Author + information
- Andrew J Hamilton, MD*,* (, )
- Shao-Ling Huang, PhD†,
- Drew Warnick, BS*,
- Mark Rabbat, BS*,
- Bonnie Kane, BS*,
- Ashwin Nagaraj, PhD*,
- Melvin Klegerman, PhD‡ and
- David D McPherson, MD*
- ↵*Reprint requests and correspondence:
Dr. David D. McPherson, Department of Medicine, Division of Cardiology, Northwestern University, 251 East Huron, Galter 8-230, Chicago, Illinois 60611-2908, USA.
Objectives Our purpose was to quantitate and confirm specific echogenic immunoliposome (ELIP) atheroma component enhancement in vivo.
Background Targeted ELIPs for ultrasonic detection and staging of active molecular components of endothelium and atheroma have been developed.
Methods In Yucatan miniswine, the endothelium was injured from one femoral and one carotid artery, and animals were fed a high-cholesterol diet for two months to create various stages of atheroma. Arteries were imaged with intravascular ultrasound (IVUS) 5 and 10 min after ELIP injection (5-mg dose). Anti-intercellular adhesion molecule-1 (ICAM-1), anti-vascular cell adhesion molecule-1 (VCAM-1), anti-fibrin, anti-fibrinogen, and anti-tissue factor (TF) conjugated ELIPs were used, and immunohistochemistry (IHC) confirmed the presence or absence of molecular expression. Two blinded observers determined if each segment was enhanced by ELIP. Three-dimensional image reconstruction and videodensitometric analysis determined the mean gray-scale (MGS) change of the luminal border.
Results To determine endothelial injury component enhancement, anti-fibrinogen ELIP enhanced exposed fibrin in all arteries (MGS increased 22 ± 5%; 6 arteries; 2 animals). To determine enhancement of molecular components in atherosclerotic arteries, observers detected enhancement 5 min after anti-VCAM, anti-ICAM, anti-TF, anti-fibrin, and anti-fibrinogen conjugated ELIPs. Furthermore, ELIP enhanced atheroma MGS by 39 ± 18% (n = 8). The IHC staining confirmed the expression of respective molecular targets in all enhanced segments.
Conclusions It was shown that ELIPs specifically enhance endothelial injury/atheroma components. This allows better characterization of the type and extent of active atheroma components and may allow more directed therapy.
We have developed targeted echogenic immunoliposomes (ELIPs) for ultrasound detection and enhancement of vasoactive and pathologic molecular components of endothelium and atheroma (1–3). These ELIPs are small (<1 μm) ultrasound contrast agents, produced in the absence of added air, that can be used to evaluate molecular components of vasoactive and pathologic endothelium/atheroma (4). Currently available ELIPs are composed of four primary lipids: phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylglycerol (PG), and cholesterol and are made using a lyophilization process (5). Once produced, the liposomes are conjugated to specific antibodies that enable targeted ultrasound enhancement of molecular components of atheroma and other tissues (4,6). Detection of molecular components may allow staging of atheroma and lead to the development of targeted therapeutic interventions.
At the endothelial surface, molecules are expressed which, if detected, can help determine the stage or degree of atherosclerosis. For example, endothelium overlying the shoulder region of atherosclerotic lesions expresses vascular cell adhesion molecule-1 (VCAM-1) and intercellular adhesion molecule-1 (ICAM-1) (7). In the early stages of plaque formation, monocytes are recruited to the injured endothelial cells by VCAM-1, an important marker of endothelial cell activation (8). VCAM-1 is expressed on the endothelium of mid-stage atheroma, and its expression declines as atheroma progresses (9). ICAM-1 has been shown to enhance recruitment of circulatory monocytes and is expressed on the surface of the endothelium in early-stage atheroma (10,11). Fibrin expression within atheroma in clefts and at the surface of the endothelium is a marker of instability of the atherosclerotic plaque and thrombogenesis (12,13). Tissue factor (TF) has also been identified as a marker of thrombosis formation (14).
The purpose of this study was to demonstrate and confirm intravascular ultrasound (IVUS) enhancement of molecular components of atheroma using targeted ELIPs in a single-step process that would allow staging of atheroma. Additionally, enhancement of atheroma components under physiologic flow conditions and confirmation by immunohistochemical (IHC) analysis could be evaluated.
The materials were provided as follows: anti-intracellular adhesion molecule-1 (Ab 1) (Labvision, Fremont, California) and anti-VCAM-1 (Ab 3) (Dako, Carpinteria, California); rabbit anti-human anti-fibrinogen (#313R), monoclonal anti-fibrin (#350), and anti-TF (American Diagnostica, Greenwich, Connecticut); Fogarty embolectomy balloon 4F catheter (Edwards LifeSciences, Irvine, California); and CVIS 20-MHz IVUS imaging catheter (Boston Scientific, Inc., Sunnyvale, California).
Echogenic liposomes were made using the component lipids: cholesterol, PC, PG, and PE (2). Echogenic liposomes were conjugated to anti-fibrin, anti-fibrinogen, anti-ICAM-1, anti-VCAM-1, and anti-TF (15).
Conjugation efficiency was determined by quantitative immunoblot and Lowry protein assays (16). Conjugation efficiency was 7.7 μg antibody (Ab)/mg ELIP lipid (20.1 molecules Ab/attamole lipid) for mouse monoclonal anti-human fibrin Ab (MAb), 12.5 μg Ab/mg lipid (32.6 molecules/amole lipid) for rabbit polyclonal anti-human fibrinogen, 11.2 μg Ab/mg lipid (29.2 molecules/amole lipid) for anti-ICAM-1 MAb, 18.2 μg Ab/mg lipid (47.5 molecules/amole lipid) for anti-VCAM-1 MAb, and 8.4 μg Ab/mg lipid (21.9 molecules/amole lipid) for anti-TF MAb.
The binding affinity (Kassoc) of rabbit anti-fibrin(ogen) ELIP for porcine fibrin was 7.4 × 107M−1(avidity index = 2.4 × 109M−1amole−1). Binding was specific, as it was >90% inhibited by the addition of soluble fibrinogen during Ab-ELIP incubation. Avidity and Kassochave been reported (15,17,18). We have demonstrated that the ELIP particle size after conjugation remained <1 μm and, in most cases, remained between 500 and 800 nm (3).
The animal protocol was approved by the Animal Care and Use Committee of Northwestern University. To create atheroma in Yucatan miniswine (20 to 25 kg), a left femoral arteriotomy was performed, a Fogarty embolectomy balloon catheter was inserted into the inguinal ligament, and a 3-cm pullback through the artery was performed to injure the endothelium. A left common carotid arteriotomy was performed, and the arterial endothelium was similarly injured.
After endothelial injury, the miniswine were fed a high-cholesterol diet (Harlan Teklad test diet) for eight weeks. At the second surgery, cholesterol levels ranged from 300 to 600 mg/dl (mean 510). In each artery, a 6F sheath was inserted at least 5 cm distal to the area of imaging for IVUS.
To determine visual endothelial enhancement in two animals, balloon endothelial injury was performed in five arteries to increase fibrin/fibrinogen expression. Anti-fibrinogen conjugated ELIPs were injected interarterially with saline injections as controls. An observer was asked to detect visually enhanced segments.
Atheroma component determination
In five animals, after atheroma formation, to determine expression of the molecular targets (VCAM-1, ICAM-1, fibrin, fibrinogen, or TF), ELIPs conjugated to either anti-VCAM-1, anti-ICAM-1, anti-fibrin, anti-fibrinogen, or anti-TF were injected interarterially. Saline and unconjugated ELIP were controls. Staining by IHC confirmed the presence of molecular targets.
To reduce non-specific binding and enhancement of previous ELIP injections to another artery, the type of Ab-conjugated ELIP injected was varied. If no enhancement was seen with the initial ELIP choice, then a second ELIP was chosen and IHC staining was performed for both molecules.
The dose of conjugated and unconjugated ELIP was calculated based on the weight of lipid in each sample. The dose of ELIP used for each artery was 5 mg of lipid. Each artery was imaged before and after each injection (at 5 and 10 min).
Intravascular ultrasound imaging
Imaging was performed with a 20-MHz, high-frequency IVUS imaging catheter attached to an automatic pullback device that withdraws the catheter at 0.5 mm/s. The ileo-femoral arteries were imaged along their length, and the left and right common carotid arteries were imaged from the level of the aorto-carotid junction to the sheath. Instrument settings for gain, zoom, compression, and rejection levels were set so that the luminal border was just visibly detectable at baseline for each artery. These settings remained constant throughout imaging. Images were recorded onto S-VHS videotape in real-time for playback and image analysis.
The IVUS images 5 min after unconjugated or conjugated ELIP injection were shown simultaneously on two monitors. After every centimeter (20 s), the observers were asked if the injured endothelium/atheroma was enhanced (+), not enhanced (−), unsure (+/−), or uninterpretable (n/a). Enhancement was defined as a visually brighter or thicker endothelial interface, compared with the control.
Images acquired onto videotape were digitized to 640 × 480-pixel spatial resolution (∼0.045 mm/pixel) and 8-bit (256 levels) gray-scale resolution and reconstructed into a three-dimensional matrix using EchoPlaque software (version 1.58, Indec Systems, Mountain View, California). Six longitudinal sections of each artery at 0, 30, 60, 90, 120, and 150 degrees were reconstructed.
Areas of enhancement were identified for the control and ELIP injections. Six 3-dimensional images at 0, 30, 60, 90, 120 and 150 degrees were selected. Each three-dimensional image had two endothelial surfaces. For each three-dimensional image, an area of interest was created over each endothelial-blood interface. The area of interest was drawn between the endothelium/atheroma and external elastic membrane. A gray-scale histogram of the endothelial/atheroma-blood interface was measured, and the mean gray-scale (MGS) value per pixel was calculated for each image using Image-Pro Plus software (version 4.1, Media Cybernetics, Silver Spring, Maryland).
The carotid and femoral arteries were removed, and each arterial section was cut into seven equal segments and placed into liquid nitrogen and frozen. Each segment was labeled 1 through 7 (1 = distal; 7 = proximal). The arterial segments were cut at 4 to 5 μm. Sections were fixed in acetone and air dried for 15 to 30 min.
Staining by IHC was performed according to the Dako EnVision IHC protocol (19). Antibodies included rabbit anti-human fibrinogen, rabbit anti-human fibrin, monoclonal anti-ICAM/CD54, anti-TF, and anti-VCAM-1 (Dako Ab, all at 1:100 dilution). One artery had staining performed for VCAM-1 and TF due to non-enhancement by TF-conjugated ELIP after the first injection. A blinded pathologist identified IHC enhancement of the target molecule (expressed [+], not expressed [−], or unsure/not assessable [n/a]). Figure 1demonstrates an IHC-ICAM-stained slide in an early atheroma-induced left femoral artery. Expression of ICAM occurs both in the intima/endothelium and within the adventitia. In nearly all positively stained segments, there was evidence of IHC staining in the dense adventitia, indicating inflammatory/atheroma components.
Data were analyzed using Sigma Stat, version 2.03 statistical software (SPSS Inc, Chicago, Illinois). The results are expressed as the mean ± SD. Differences between groups were analyzed using one-way analysis of variance with the Student-Newman-Keuls method. A p value <0.05 was considered statistically significant.
IVUS detection of anti-fibrinogen ELIP attached to injured endothelium
In the first two animals, blinded observers were able to visibly detect injured endothelial enhancement by anti-fibrinogen ELIP, 5 min after intra-arterial injection, in six of eight arteries. Figure 2demonstrates IVUS images of a right femoral artery, showing enhancement of injured endothelium by anti-fibrinogen conjugated ELIP. On the left is an image after saline injection and on the right 5 min after anti-fibrinogen ELIP (8 mg). With ELIP enhancement, the intima is better identified, separated from the black media and the outer white adventitia, which is also more clearly identified. Blooming from the ELIP intimal enhancement makes the media smaller than in actuality, although separation of the intima, media, and adventitia remains.
Videodensitometric analysis of balloon-injured endothelium
The MGS changes after anti-fibrinogen ELIP injection at 5 min are shown in Table 1. The MGS increase of all injured segments observed to be enhanced by ELIP versus saline was 22 ± 5%, and that of all non-injured, non-enhanced segments was −2 ± 6% (p < 0.001).
Atheroma enhancement by ELIP/molecular expression
Visual detection of atheroma after conjugated ELIPs
Visible detection of atheroma enhancement with IVUS was seen 5 min after intra-arterial injection of anti-VCAM, anti-ICAM, anti-TF, anti-fibrin, and anti-fibrinogen conjugated ELIPs. The observers saw no difference between the 5- and 10-min time points. In 30 segments (9 arteries), atheroma enhancement by conjugated ELIP was seen. After unconjugated ELIP injection, the observers reported no enhancement of any of the 140 segments assessed. Figure 3demonstrates the IVUS images of a left carotid artery, showing atheroma expressing VCAM by anti-VCAM conjugated ELIP, and Figure 4demonstrates the IVUS images of a right femoral artery, showing enhancement of atheroma expressing ICAM by anti-ICAM conjugated ELIP (part a, after saline injection; b, 5 min after unconjugated ELIP [8 mg]; and c, 5 min after conjugated ELIP [8 mg]). Figure 4demonstrates adventitial enhancement of anti-ICAM conjugated ELIP in addition to intimal highlighting, as confirmed by IHC staining. As in Figure 2, in both of the Ab-conjugated ELIP images, the intima is clearly outlined and separated from the black intima, although some blooming from the intimal enhancement makes the media smaller than in actuality. There was variable highlighting of the adventitia with the various Ab-conjugated ELIPs, compatible with variable inflammatory component deposition in the dense adventitia and ELIP access to the dense adventitia via the vasovasorum.
Videodensitometric analysis of ELIP atheroma enhancement
The IVUS MGS values at 5 min for visually enhanced segments are outlined in Table 2. Saline and unconjugated ELIP were used as controls. Nine arteries had segments visually observed to be enhanced. In one artery, the MGS of conjugated ELIP, and in two arteries, the MGS of unconjugated ELIP, could not be assessed reliably due to stagnant flow obscuring the endothelium. All enhanced segments had MGS different from saline and unconjugated ELIP. The combined average MGS change of all the enhanced segments after conjugated ELIP injection versus saline was 39 ± 18% (p < 0.001; n = 8 arteries), compared with the combined average MGS change of the unconjugated ELIP versus saline of 1 ± 4% (p = NS; n = 6 arteries).
The IVUS MGS values for non-enhanced segments at 5 min are presented in Table 3. Seventeen arteries contained at least one segment not observed to have been enhanced after ELIP injection. The MGS values of all segments after unconjugated ELIP and conjugated ELIP were not different from those with saline. The average MGS change of the non-enhanced segments after conjugated ELIP injection versus saline was 4 ± 6% (p = NS; n = 17 arteries), compared with the average MGS change of the unconjugated ELIP versus saline (1 ± 7%; p = NS; n = 14 arteries).
Enhancement by IVUS imaging versus IHC molecular expression
A comparison of ELIP enhancement and IHC molecular expression is shown in Table 4. All arteries observed to have segments that were enhanced by ELIP had corresponding IHC expression of the target molecule. In nine of 11 arteries that did not have ELIP atheroma enhancement, no expression of the target molecule was detected. Two of 11 arteries that did not have ELIP atheroma enhancement had expression of the target molecule in one (VCAM-1) and three (ICAM-1) segments, respectively.
We have demonstrated and confirmed by IHC that acoustically reflective ELIP can produce rapid, specific ultrasound enhancement of molecular endothelial/atheroma targets. Our discussion focuses on the implications of our results, factors influencing our results, and potential applications.
We have previously quantified the ELIP enhancement of fibrin in vitro by IVUS imaging (20). We have qualitatively demonstrated in vivo non-specific acoustic image enhancement with anti-fibrinogen and anti-ICAM conjugated ELIPs using transvascular and IVUS (4). We have performed previous experiments to demonstrate the link between Ab-labeled liposomes and fibrin using scanning electron microscopy (3). We have demonstrated in a flow chamber that 1) anti-fibrinogen conjugated liposomes retain their adherence to fibrin under physiologic sheer/flow (17)and 2) ELIPs retain their echogenic properties after 9 min of physiologic sheer (20). These studies provide evidence for ELIP targeting under physiologic conditions while retaining echogenic characteristics.
Our previous studies involved a more aggregated preparation of conjugated ELIP. By examining the components and physical properties of ELIPs, we have altered our original ELIP preparation and developed a new preparation that is subject to less aggregation (unpublished data) and also has higher acoustic reflectivity (21).
Other investigators have been studying ultrasound contrast agents for targeted highlighting. Lanza et al. (22–24)first demonstrated that a multi-step, acoustically active biotinylated, lipid-coated, perfluorocarbon nanoemulsion, as an ultrasound contrast agent, could be successfully targeted to dog thrombi in vivo. They have also characterized this contrast agent and demonstrated that it can infiltrate arterial walls and localize TF expression (25).
The time frame of image enhancement with our ELIP preparation is more rapid than others have previously described with other contrast agents requiring multiple-step processes, which require 20 to 30 min to produce discernable enhancement (24,28). The rapid enhancement seen with ELIP may be partly due to the single-step targeting. Enhancement continued at least to 10 min after injection, demonstrating continuing enhancement over time. This implies ongoing binding of ELIP under physiologic flow conditions. The maintenance of echogenicity also demonstrates structural stability of the ELIPs once they attach to the target.
Some limitations and unexpected observations were noted. In a few arteries, the endothelium/atheroma was unable to be visualized due to either slow flow obscuring the image or occlusive disease making the area of disease difficult to see if immediately adjacent the IVUS catheter. Additionally, interpretation of the IHC staining in some segments was difficult because of preparation artifacts.
The choice of control was difficult. We have demonstrated in previous work that non-specific binding of immunoglobulin G to fibrin was enough to produce attachment to target molecules. In vitro, this can be controlled with bovine gamma globulin that blocks non-specific binding. In vivo, the cross reactivity of nearly all unrelated Abs in this animal model has not been assessed. As a result, controlling for unrelated Ab cross reactivity with the specific molecular target was very difficult. For this reason, unconjugated ELIPs were used as controls. The lack of enhancement by conjugated ELIP in the absence of positive IHC staining confirms minimal non-specific binding.
We were fortuitous in finding that the rabbit anti-fibrinogen conjugated ELIP not only binds fibrin specifically, but can also be injected intravenously to target left ventricular thrombi (25). We have also demonstrated no coagulation of the rabbit anti-fibrinogen conjugated ELIP in the presence of free fibrinogen. Although useful for our experiments, we will have to use a monoclonal fibrin-specific Ab for ELIP targeting in humans.
Due to histologic processing artifacts and shrinkage, we were unable to make an exact spatial correlation between IVUS images and IHC sections. This reduced the ability of the protocol to quantitate specificity and sensitivity with respect to IHC binding. Despite this, enhanced segments did appear to correspond to IHC staining data, and non-enhanced segments corresponded to non-IHC staining within each arterial segment, which provides good evidence of molecular imaging. Two arteries did express target molecules in one and three segments but were not enhanced by ELIP. Slow luminal flow in these two arteries may have obscured IVUS atheroma visualization. Further studies will need to be conducted to demonstrate the potential of ELIPs to provide molecular atheroma imaging when injected intravenously and to quantitate molecular atheroma enhancement using transcutaneous ultrasound.
The implications and potential applications of these data are important. Echogenic immunoliposomes may now be developed for use in the cardiac catheterization laboratory to stage atheroma. Recent advances by others in advanced IVUS image analysis, combined with these targeted molecular contrast agents, may provide powerful tools to better localize and quantitate plaque components (29,30). This includes characteristics of plaques (e.g., fibrinous craters, lipid lakes under small rims of atheroma) that may make them vulnerable to plaque rupture and embolism. The ability to stage atheroma components may facilitate application of more directed therapy, as non-echogenic liposomes have already been evaluated as drug (31)and gene delivery systems (32). Recent studies have demonstrated other cardiovascular uses of targeted ELIPs for transthoracic imaging of left ventricular thrombi (15). Furthermore, targeted ultrasound contrast agents are not limited to cardiovascular medicine, and by interchanging Abs, ELIPs can be targeted to any molecule for pathologic ultrasound image enhancement.
We have demonstrated rapid IVUS image enhancement of injured endothelium/atheroma by ELIPs in an in vivo miniswine atheroma model. Image enhancement was confirmed quantitatively. Confirmation by IHC documented ELIP targeting. These data help to further the utility of targeted ultrasound contrast agents for molecular atheroma component enhancement.
☆ Supported in part by grant no. HL-59586 from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland, and the Feinberg Cardiovascular Research Institute, Chicago, Illinois.
This study was presented in part at the American College of Cardiology Scientific Sessions, Atlanta, Georgia, March 17 to 20, 2002.
- echogenic immunoliposome
- intercellular adhesion molecule-1
- intravascular ultrasound
- (monoclonal) antibody
- mean gray scale
- tissue factor
- vascular cell adhesion molecule-1
- American College of Cardiology Foundation
- ↵Lanza GM. Acoustically reflective liposomes and methods to make and use the same. Patent no. 5,612.057, 2000. University of Illinois and Northwestern University
- Demos S.M.,
- Alkan-Onyuksel H.,
- Kane B.J.,
- et al.
- MacDonald R.C.,
- MacDonald R.I.
- Verstraete M.
- Falk E.
- Marmur J.D.,
- Thiruvikraman S.V.,
- Fyfe B.S.,
- et al.
- Hamilton A.,
- Huang S.L.,
- Warnick D.,
- et al.
- Klegerman ME, Hamilton AJ, Tiukinhoy SD, et al. Fibrin association characteristics of atheroma-targeted echogenic immunoliposomes (abstr). Presented at the American Heart Association's Third Annual Conference in Salt Lake City, Utah, April 2002. Arterioscler Thromb Vasc Biol 2002;214
- Boenisch T.
- Lanza G.M.,
- Wallace K.D.,
- Scott M.J.,
- et al.
- Unger E.C.,
- McCreery T.P.,
- Sweitzer R.H.,
- Shen D.,
- Wu G.
- Nair A.,
- Kuban B.D.,
- Tuzcu E.M.,
- Schoenhagen P.,
- Nissen S.E.,
- Vince D.G.
- de Korte C.L.,
- Pasterkamp G.,
- van der Steen A.F.,
- Woutman H.A.,
- Bom N.