocardial infarction in rat models (1, 3, 8, 11), measurement reproducibility may be limited by the variability in transducer placement and/or alignment and the limits of spatial resolution of transducers in small animals. In addition, TTE cannot be performed continuously during an experimental protocol, requiring interruption and repositioning of the transducer for serial measurements. Furthermore, TTE does not provide optimal imaging of some cardiovascular structures such as the great vessels.

Transesophageal echocardiography (TEE) provides an unobstructed view of the heart and great vessels without the acoustic barriers of the ribs, skeletal muscles, or lungs (10). Once positioned, a TEE probe can provide uninterrupted monitoring of cardiac function during an extended experimental protocol. At present, TEE studies are only feasible in humans and large animals because commercially available TEE probes are too large for use in small animals. The small catheter size and high frequency of commercially available intravascular ultrasound (IVUS) catheters afford the potential for superior cardiac imaging in small animals, if such a device could be adapted for use in TEE. Also, with the transducer inserted to a known distance in the esophagus, the orientation of the imaging plane may be more reproducible than that of TTE in the rat. The spatial resolution of a 20-MHz IVUS catheter is ~0.1 mm compared with 0.2 mm for a 7.5-MHz TTE transducer. Thus many of the shortcomings encountered in rat TTE might be overcome with appropriate TEE instrumentation.

Previous studies have shown the feasibility of adapting IVUS technology to perform TEE in fetal lambs (5, 6) and intrapericardial imaging in rabbits (9). The present study demonstrates the feasibility of rat TEE with a high degree of reproducibility and without significant trauma to the animal. Furthermore, data from TTE and TEE in the same rats are compared.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

The study was performed according to the guidelines of the American Physiological Society. The protocol was approved by the Committee of Animal Research at the University of California, San Francisco, and conformed to the Position of the American Heart Association on Research Animal Use adopted November 11, 1984, by the American Heart Association. Eighteen female Sprague-Dawley rats (Simonsen, Gilroy, CA) weighing 270-370 g were used in this study (6 for pilot studies and 12 to obtain the data reported herein). All studies were performed under anesthesia with pentobarbital sodium (40 mg/kg ip).

TEE. A commercially available IVUS system (CVIS, Cardiovascular Imaging Systems, Sunnyvale, CA) with an 8-Fr, 20-MHz IVUS catheter was used. The catheter consists of a fixed piezoelectric crystal and a rotating mirror to reflect the ultrasound beam. The rotational frequency (image frame rate) is 30 Hz. There are two lumens for a guide wire and for pressure monitoring. To facilitate apposition of the catheter with the esophagus, the tip of the catheter was preshaped with a 0.018-inch-diameter guide wire to produce a gentle curvature of the distal portion of the catheter. We found that flushing of the catheter could cause reflux of fluid from the esophagus into the bronchial tree. To avoid this complication, the catheter was flushed with saline before insertion and the distal opening of the pressure lumen was sealed with wax. After lubrication with lidocaine gel, the catheter was carefully inserted into esophagus. As the catheter was slowly advanced, a long-axis view of the aortic arch, short-axis (cross-sectional) view of the ascending and descending aorta, and longaxis view of the pulmonary artery were consistently obtained. A short-axis view of the mitral valve was obtained at the base of the LV before the catheter was advanced further to acquire a short-axis view of the LV at the midpapillary muscle level. The latter view was used to obtain all measurements reported in this study. Transgastric images of the heart could not be reliably obtained. Usually, it took ~15 min to manipulate the catheter to obtain an optimal short-axis view of the LV at the midpapillary muscle level. Once obtained, the imaging plane could be maintained without further manipulation or adjustment. The length that the catheter was inserted was recorded for future repeat examination. Four of twelve rats underwent two or three TEE examinations on separate occasions 1-2 wk apart to determine the reproducibility of the technique. Three rats also underwent TEE examination with a 4.3-Fr, 30-MHz IVUS catheter. This catheter provided good images of the great vessels but had insufficient depth of field for imaging of the LV. All the TEE images were obtained at the maximum frame rate of the catheters (30 Hz).

TTE. TTE was performed preceding and following TEE in 6 of 12 rats with a Hewlett-Packard Sonos 1500 ultrasound system equipped with a 7.5-MHz pediatric cardiac phased-array transducer (Hewlett-Packard, Andover, MA). The chest was shaved, and the rats were placed either supine or prone. A two-dimensional short-axis view of the LV was obtained at the level of the papillary muscle at either a 30- or 60-Hz frame rate. The roundness of the ventricular cavity was used as the indicator of an optimal cross-sectional image (7).

Data analysis. All images were recorded on super VHS format videotape for off-line measurement. End-diastolic frames [identified by the maximal LV cavity dimension (diameter)] and end-systolic frames (identified by the minimal LV cavity dimension) were digitized by Video Monitor software (Apple Computer, Cupertino, CA) on an Apple Power Mac 7100 computer (Apple Computer). In each rat, end-diastolic and end-systolic measurements were averaged from those obtained in two to four cardiac cycles. Quantitative analysis of digitized images was measured by National Institutes of Health Image Version 1.55 software (National Institutes of Health, Bethesda, MD).

LV mass was calculated with a standard cube formula (3): LV mass (in mg) = 1.05 × (ED³_ed  ID³_ed), where 1.05 is the specific gravity of muscle, ED_ed is the external dimension of the LV at end diastole, and ID_ed is the internal dimension of the LV cavity at end diastole, both measured in millimeters.

Systolic ventricular function was quantified as fractional shortening (FS), which was calculated as FS (in %) = [(ID_ed  ID_es)/ID_ed] × 100, where ID_es is the internal dimension of the LV cavity at end systole measured in millimeters.

Necropsy. At the conclusion of all TEE and TTE examinations, the heart was arrested by intracardiac injection of 1 ml of 10% potassium chloride. After removal of the heart from the chest, the atrium, aorta, and right ventricle were excised. The remaining LV was blotted dry and weighed on a calibrated balance. The esophagus was incised with a longitudinal incision to inspect its mucosa for evidence of luminal injury from the catheter.

Reproducibility and statistical analysis. To determine interobserver and intraobserver variability, captured images from each study were measured independently by two observers (L. Lu and E. Ko), each blinded to necropsy LV mass, and twice by one observer (L. Lu). To determine the reproducibility of rat TEE, two or three serial TEE examinations were performed 1-2 wk apart in four rats. To determine the correlation between mass calculated by either TEE or TTE and actual mass determined at necropsy, the linear regression equation, correlation coefficient (r), and standard error of estimate (SEE) were determined. Differences between TEE and TTE measurements of LV dimensions and calculations of fractional shortening and mass in the same heart were determined by paired Student's t-tests. All values are shown as means ± SD.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

During six pilot studies, rats died from respiratory failure due to aspiration of fluid resulting from flushing the catheter lumen (n = 2) or an overdose of the anesthetic (n = 2). After the techniques for flushing and sealing the lumen of the catheter were refined, all 12 subsequent rats survived the TEE procedure without apparent complications, regaining normal activity and feeding. Postmortem examination of the esophagus revealed no apparent visually detectable injury of the mucosa.

Examples of TEE images of the great vessels are shown in Fig. 1. Figure 1A shows an image of the ascending and descending aorta obtained with a 4.3-Fr, 30-MHz catheter just caudal to the level of the aortic arch. Figure 1B is a short-axis view of the aortic root, left atrium, and pulmonary artery obtained with an 8-Fr, 20-MHz catheter.

View larger version (73K): [in this window] [in a new window]   View larger version (146K): [in this window] [in a new window]   Fig. 1.   A: cross-sectional view of ascending and descending aorta obtained with a 4.3-Fr, 30-MHz catheter without preshape. DAO, descending aorta; AAO, ascending aorta. B: cross-sectional view at aortic root obtained with a preshaped 4.3-Fr, 30-MHz catheter. LA, left atrium; AO, aorta; PA, pulmonary artery.

Figure 2 is a representative set of short-axis images of the LV at the midpapillary muscle level. Figure 2, A and C, shows images obtained by TEE with an 8-Fr, 20-MHz catheter at a frame rate of 30 Hz. Figure 2, B and D, shows images from the same heart obtained by TTE with a 7.5-MHz pediatric transducer at a frame rate of 60 Hz. End-diastolic images are shown in Fig. 2, A and B, and end-systolic images in Fig. 2, C and D. Endocardial and epicardial borders can be readily determined in all four images. There are echoes originating from the LV cavity in the TEE images compared with the TTE images in which the LV cavity is relatively echo free. This difference is most likely due to the greater reflection of ultrasound from blood cells at the higher frequency of the TEE catheter.

View larger version (178K): [in this window] [in a new window]   View larger version (137K): [in this window] [in a new window]   View larger version (170K): [in this window] [in a new window]   View larger version (133K): [in this window] [in a new window]   Fig. 2.   Short-axis views of left ventricle at midpapillary muscle level. A and C were obtained by transesophageal echocardiography (TEE) with an 8-Fr, 20-MHz catheter at 30-Hz frame rate; B and D were obtained by TTE in the same heart with a 7.5-MHz transducer at 60-Hz frame rate. A and B: end diastole. C and D: end systole. Top of each image is anterior aspect. PM, papillary muscles.

LV dimensions, fractional shortening, and calculated LV mass obtained by TEE and TTE, along with the actual postmortem LV mass, are shown in Table 1 for the six rats that were examined by both imaging techniques. Calculated masses were similar to those reported by De Simone et al. (3) and Pawlush et al. (11) using two-dimensional TTE. Differences in LV dimensions and calculated LV mass by TEE and TTE were minor. The linear regressions and 95% confidence intervals between TEE- or TTE-calculated and actual LV masses (Figs. 3 and 4) show good correlation with both techniques (TEE: r = 0.94, r² = 0.88, SEE = 15 mg; TTE: r = 0.88, r² = 0.77, SEE = 14 mg). The difference between TEE-estimated and actual LV masses was 7 ± 13 mg (P = 0.31), whereas the difference between TTE-estimated and actual LV masses was +14 ± 3 mg (P = 0.41) for hearts with an actual LV mass of 627 ± 18 mg.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Linear regression of TEE-calculated LV mass with actual LV mass. y = 1.02x  7.7; r = 0.94; standard error of estimate = 15 mg.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Linear regression of transthoracic echocardiography (TTE)-calculated LV mass with actual LV mass. y = 1.09x  58.8; r = 0.88; standard error of estimate = 14 mg.

There were no significant differences between TEE and TTE measurements of LV end-diastolic and end-systolic dimensions (Table 1). LV fractional shortening was 56 ± 5% as determined by TEE and 50 ± 5% as determined by TTE.

The coefficients of variation for ED_ed, ID_ed, and calculated LV mass in the four rats in which two to three serial TEE studies were performed were 0.04, 0.05, and 0.07, respectively. The intraobserver coefficients of variation for ED_ed, ID_ed, and calculated LV mass by TEE were 0.03, 0.05, and 0.08, respectively. The interobserver coefficients of variation for ED_ed, ID_ed, and calculated LV mass by TEE were 0.05, 0.06, and 0.08, respectively.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The results of this study show that TEE with commercial IVUS instrumentation in rats is feasible and provides useful information. IVUS TEE is simple to perform and has a high resolution due to the anatomic relationship between the esophagus and the heart and the high frequency of the IVUS catheter, although the near-field effect and high frequency of the TEE catheter may contribute to artifactual blood echoes (2). With the catheter advanced a fixed distance into the esophagus, a reproducible imaging plane can be achieved in serial studies. Both TEE- and TTE-calculated LV mass showed good agreement with the actual LV mass determined at necropsy. In the six rats in which both TEE and TTE were performed, the correlation between the calculated and actual LV masses was nearly identical (r = 0.89 and 0.88, respectively), and the mean difference between calculated and actual LV mass was small. The coefficients of variation of LV dimensions and derived parameters were <0.10 in serial studies in the same rats. Similarly, coefficients of intra- and interobserver variation were <0.10. These findings indicated acceptable reproducibility of the IVUS TEE technique. Therefore, TEE with IVUS instrumentation may offer an alternative technique for the accurate, serial assessment of LV dimensions, mass, and systolic function in small laboratory animals. Potential advantages of this technique include the ability to continuously monitor LV function during an experimental protocol when interruption for repeated TTE examinations would be undesirable and the reproducibility of the image plane afforded by insertion of the catheter to the same distance in serial examinations, thereby avoiding potential differences in probe placement and orientation associated with TTE. Also, TEE provides a superior ability to image the aorta and pulmonary artery of rats compared with TTE.

The present data in normal rats indicate that IVUS TEE provides measurements of LV size and function that are at least comparable to those obtainable by TTE in normal rats. Further studies will be necessary to determine if IVUS TEE is advantageous to TTE in assessing cardiac structure and function in models of heart disease in small animals.

Limitations. The catheter employed in this study is not steerable and has only one imaging plane. Consequently, even with the catheter preshaped with a guide wire, it was difficult to get standard, reproducible long-axis two- or four-chamber images of the heart, and it was not possible to obtain transgastric images comparable to TEE in humans. The crystal-mirror assembly in the catheter employed is located 1.5 cm from the catheter tip. A catheter with the transducer closer to the tip may be of help in achieving additional imaging planes. Another shortcoming of the present study is that the catheter we used has a relatively low frame rate (30 Hz) compared with the rapid heart rate in rats (~5 Hz). Ideally, a frame rate of 10 times the cardiac frequency (i.e., 50 Hz) would be optimal to determine the maximum diastolic and minimum systolic dimensions of the cardiac cycle.