Author + information
- Received June 20, 2001
- Revision received September 27, 2001
- Accepted November 7, 2001
- Published online February 6, 2002.
- Nicholas G Fisher, MBBS∗,
- Jonathan P Christiansen, MB, ChB∗,
- Howard Leong-Poi, MD∗,
- Ananda R Jayaweera, PhD∗,
- Jonathan R Lindner, MD∗ and
- Sanjiv Kaul, MD∗,* ()
- ↵*Reprint requests and correspondence:
Dr. Sanjiv Kaul, Cardiovascular Division, Box 158, Medical Center, University of Virginia, Charlottesville, Virginia 22908, USA.
Objectives This study sought to investigate the myocardial and microvascular kinetics of BR14, a novel third-generation ultrasound contrast agent.
Background BR14 produces persistent myocardial opacification after the administration of a single intravenous bolus when the left ventricular cavity contrast has considerably diminished. The mechanism of this finding is unknown.
Methods Nine open-chest dogs with non-critical stenosis of a single coronary artery were given intravenous bolus injections of BR14 during coronary hyperemia. Time versus acoustic intensity (AI) plots were generated from the normal and stenosed beds and myocardial blood flow (MBF) was measured with radiolabeled microspheres. Intravital microscopy was performed on an exteriorized cremaster muscle in 11 wild-type mice to study the microvascular kinetics of the agent.
Results At peak contrast enhancement, the ratio between AI in the stenosed and normal bed was 0.44 ± 0.23, which was similar to the radiolabeled microsphere-derived MBF ratio between the two beds (0.45 ± 0.20). At 400 s after injection, the AI ratio between the two beds approximated unity (0.99 ± 0.07) despite no changes in MBF, indicating redistribution of the agent. The myocardial kinetics of BR14 was best characterized by a modified lagged normal density function. Only about 3% of administered microbubbles were estimated to be retained in the myocardium. Intravital microscopy showed that most of these bubbles were retained only transiently (2 to 3 s) within capillaries.
Conclusions BR14 demonstrates redistribution because of transient retention within capillaries. Therefore, similar to 201Tl, it could potentially be used to detect both coronary stenosis and myocardial viability after a single injection during stress.
BR14 is a new ultrasound contrast agent that has undergone pre-clinical studies and is currently undergoing phase II studies in humans. It consists of perfluorocarbon-containing microbubbles stabilized by a phospholipid monolayer (1). When this agent is injected intravenously as a single bolus and imaging is initiated several minutes later, persistent myocardial opacification is noted at a time when the agent has mostly cleared from the left ventricular (LV) cavity. This finding implies myocardial retention of at least a portion of the agent. Our study pursued the following aims: 1) calculation of the myocardial retention fraction of the agent; 2) modeling of the myocardial kinetics of this agent in beds supplied by normal coronary arteries and those with non-critical stenosis; and 3) study of the microvascular rheology of the agent. For the first two aims we performed experiments in an open-chest canine preparation. We performed intravital microscopy of the mouse cremaster muscle for the third aim.
The study was approved by the Animal Research Committee at the University of Virginia and conformed to The American Heart Association Guidelines for the Use of Animals in Research.
Nine adult mongrel dogs were used for the study. They were anesthetized with 30 mg/kg−1of sodium pentobarbital (Abbott Laboratories, North Chicago, Illinois), intubated and ventilated with room air. Catheters were placed in both femoral arteries for withdrawal of reference samples during radiolabeled microsphere injections and in both femoral veins for administration of drugs, fluids and BR14.
A left lateral thoracotomy was performed, and the heart was suspended in a pericardial cradle. Catheters were placed in the aortic root as well as the right and left atria for pressure measurements. The left atrial catheter was also used for the injection of radiolabeled microspheres. The proximal sections of the left anterior descending coronary artery (LAD) and left circumflex coronary artery (LCx) were dissected from the surrounding tissue. Ultrasonic flow probes (series SC, Transonics, Mount Pleasant, Wimbley, UK) were placed on both arteries and connected to a digital flow meter (model T206, Transonics) to monitor epicardial coronary blood flow. A custom-designed screw occluder was also placed around the coronary arteries to enable creation of coronary stenoses of varying severities. A 20-gauge catheter (Becton-Dickinson) was introduced into the distal branch of the coronary artery to measure distal coronary pressure and determine the severity of the stenosis.
Myocardial contrast echocardiography
Myocardial contrast echocardiography (MCE) was performed using power pulse-inversion imaging with a HDI 5000 system (Philips-ATL, Bothell, Washington). The ultrasound transducer (transmit frequency of 1.67 and receive frequency of 3.3 MHz) was fixed in position within a saline bath that served as an acoustic interface between the transducer and the heart. Imaging was performed at the mid-papillary muscle short-axis level caudal to the occluder. The transmit power, image depth and color gains were optimized at the beginning of each experiment and held constant throughout. The focal plane was placed at the level of the LV posterior wall. A pulse repetition frequency of 2,500 Hz and a mechanical index of 0.1 were used. Only end-systolic images (gated to the T wave on the electrocardiogram) were acquired and stored digitally using HDI lab. This system allows direct measurement of AI in linear units before log compression and post-processing.
BR14 (Bracco Diagnostics, Geneva, Switzerland) was administered as a controlled 2-ml bolus over 30 s using a power injector (Medrad, Pittsburgh, Pennsylvania). This is a third-generation ultrasound contrast agent consisting of perfluorocarbon-containing microbubbles stabilized by a phospholipid monolayer. The mean diameter of the microbubbles is 2.5 to 3.0 μm, and their mean concentration is 2 to 5 × 108ml−1(1). Every end-systolic frame was captured for the first 30 s after the injection. The pulsing interval was then increased to every 10 cardiac cycles for 1 min and to every 20 cardiac cycles for the next 10 min in order to minimize microbubble destruction. After alignment of all images from a single injection sequence (2), regions of interest were placed over the LAD and LCx beds for measurement of AI, which was then plotted against time. The resultant curves were smoothed using a moving average function.
Radiolabeled microsphere analysis
Myocardial blood flow (MBF) was measured using left atrial injections of ∼2 × 106–11 μm radiolabeled microspheres (Dupont Medical Products, Wilmington, Delaware) suspended in 4 ml of 0.9% saline and 0.01% Tween-80. Duplicate arterial reference blood samples were collected from the femoral arteries. At the end of the experiment, the post-mortem LV short-axis slice corresponding to the MCE image was cut into 16 wedge-shaped pieces, and each piece was further divided into epi-, mid- and endocardial portions. The tissue and blood samples were counted in a gamma well scintillation camera with a multichannel analyzer (model 1282, LKB Wallace, Washington, DC), and corrections were made for activity spilling from one energy window to another with custom-designed software (3).
Myocardial blood flow to each segment was calculated from the equation Qm=(Cm·Qr)/Crwhere Qmis blood flow to the segment (ml/min−1), Cmis tissue counts, Qris the rate of arterial blood sample withdrawal (ml/min−1) and Cris arterial reference sample counts (4). Transmural MBF (ml/min−1/g−1) to each of the 16 wedge-shaped pieces was calculated as the quotient of the summed flows to the individual segments within that piece and their combined weight. Mean MBF to both beds were then calculated by averaging the transmural MBF in all segments in those beds.
Noncritical stenoses of varying severities (judging from the trans-stenotic pressure gradient) were placed on either the LAD or LCx. Maximal coronary hyperemia was induced with a continuous infusion of 1 μg/kg−1/min−1of a selective adenosine A2areceptor agonist-2-cyclohexyl-methyl-idenehydrazino-adenosine (WRC-0470). Radiolabeled microspheres were injected, a bolus of BR14 was administered, and MCE was performed. Post-mortem, the heart slice corresponding to the MCE image was removed for tissue staining for presence of infarction and radiolabeled microsphere MBF analysis.
Intravital microscopy experiments
Eleven mice were anesthetized with an intraperitoneal injection (12.5 μl/g−1) of a solution containing ketamine hydrochloride, xylazine and atropine. Body temperature was maintained at 37°C with a heating pad. A jugular vein was cannulated for the administration of microbubbles and drugs. The right or left cremaster muscle was exteriorized, a longitudinal incision was made in it, and the edges were secured to a translucent custom-designed stage. The preparation was superfused continuously with isothermic bicarbonate-buffered saline (5).
Intravital microscopy was performed with an Axioskop2-FS microscope (Carl Zeiss, Inc., Thornwood, New York) using a saline immersion objective (X20/0.5 N.A., X40/0.8 N.A. or X 63/0.9 N.A.) under combined trans- and fluorescent epi-illumination with a 469–500 nm excitation filter. Video recordings were made with a high-resolution charge-coupled device camera (C2400, Hammamatsu Photonics, Bridgewater, New Jersey) interfaced with a video time display unit and connected to an S-VHS recorder.
Arteriolar blood velocities were measured with a dual-slit photodiode (6)and converted to mean blood velocities (Vb) by multiplication by 0.625 (7). Shear rates (γw) were determined by γw= 2.12(8 Vb)/d, where d is the vessel diameter and 2.12 is a correction factor for the shape of the velocity profile.
Video recordings of arterioles feeding the tissue were made, and their diameter and blood velocity measurements were calculated at baseline. Approximately 4 × 107fluorescently labeled microbubbles (provided by the manufacturer) were then injected intravenously as a slow bolus. In five mice, random optical fields were observed at ×40 and ×63 for 30 min to determine the fate of the microbubbles. In three mice, 20 optical fields (×40) were sequentially observed at 2, 10, 20 and 30 min to determine the temporal behavior of the initially stationary bubbles. In three additional mice, the preparation was observed continuously for 20 min under ×20 objective. The arrival times and durations of retention were recorded for each microbubble. Video recordings as well as arteriolar velocity and diameter measurements were repeated at 10-min intervals.
Data are expressed as mean ± 1 standard deviation. Differences between >2 stages were assessed using analysis of variance. When differences were found, any two stages were compared using Student’s ttest with Bonferroni correction. Correlations were calculated using least-fit regression analyses. Differences were considered significant at p < 0.05 (two-sided).
Figure 1illustrates a time-versus-AI plot from the LAD and LCx beds from a dog with a non-critical LAD stenosis during maximal hyperemia induced by WRC-0470. The initial and delayed images are also shown. The AI in the stenosed bed is less than the normal bed during the peak contrast effect of BR14, a finding that has been reported for all intravenous ultrasound contrast agents. The unique findings of BR14 are that myocardial opacification persists at a time when microbubbles have all but cleared from the LV cavity and that, despite ongoing hyperemia, the AI in both beds becomes identical over time. This latter finding is very similar to that seen after 201Tl injection and is termed “redistribution” (8). That is, there is faster washout of the tracer from the bed with higher initial microbubble concentration and slower washout from the bed with lower initial microbubble concentration. Although there is less hyperemic flow in the region showing less AI initially, the presence of redistribution indicates that the myocardium distal to the stenosis is normal (not infarcted) (8).
At peak contrast enhancement, the ratio between AI in the stenosed and normal bed in all nine dogs was 0.44 ± 0.23, which was similar to the radiolabeled microsphere-derived MBF ratio between the two beds (0.45 ± 0.20). Despite no change in MBF in the two beds, however, the AI ratio 400 s later between the two beds was 0.99 ± 0.07, indicating redistribution of BR14. No infarction was seen in these dogs on post-mortem tissue staining.
Microbubble injections were well tolerated in all mice, and no significant changes in arteriolar flow or shear rates were observed after injection (Table 1). Within 2 min of injection, stationary microbubbles (defined as those retained at a single site for >5 s) were observed in both capillaries and venules. The number of the injected microbubbles that were stationary was very small compared with the number showing normal flux. Early after injection, most stationary microbubbles appeared to be lodged in capillaries, whereas by 30 min most stationary microbubbles were found in venules (Fig. 2). Venular retention resulted from the adherence of microbubbles to activated leukocytes, which has been seen in this experimental model with other agents as well (8)and can, in part, be explained by activation of leukocytes from surgical trauma (5).
The unique feature of BR14 is its retention within the capillaries, which has the following characteristics. First, very few of the injected bubbles are retained in the capillaries (Fig. 3). Second, this retention is transient. Figure 4illustrates the time course of individual bubble retention from a single experiment when continuous observations were made at ×20 magnification. Third, the stationary microbubbles retain a uniform spherical shape and, unlike other bubbles, produce no distortion of the capillaries in which they were retained. Fourth, the diameter of stationary microbubbles does not change over time (3.9 ± 0.5 μm vs. 4.6 ± 0.7 μm for observations at 0 to 10 and 10 to 20 min, p = 0.42), indicating that neither entrapment nor dislodgment is related to bubble size. Fifth, in the majority of cases, microbubble retention in a capillary results in temporary cessation of red cell flux through that vessel, although slow red blood cell flux can still persist around the microbubble. Sixth, the dislodged microbubbles are frequently observed to stagger through the capillaries, which is also seen with freely circulating microbubbles with retention times <5 s. Finally, microbubbles that persist for more than 20 min are commonly phagocytosed by leukocytes resident within the capillaries, resulting in restoration of red blood cell flux.
We have described the myocardial and microvascular kinetics of BR14, a novel third-generation ultrasound contrast agent. We believe that the unique properties of this agent will allow it to be used like 201Tl. That is, after a single injection at peak stress, early and delayed images will provide information on both the presence of coronary stenoses as well as myocardial viability. Because of its myocardial retention, we believe that this agent could be used during exercise stress with imaging performed after the stress protocol is complete. This agent could, therefore, not only provide information on myocardial perfusion but also enhance regional function assessment during exercise stress.
Myocardial kinetics of BR14
Because the lung acts as a large mixing chamber, a venous bolus injection of a contrast agent will spread in a normal or Gaussian form when the contrast reaches the coronary arteries (Fig. 5A). This Gaussian input function is of the formwhere ais the area under the curve that depends on the amount of contrast injected, and σ is the width of the curve, which depends mostly on cardiac output if duration of injection is short.
The output function of the mixing chamber in this model can be derived by two methods. one can either use a convolution of the input and transfer functions (9)or solve the following differential equation:where(vis the volume of the mixing chamber, fis the flow through it, and ctis the concentration of contrast in the myocardium). the solution to this equation, which does not have an analytic form, is termed a lagged-normal density function (lndf) (10)and is obtained with intravenous contrast agents that are not retained within the myocardium (fig. 5b) (11).
If after administration, all microbubbles were retained within the myocardium (as in the case of radiolabeled microspheres), the ai would be represented by the integral of a lndf (lndfi). if only a fraction (f1) of the microbubbles are retained (fig. 1), the resulting ai attributable to retained bubbles can be given bywhere f2is the myocardial blood volume fraction (percent of myocardial mass that is blood).
Therefore, the ai due to both stuck and free bubbles is given bywhere
This model accurately fits the experimental time-versus-ai curves generated in the open-chest dogs (fig. 6). on the basis of this model, the fraction of br14 being retained in the myocardium is very small (≃3%).
Comparison with other contrast agents
Most second-generation ultrasound contrast agents behave like red blood cells in vivo and are not retained in the normal microcirculation (8,12,13). Myocardial redistribution of an ultrasound contrast agent in normal myocardium was first reported with AIP-201, a 10 μm air-filled bubble with a thick albumin shell (14). Because of its size, this agent could be injected only into the arterial system. It was found that during coronary occlusion, a left atrial injection of this agent resulted in no opacification of the occluded bed, but opacification was noted within minutes after perfusion was re-established, without performing a second injection of contrast, and at a time when there was no contrast in the LV cavity. Subsequently, using intravital microscopy, we found that these bubbles entered arterioles supplying the occluded bed via collateral channels. When flow was re-established, the bubbles dislodged from these feeding arterioles and went to regions with adequate microvascular flow. We found that redistribution occurred only to normal, non-infarcted tissue and not to infarcted regions with no reflow (14). Because of the limited use of ultrasound contrast agents in the cardiac catheterization laboratory, however, AIP-201 is no longer under clinical development.
Persistent myocardial opacification from a single venous injection was also seen with Echogen, an ultrasound contrast agent produced using the “phase shift” technology, where a liquid at room temperature turns to gas at body temperature (15). The problem with this agent was that the size of the microbubble could not be controlled in vivo and the microbubbles became larger in blood from diffusion of air normally present in blood into the non-encapsulated perfluorocarbon emulsion. This agent also caused complement activation. Because of safety concerns, this agent is also no longer in clinical development.
BR14 overcomes the problems associated with its two predecessors. First, it consists of stabilized microbubbles of a small size (1). All pre-clinical studies have attested to the safety of this agent administered in very large amounts to animals. Second, unlike the larger bubbles that were entrained within larger vessels (arterioles and venules), these microbubbles are transiently retained within capillaries. Therefore, their presence in capillaries minutes after injection indicates capillary viability. Third, its kinetics are identical to 201Tl, an agent that has been used clinically for more than 20 years and has provided important information in patients with coronary artery disease. Thus a clinical paradigm already exists for its use.
From this study, we cannot explain the mechanism of transient microbubble retention within the capillaries. It cannot be explained on the basis of size, because these bubbles do not seem to plug capillaries. Instead, they are transiently retained before dislodgment, with no change in capillary dimensions either at or proximal to the site of entrapment. Another mechanism may be a rigid shell preventing easy capillary passage. This is unlikely because these bubbles have successfully transited the pulmonary circulation and dislodge easily from the site of capillary retention. Bubble charge and coating characteristics are other factors that are under investigation in our laboratory. Finally, the results of findings in the skeletal muscle may not completely explain those noted in the beating canine heart.
It is obvious from our results that only a fraction of the microbubbles is retained within the myocardium. Myocardial opacification, therefore, is orders of magnitude lower in delayed compared with initial images (Fig. 1). We therefore propose the following imaging protocol with this agent. The initial images should be obtained using low-mechanical-index intermittent imaging (once every cardiac cycle). Both the low-power and intermittent protocols will minimize microbubble destruction. The delayed image should be obtained using a high mechanical index. In this manner, myocardial signal will be good in the delayed image despite a lower number of bubbles present in the myocardium. These images could then be displayed side by side as shown in Figure 1. Because a number of images would have been recorded, time-versus-AI plots from several beds could also be derived.
We feel that that the best application of this imaging protocol would be during exercise stress where a single injection could be performed at peak stress. Initial images could then be obtained within 1 to 3 min after cessation of exercise, and delayed images could be obtained 5 to 7 min later. If the delayed images are inadequate, a second injection of the agent could be made at rest. This protocol would therefore address both coronary disease detection and myocardial viability from a single bolus injection. The mode of administration (bolus rather than continuous infusion ) is also suitable for exercise, compared with other forms of stress.
☆ Supported in part by grants (3RO1-HL-48890) from the National Institutes of Health, Bethesda, Maryland, and Bracco Research SA, Geneva, Switzerland. Dupont Medical Products (Wilmington, Delaware) provided the radiolabeled microspheres and the ultrasound equipment was provided by Phillips-ATL, Bothell, Washington. Drs. Christiansen and Leong-Poi are the recipients of Fellowship Training Grants from the Mid-Atlantic Affiliate of the American Heart Association, Baltimore, Maryland, and the Canadian Institutes of Health Research and the Heart and Stroke Foundation of Canada, Ottawa, Canada, respectively. Dr. Lindner is the recipient of the Mentored Clinical Scientist Development Award (K08-HL03909) from the National Institutes of Health.
Presented in part at the 50th Annual Scientific Session of the American College of Cardiology, March 2001, Orlando, Florida.
- acoustic intensity
- left anterior descending coronary artery
- left circumflex coronary artery
- lagged normal density function
- left ventricular
- myocardial blood flow
- myocardial contrast echocardiography
- Received June 20, 2001.
- Revision received September 27, 2001.
- Accepted November 7, 2001.
- American College of Cardiology
- Wei K,
- Ananda J.R,
- Firoozan S,
- Linka A,
- Skyba D.M,
- Kaul S
- Heyman M.A,
- Payne B.D,
- Hoffman J.I,
- Rodolf A.M
- ↵Lindner JR, Song J, Jayaweera AR, Kaul S. Microvascular rheology of a second-generation gas-filled microbubble: Implications regarding safety. J Am Soc Echocardiogr. In Press
- Pries A.R
- Pohost G.M,
- Okada R.D,
- O’Keefe D.D,
- et al.
- Jayaweera A.R,
- Edwards N,
- Glasheen W.P,
- Villanueva F.S,
- Abbott R.D,
- Kaul S
- Bassingthwaighte J.B,
- Ackerman F.H,
- Wood E.H
- Skyba D.M,
- Jayaweera A.R,
- Goodman N.C,
- Ismail S,
- Camarano G.P,
- Kaul S
- Keller M.W,
- Segal S.S,
- Kaul S,
- Duling B
- Skyba D.M,
- Camarano G,
- Goodman N.C,
- Price R.J,
- Skalak T.C,
- Kaul S
- Linka A.Z,
- Skyba D.M,
- Price R.J,
- Wei K,
- Skalak T.C,
- Kaul S
- Grayburn P.A,
- Erickson J.M,
- Escobar J,
- Womack L,
- Velasco C.E