Effects of ultrasound and ultrasound contrast agent on vascular tissue.

Background

Ultrasound (US) is widely used clinically: applications include fetal development monitoring [1] and monitoringof cerebral hemorrhages [2]. The clinical use of US contrast agents produces few side effects and the safety profile drives the technology. US contrast agents have beendeveloped to enhance imaging via the generation of echosignals at the tissue-gas interface and following microbubble collapse. Intravascular ultrasound (IVUS) [3] hasrevolutionized the imaging of coronary circulation andvirtual histology is the product of the IVUS imaging. Inaddition, trans-esophageal echocardiography studies arefrequently used to assess ventricular function. Intravascular gene transfer using microbubbles has beenachieved, which has the potential to interdict into disease processes via gene therapy [4].

Optison is a first generation US imaging agent thatis comprised of a suspension of microspheres. Thesemicrospheres, 3–4.5 uM in diameter (32 uM maximum), consist of the insoluble gas Perflutrene, [5]surrounded by a shell of human serum albumin. US,when coupled with US contrast agents, can havemarked effects upon cells and vascular tissue. Notably,hemolysis of red blood cells (RBCs) can occur [6] andtumor ablation by anti-tumor tagged microbubbles andultrasound has been demonstrated. In addition, microvascular hemorrhaging can be induced by these combination treatments [7] and the blood brain barriercan be disrupted [8].

The clinical use of Optison has resulted in adversereports. For example, the generation of cardiac arrhythmias is an ongoing concern. Currently, US contrastagents are not recommended in the context of acutecoronary syndromes, acute myocardial infarction, or unstable cardiopulmonary disease. Following an US procedure, patients should be electrocardiographicallymonitored for 30 min [9].

In this study, we investigated the effects of both USand Optison on vascular function in anex vivoin an isolated aorta preparation. In addition, we examined howUS and US contrast agents altered vascular morphologyand apoptosis. Data from these studies suggest that theremay be a potential interaction of US and the US contrastagents on vascular function.

Methods

Animals

Sprague Dawley male rats weighing 200–250 g wereobtained from Harlan Laboratories, Inc., Indianapolis,IN. Upon receipt, the rats were held for 7 days for acclimation in an AAALAC-approved facility with ad libitumaccess to food and water. All experiments were performed with the approval of the CDRH IACUC. Ratswere housed singly, and the lights were on from 8 AMto 8 PM. For perfusion and removal of the aortas, therats were anesthetized with isoflurane (Halocarbon,River Edge, New Jersey), the thorax was opened, and theanimals were exsanguinated via heart puncture. Next,the heart and lungs were removed en bloc. A ligaturewas loosely placed at the proximal and distal end of thedorsal aorta and a small incision was made at the proximal end. The distal end was transected beyond the ligature. An 18 gauge intravenous catheter was gentlyinserted into the incision and care was taken to notintroduce the catheter deep into the aorta to avoid damaging the endothelium. The residual blood in the aortawas flushed with cell culture media, RPMI 1640 with10% FBS (GIBCO, Grand Island, NY), then the aortawas gently perfused with approximately 1 ml of cell culture media either with or without 1% Optison. Thelower ligature was tightened during the perfusion, andthen the proximal ligature was tightened while the catheter was removed, in an effort to retain media in the vessel. The aorta was then gently removed from the animaland placed in cell culture media for transport to the lab.

Following removal from the animal, the explantedaorta was mounted in an exposure chamber with a2.5 cm diameter opening bounded by two thin (12μm)plastic membranes separated by 3 mm. The exposurechamber is described in detail below. The time from tissue harvest to placement in the exposure chamber tosonication was about 20 minutes. The animal protocolwas approved by the FDA White Oak Institutional Animal Care and Use Committee.

US waveform and exposure set up

All ultrasound exposures were performed using a systemthat simulated a clinical ultrasound beam in pulsed Doppler mode. A spherically focused, 2 MHz transducerhaving a diameter of 2.5 cm and a focal length of 6 cm(Valpey, Fisher, Hopkinton, MA) was excited by a highvoltage pulser-receiver (Gammell Applied Technologies,Exmore, VA) using a four cycle burst and a pulse repetition frequency of 1 kHz. The exposure levels were measured with a spot-poled piezoelectric polymermembrane hydrophone that was constructed in-house[10]. It had an active diameter of 0.5 mm. The hydrophone’s sensitivity calibration was traceable to a nationalstandards laboratory. Ultrasonic pressure–time waveforms were recorded, and temporal peak pressures alongwith lateral focal beam dimensions were measured. Theoutput was calibrated in terms of the Mechanical Index(MI), a standardized quantity for predicting the potentialfor mechanical biological effects related to cavitation.The pulse duration and−6 dB focal beam width were2.3μs and 3 mm, respectively.

The exposure tank was 30 cm wide × 60 cm long × 30 cmdeep and thewater depthwas about 20 cm. Sound absorbingrubber was used to minimize reflections. The center of thechamber was positioned at the 6-cm focus of the ultrasoundbeam via a pulse-echo measurement at low MI (<0.2), andthen the exposures were made at MI = 1.9. The exposuresystem is shown in Figure 1A (transducer and chamber withaorta) along with the temporal pressure waveform inFigure 1B. Aortas were sonified for 30 seconds at 2.5 mmintervals down thelength of the vessel.

US treatment

Aortas were randomly divided into four groups: control,Optison, ultrasound and ultrasound plus Optison. Twoaortas were filled with RPMI 1640 and two were filledwith 1% Optison in RPMI 1640. After the US or shamexposure, both the control and Optison-treated aortaswere removed after five minutes and placed in a separatedish containing cell culture media. One rat dorsal aortawas used per treatment group

Myobath experiments

‍The ability of aortas to contract and relax was assessedusing a tissue myobath. Briefly, the ligatures to theaortas were cut and the adventitial fat was removedunder a dissecting microscope using fine scissors inice cold oxygenated Krebs buffer. The aortas were cutinto 3 mm sections mounted on hooks in 10 ml,water-jacketed myobath vessels (World PrecisionInstruments, Sarasota, FL). Aortic contraction and relaxation forces were detected using model Fort 25force transducers (WPI). The force signal was converted into a digital signal via Lab Trax 4/16 (WPI)and recorded via WPI Data Trax 2 2.05 software. Aortas were subjected to a stretch preconditioning byloading them with 0.5 gms for 15 minutes in oxygenated Krebs buffer which was followed by a washout.Krebs buffer was quickly added. Next, phenylephrine(Sigma) dissolved in 0.1 M bisulfite buffer (Sigma) was

added to the aortic rings and the bath concentrationsranged from 10-9to 10-4M. Relaxation was inducedby the addition of acetylcholine (Sigma) in concentrations ranging from 10-9to 3 × 10-5M. KCL (Sigma),2μM final, induced maximal contraction while 10-5Msodium nitroprusside (Sigma) induced maximal relaxation. These experiments were repeated three timesand representative results are shown.

Staining for endothelial markers

Unused 3 mm aortic segments were immediately fixedin 10% formalin/PBS, embedded and sectioned byAmerican Histolabs (Gaithersburg, MD). Sections weredewaxed in xylene (Sigma) and rehydrated in gradedethanol (Pharmaco) solutions for 2 min. Antigen retrieval was performed by microwaving on a high settingfor 300 seconds in 0.1 M citrate buffer (Sigma). Theslides were blocked with 5% horse sera for 20 min atroom temperature, which was provided in the ImpressKit (Vector labs, Burlingame, CA). Following a washingwith TBST, anti-VEGF, anti-FLT-1 and anti-FLK-1 (SantaCruz Biotechnology, Santa Cruz, CA) were diluted 1:50in 0.1% BSA/HBSS, and placed on the aortic sections. Acoverslip was added and the slides were incubatedovernight at 4°C. Next, the slides were washed threetimes with TBST and the slides were treated with 0.03%H2O2in ethanol for 15 min. The slides were washedthree times with TBST and strepavidin HRP (JacksonImmunoresearch, West Grove, PA), diluted 1:500 inTBST, was pipetted onto the sections. Following a onehour incubation, the slides were washed and diaminobenzidine (DAB) (Vector Labs) was prepared. Followinga10 minute incubation, the slides were quenched withdistilled water.

Apoptosis

TUNEL assay was performed according to the manufacturers’directions (Genscript Corporation, Piscataway,NJ). The slides were de-waxed with xylene and dehydrated with ethyl alcohol. The slides were treated withProteinase K and then the endogenous peroxidase wasinhibited by 3% peroxide (Sigma) in methanol (Sigma).The TUNEL reaction mixture was prepared and addedto the slides for 1 hour. Following washing, streptavidinHRP was added for 45 minutes at 37°C. After washingthe slides, DAB substrate (Vector Labs) was added withnickel enhancement for ten minutes. Finally, the slideswere immersed in water and counter-stained withHemotoxylin. Images were captured as described above.

Statistics

All values are expressed as means ± SE. Differences between individual mean values were determined usingANOVA and Dunnett’s test using SAS for Windows(SAS, Cary, NC). Values of p less than 0.05 wereregarded as significant.

Results

Vascular contraction and relaxation studies were used toassess the effects of US and US contrast agent on arterial function. The contractile response to phenylephrinewas examined initially. As shown in Figure 2, contraction of a representative segment of the aorta was compromised by either US or the US contrast agent. Thecombination of US and US contrast agent dramaticallyaltered the contractile response. When the response wasnormalized to % of control (Figure 3), a marked reduction in the contractile response was noted in a statistically significant manner, implying that smooth musclefunction was impaired.

The ability of aortic segments to relax was examinednext. The rings were exposed to increasing concentrationsof acetylcholine. In the representative results shown inFigure 4, the control aortas readily relaxed, while the degree of relaxation was reduced in both US- and contrastagent-treated aortas. The combination of US and contrastagent had a marked effect upon relaxation. When the datawere normalized to % of control, the reduced vascular response to acetylcholine was apparent and was statisticallysignificant in the US + contrast agent group. Given thatrelaxation, as shown in Figure 5, is dependent upon therelease of soluble factors from endothelial cells, it can beinferred from these data that the endothelium wasdamaged following the combination treatment.

To assess endothelial integrity, the sections from control and treated aortas were stained with a combinationof the FLT-1, FLK1-1 and anti-VEGF. FLT-1 and FLK1-1are receptors for VEGF, and VEGF is synthesized andreleased by endothelial cells. Prominent DAB endothelialstaining was observed in control sections. Aortas treatedwith both the US and US contrast agents showed a reduction in staining. Again the combination of US andUS contrast agent suppressed the expression of endothelial markers, which is shown in Figure 6.

Apoptosis may be responsible for the alteration ofthe endothelial layer. Sections were subjected toTUNEL analysis. No TUNEL staining was seen in thecontrol and minimal endothelial staining was seen inthe US and US + contrast agent-treated aortas. In contrast, TUNEL staining of the endothelial cells was seenwhen US is combined with US contrast treated aortas,as shown in Figure 7.

Discussion

Ultrasound contrast agents are widely used for diagnostic ultrasound imaging procedures, but questions havebeen raised about their safety based on the results ofin vitroand animal studies. Notably, the combination ofUS and US contrast agents has been shown to producevascular damage. A number of these studies report damage to endothelial layer of blood vessels exposed to USand US contrast agents. In this paper, we have exploredthe physiological consequences of endothelial damagefollowing combined exposure to US and Optison. Whencombined, this treatment can have marked vasculareffects in anex vivovascular preparation.

As shown in Figures 2 and 3, a marked reduction incontractile response was noted when Optison was combined with US treatment. Phenylephrine is an-adrenergicagonist that induces smooth muscle contraction. Theinability of the aortic segment to contract implies thatthere is a profound injury to smooth muscle. This observation is in agreement with an earlier study whichshowed dose-dependent smooth muscle damage inex vivoporcine carotid arteries exposed to both US andOptison [11]. Similarly, cardiomyocyte viability has alsobeen compromised by the combination of US and Optison followingin vivoexposure [12].

Relaxation of arteries by acetylcholine is due to the release of soluble factors from endothelium. As shown inFigures 4 and 5, impaired relaxation was noted when USand Optison were combined. Brayman et al. noted thatendothelial monolayers were readily damaged by US andOptison [13]. In rabbit ears, adherence of platelets wasobserved following US and Optison treatment. Plateletaggregation and injury to endothelial cells were more severe when the contrast agent and US were combined

[14]. Importantly, US and Optison can induce markedmicrovascular damage in rat mesentery [7,15].Endothelial cells synthesize and release multiple formsof VEGF, which is required to maintain vascular integrityas well to promote angiogenesis [16]. The most commonforms of VEGF interact with tyrosine coupled receptors,FLT-1 and FLK-1, on the surface of the endothelium tomodulate function in an auto-regulatory fashion.

Staining for these three entities was markedly reducedin the US plus Optison group (Figure 6). A slightlydecreased DAB signal was seen in either US or Optison treatment.

In this study, endothelial apoptosis was noted duringthe combined treatment of US and Optison (Figure 7).Importantly, endothelial apoptosis has been found inrabbit corneal endothelial cells treated with both USand Optisonin vivo[17]. Using confocal microscopy,US and Optison was observed to induce apoptosis atlow energy. As emphasized in their study, damage waslimited to the endothelial layers, and the internal elastic lamina may protect the smooth muscles cells frominertial cavitation [11]. It should be noted that agentssuch as Optison have the propensity to bind to activated endothelial cells. If the binding of Optison takesplacein vivoas well, atherosclerotic lesions may be atrisk during US imaging.

Optison, in conjunction with US, has been used forgene therapy, drug delivery, angiogenesis studies, imaging vascular injury and evaluating cardiac function. Aswith any therapy, there are often untoward effects thatneed to be balanced the potential benefits. Given thatthere is the potential for widespread collateral effects,preclinical evaluation of ultrasound contrast agents iswarranted which is consistent with the results presentedin Figures 3, 5, 6 and 7.

The goal of US directed gene therapy is to identifyand transfect selected anatomical structures with thegene(s) of interest. Vascular beds are ideal targets giventhe relative ease in identification and the delivery ofthe DNA of interest to the relevant tissues. The transfer of genetic material such as plasmid into a cellrequires the brief disruption of the membrane. Additionally, signaling pathways such as ERK are activatedand may play a role in the eventual expression of thetransfected DNA, probably through mechanical sensing via integrins. US activates ERK ½ signaling via rockin skin fibroblasts [18].

Both endothelial cells and smooth muscle cells canbe transfected with plasmids by Optison and US [19].C-myc expression was decreased following transfectionwith anti-sense morpholino oligomers in porcine arteries treatedex vivo[20]. Neointimal proliferation wasinhibited following balloon injury when anti-sense p53plasmids or decoy E2F decoy oligo nucleotides, Optison and ultrasound were used to transfect rat carotidarteries [21,22]. The contractile response to prostaglandin was reduced in porcine carotid arteries thatwere transfected with eNOSex vivo[23]. Plasmid DNAand viruses can be transduced into skeletal musclethrough an intra vascular route using US and US contrast agents [24]. In sum, the use of both Optison andUS enhances the nonviral gene transfer is analternative to using viral vectors [25]. While thesestudies focused upon transfer of genetic material, therewas no attempt to evaluate untoward thrombus  formation.

Inflamed endothelial cells express cell adhesion molecules such as Intracellular adhesion molecule (ICAM)and vascular cell adhesion molecule (VCAM). BothICAM and VCAM can be used to image inflammationand the inflamed tissue can be targeted forin vivodrugdelivery as well. A gas-filled microbubble with antiICAM-1 antibody on its shell specifically binds toactivated endothelial cells over expressing ICAM-1 [26].Significantly, the endothelial BBB can be altered byultrasound and contrast agents and may be a means ofdelivering drugs to the CNS due to altered permeability[27,28]. Other vascular effects include micro vessel rupture, and cell death in the rat spinotrapezius muscleusing Optison [29]. Further, petechiae, and capillaryleakage was observed in the mouse abdomen followinguse of Optison [30]. Clearly, the coupling of integrinspecific MAB to US contrast reagents may generate tissue specific contrast reagents but, there are nonspecificand collateral damage that results from the combineduse of Optison and Ultrasound.

Tumors are dependent upon the formation of neovessels for continued growth. Noninvasive,in vivo,imaging of vasculature is extremely important for identifying tumors. Most importantly, this permits a meansof screening anti-tumor regimens in preclinical modelsand clinical applications. These imaging techniques alsopermit high-resolution, volumetric assessments oftumor vascularity. In a preclinical model of breast cancer is shown that correlates with other ultrasonographicmeasures of blood flow, which may provide greater sensitivity to the microvasculature in real time [31]. Theendothelium of tumor neo-vessels express vascular cellendothelial growth factor receptor 2 (VEGFR2). UCAMicroMarker has been with conjugated to anti VEGFR2have been used to follow angiogenesis in a preclinicalmurine model for breast tumors [32]. In a human xenograft melanoma model, immuohistochemical COX-2staining of excised tumors correlated with the contrastenhanced ultrasound image [33]. The imaging of angiogenesis is dependent upon the expression of tumor orendothelial markers such as VEGFR2. The expression ofVEGFR2 may be variable, depending upon the growthstage/size and in humans, clonality. These imagingtechniques also permit high-resolution, volumetricassessments of tumor vascularity. The utility of following VEGF receptor and signaling kinases as a marker ofendothelial integrity is amply demonstrated in thispaper.

Inappropriate thrombus formation in the heart, brainor in a peripheral site is the hallmark of vascular disease.

Imaging thrombus is an important application of Optison and US. Abciximab, which recognizes glycoproteinIIb/IIIa receptor was conjugated to Optison and theimmuno bubbles enhance the image of arterial thrombusin vivo[34]. In contrast, US and USCA can be used incombination to break apart moderate sized clots [35]. Inthe rabbit ear, US and contrast agent were directedagainst the auricular vein. When fibrinogen was administered, the vein was occluded by an a thrombus [36].These examples illustrate the usefulness in identifyingthrombus as well as inducing thrombus to inhibit bloodflow to a target lesion. The loss of vascular function thatwe observed in this study illustrates the collateral damage that can be induced.

IVUS imaging of coronary and peripheral arteries isextremely useful technique to image plaque formationand vessel patency. Vulnerable plaque is an arterial lesion that has a propensity for rupture and thrombus formation. IVUS and contrast agent permits thevisualization of vasa vasorum density and a combinationof lipid core, cap thickness and calcification may helpidentify the plaques most likely to rupture [37].

Valvular stenosis can easily be visualized by ultrasoundexamination of the heart. The micro bubbles enhancethe image of the ventricle making it easier to identifythrombus, calculate the volume of ejected blood andvisualize wall motion. These functional studies are crucial for the clinical assessments of patients [38].

There are contradictory reports regarding the effectOptison on human cardiac function [39]. Perventricularcontractions (PVCs) were noted in the human heart[40]. Troponin T was elevation was seen with Optisonhowever, there were no negative histologic findingswere seen [41]. In an other study, no changes wereseen in PVCs, Troponin I, CK, CK-MB in the humanheart [42].

In a preclinical model using the rat heart, PVCs, andmicrovascular leakage was noted with Optison [43,44].Importantly, micro lesions were seen histologically withinflammatory infiltrates 24 hours post exposure in therat heart [44,45]. In glass catfish model, US and USCArevealed focal damage in the tail of fish. Importantly, thiswas a real time assessment of the damage, which wassignificant [46]. Our results from this study are consistent with vascular damage that may contribute to thearrhythmias seen in vivo.

Rat hearts were subjected to both US and Optison,and cardiac arrhythmias were induced. Cessation of UStreatment reversed the effect. However, no histologicaleffect was seen [47]. Rat hearts were treated with ultrasound and Optison, and the RNA was prepared formicroarray analysis. The only gene up-regulated wascarbonic anhydrase, so there was not dominant geneinduction [48].

Use of ultrasound contrast agents, newly-developedmicrobubble-based products which are administeredintravenously to enhance the ultrasound image quality,present new challenges with regard to clinical safetybecause of the locally destructive forces of inertialcavitation caused by the interaction of ultrasound withthe micro bubbles. These destructive forces can damage the endothelium and smooth muscle of the vascular wall. The target patient population that may beexposed to microbubble/ultrasound is large and continues to grow as new applications and new productsin this class are developed.

For example, increased permeability due to contrastagent-induced vascular damage can capitalized upontherapeutically to deliver genes and other large molecules across endothelium. However, an adverse eventthat may occur from microbubble-induced vascularlesions may be the initiation or acceleration of atherosclerotic progression. Since this modality is being considered for delivery of drugs and genes, an even largerpatient population (who originally had no cardiovasculardisease) will be exposed to long-term risks. Therefore, itis critical to identify the ultrasound and microbubble exposure conditions which cause damage to the vascularendothelium and determine whether microbubbleinduced vascular damage increases the risk of atherosclerosis in selected populations.

There are several limitations in the present study:First, it was performedex vivounder static conditionswith no blood flow. Second, only one concentration ofOptison was employed. In earlier studies (Miller, Dou,and Song 601–07; [6,12,39,49], the microbubble concentration ranged from 0.01% to 2% and maximal RBChemolysis was seen at 1%. In our experiment, we used1% Optison, so we are using concentrations that areconsistent with previous work. In clinical use, therecommended doses range from 0.5 to 5 ml of Optisoninfused in a 10 min period.

For these studies, we used a 2 MHz US wave employing a four-cycle tone burst simulating a pulsed Dopplermode and having an MI of 1.9, the maximum settingavailable on a clinical imaging unit. It is not clearwhether similar effects could occur at MI values lessthan 1.9.

In conclusion, 2 MHz US with an MI of 1.9 and 1%Optison altered contraction and relaxation in rat dorsalaortas exposedex vivo. The changes in arterial functionmay be due to damage of the endothelium and smoothmuscle. This study provides insight into functional parameters of vascular function that may be compromisedby US and Optison treatment.