HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Appendix Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Stieger, S. M. Articles by Ferrara, K. W. Search for Related Content PubMed PubMed Citation Articles by Stieger, S. M. Articles by Ferrara, K. W.

Enhancement of Vascular Permeability with Low-Frequency Contrast-enhanced Ultrasound in the Chorioallantoic Membrane Model¹

¹ From the Department of Surgical and Radiological Sciences, School of Veterinary Medicine, University of California, Davis, 2112 Tupper Hall, Davis, CA 95616 (S.M.S., E.R.W.); Department of Biomedical Engineering, University of California, Davis, Calif (C.F.C., S.Q., K.W.F.); and Department of Physiology and Membrane Biology, School of Medicine, University of California, Davis, Calif (R.H.A., F.E.C.). Received January 27, 2006; revision requested March 28; revision received April 25; accepted May 31; final version accepted August 1. Supported by National Institutes of Health CA 103828. Address correspondence to S.M.S. (e-mail: smstieger{at}ucdavis.edu).

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
Purpose: To characterize the effect of low-frequency contrast material–enhanced ultrasound on the vascular endothelium and to determine the parameters and techniques required to deliver a therapeutic agent by using the chorioallantoic membrane (CAM) model.

Materials and Methods: All in vivo animal procedures were conducted with institutional Animal Care and Use Committee approval. Extravasation of 8.5-nm-diameter fluorescein isothiocyanate–labeled dextran was evaluated in the vasculature of a chick CAM model. Intravital microscopy was performed during contrast-enhanced ultrasound exposure (1.00 or 2.25 MHz); results were compared with results of electron microscopy of the insonated regions. Data acquired after insonation with greater mechanical stress (n = 30 animals) (mechanical index [MI] > 1.3) and with lower mechanical stress (n = 86 animals) (MI < 1.13) were compared with measurements in control conditions (n = 46 animals). The diameter of affected vessels; number of extravasation sites; extravasation rate, area, and location; and changes in endothelial cells and basement membrane were evaluated. Differences were tested with analysis of variance or the Student t test.

Results: After ultrasound application, convective transport of the model drug was observed through micron-sized openings with a mean fluid velocity of 188.6 µm/sec in the low-stress class and 362.5 µm/sec in the high-stress class. Electron microscopy revealed micron-sized focal endothelial gaps and disseminated blebs, vacuoles, and filopodia extending across tens of microns. The threshold pressure for extravasation was 0.5 MPa for a transmitted center frequency of 1.00 MHz (MI = 0.5) and 1.6 MPa for a frequency of 2.25 MHz (MI = 1.06); thus, the frequency dependence of the threshold was not predicted simply by the MI.

Conclusion: Low-frequency contrast-enhanced ultrasound can increase vascular permeability and result in convective extravasation of an 8.5-nm-diameter model drug.

Supplemental material: http://radiology.rsnajnls.org/cgi/content/full/243/1/112/DC1

© RSNA, 2007

Although ultrasonographic (US) contrast agents oscillate by a few percent of their mean diameter at a low transmission pressure, increased pressure or decreased frequency has been shown to increase the oscillation amplitude and the effect on the endothelium. In preclinical studies, endothelial changes have been observed to include transcytosis, fenestration, channel formation, opening of tight junctions, and free passage across the vascular wall (1–3) and to result in enhanced local delivery of colloidal particles and plasmids mediating angiogenesis (4). New interventional radiology approaches may exploit these effects to locally increase vascular permeability and deliver an intravenously injected drug to a target site (5–11). Contrast material–enhanced diagnostic US involving center frequencies near 1.00 MHz must also be performed with care, recognizing that the vascular endothelial effects that may be exploited for targeted therapeutic applications could potentially result in unintended consequences. The mechanical index (MI) is the measure of the relative risk of mechanical effects (streaming and cavitation) of ultrasound currently provided by commercial systems; however, to our knowledge, the propensity for vascular effects in the presence of contrast agents has not yet been systematically evaluated as a function of MI (12).

The mechanism for the effect on vascular endothelium is assumed to be cavitation related because of the great increase in the effect at low transmission frequencies; however, the exact mechanism is not well understood. Proposed mechanisms include vessel wall rupture produced by expansion of a microbubble beyond the diameter of the vessel (fibers rupture along a vessel axis cleft during lithotripsy) or regional endothelial disruption from a fluid jet or shock wave directed into the wall and resulting from the rapid collapse of a microbubble (13,14). Results of in vitro experiments indicate that an ultrasound pulse with a center frequency of 1.00 MHz and a peak negative pressure (PNP) of 1.3 MPa will expand a microbubble to a maximum instantaneous diameter of approximately 18 µm within a large vessel; this is followed by rapid collapse of the microbubble and fragmentation into smaller bubbles (15). On the basis of results of these in vitro studies and previous reports that a bubble-induced shock wave will affect a boundary (vessel wall) if the distance to the boundary at peak expansion is less than the peak diameter of the bubble (16), a gas bubble that expands to 18 µm might affect the entire wall circumference if it is centered in a vessel with a diameter as large as 54 µm (ie, three times the peak diameter).

The use of specific parameters, including a multisecond pulse interval, a selective arterial injection site, and increased microvascular pressure (17,18), has been shown to maximize local delivery of particles in vivo; however, to our knowledge, the kinetics of delivery have not been previously characterized. The chorioallantoic membrane (CAM) model is a suitable system for in vivo visualization of the effects of ultrasound on the permeability and ultrastructure of the vessel wall and the transport of a model drug from the vasculature to the interstitium. The CAM model has been used extensively over the past 30 years in angiogenesis research and in ion transport and tumor transplantation experiments (19). Thus, the purpose of our study was to characterize the effect of low-frequency contrast-enhanced ultrasound on the vascular endothelium and to determine the parameters and techniques required to deliver a therapeutic agent, by using the CAM model.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
Contrast Medium
ImaRx Therapeutics (Tucson, Az) provided the US contrast agent for this study. The experimental agent is a lipid-shelled perfluroropropane-filled US contrast agent that contains microbubbles of approximately 1.8 µm ± 1.5 (standard deviation) in diameter. The authors had control of the inclusion of the data and the information submitted for publication.

Ultrasound System
A cylindrically focused single-element transducer with a center frequency of 1.00 or 2.25 MHz was mounted confocally with the imaging plane of a microscope (Fig 1, A) (see Appendix E1, http://radiology.rsnajnls.org/cgi/content/full/243/1/112/DC1, for details).

View larger version (121K): [in this window] [in a new window] [Download PPT slide]   Figure 1: A, Diagram shows experimental setup for optical imaging of the CAM. B, Image of CAM acquired at day 15. C–J, Optical images from 2.25-MHz study with PNP of 1.7 MPa. C was acquired before insonation with 2.25 MHz, and D–J were acquired 0.06, 0.12, 0.24, 1.24, 2.24, 3.24, and 4.24 seconds after insonation began, respectively. K–P, Optical images from 1.00-MHz study at PNP of 1.3 MPa. K was acquired before insonation with 1.00 MHz, and L–P were acquired at the same times as C–J. A smaller number of extravasation sites (arrows) and a more gradual, spherical extravasation is observed with 2.25-MHz insonation as compared with 1.00-MHz insonation. Scale bar = 100 µm.

Ultrasound at a center frequency of 1.00 or 2.25 MHz was applied in 10-cycle pulses at a pulse repetition frequency of 500 Hz over a range of PNP values for 5 seconds (S.M.S., C.F.C.). The studies were performed and the results were analyzed by two observers, one of whom (S.M.S.) was a board-certified veterinary radiologist with 8 years of experience in veterinary diagnostic imaging and the other of whom (C.F.C.) was a biomedical engineer with experience in contrast-enhanced US imaging and modeling. PNP values ranged from 0.5 to 2.3 MPa in the 1-MHz study group and from 1.5 to 1.7 MPa in the 2.25-MHz group (Table 1). In addition to being described in terms of the control subgroups, the results will be described on the basis of two classes: one with application of greater mechanical stress (with subgroups 1 [insonation at 1.00 MHz and 1.3 MPa] and 2 [insonation at 1.00 MHz and 2.3 MPa]), and one with application of less mechanical stress (with subgroups 1 [insonation at 1.00 MHz and 0.5 MPa], 2 [insonation at 1.00 MHz and 0.7 MPa], and 3–5 [all 2.25-MHz insonation groups]). The application of mechanical stress by ultrasound is typically summarized by the MI (pressure normalized by the square root of the center frequency); in this study, the high-stress groups include MI values greater than 1.3, and the low-stress groups include values of 1.13 and lower. The insonated area of the CAM was recorded on a digital video disk for later evaluation, and the observed field of view was 2.0 mm².

Electron Microscopy
Fifteen samples from all 1.00-MHz studies and samples from each control group were evaluated with electron microscopy. Samples in the control subgroups that were not exposed to ultrasound were randomly chosen, whereas in the subgroups that were exposed to ultrasound, the insonated tissue was processed. The insonated samples were fixed for transmission electron microscopy within 30 seconds after ultrasound exposure. The perimeter and luminal area of each vessel were recorded (Appendix E1, http://radiology.rsnajnls.org/cgi/content/full/243/1/112/DC1). Events were measured (R.H.A.) as perimeter lengths exhibiting unusual features (eg, gaps through the endothelium, large vacuoles, multiple filopodia, separation between endothelial cells or between endothelium and basement membrane).

Modeling of FITC Dextran Extravasation through Microvessel Wall
Finite-element analysis of the transport of dextran from a small vessel (arteriole) to the interstitium was performed (Fig 2a; Appendix E1, http://radiology.rsnajnls.org/cgi/content/full/243/1/112/DC1) (20). The following parameters were used: a length of 80 µm, an intravascular pressure of 35 mm Hg (based on previously published data [21,22]), a density of 1060 kg/m³, a viscosity of 0.004 Pa · sec, a flow velocity of 3 mm/sec, a dextran concentration of 1.0 x 10^–5 mL/g, and a diffusion coefficient of 10^–11 m²/sec. Calculations (S.Q.) were performed for vessel diameters of 10 and 20 µm and endothelial gap diameters of 0.5 and 1.0 µm. The mean velocity of extravasation through the gaps, as well as the flow pattern of the extravasated fluid, was evaluated. The results of the finite-element analysis were compared with the results of the optical imaging studies and electron microscopy (K.W.F., S.Q., S.M.S.).

View larger version (5K): [in this window] [in a new window] [Download PPT slide]   Figure 2a: (a) Schematic of parabolic blood flow in vessel. d = Opening in vessel wall, Di = inner diameter of vessel (10 µm), Do = outer diameter of vessel (12 µm), Pi = intravascular pressure (35 mm Hg), umax = maximum flow velocity (3 mm/sec). (b) Graph shows mean velocity of dextran traveling through 0.5- and 1-µm openings of vessel. * = Mean velocity could not be calculated any further because extravasation traveled beyond the computational domain. (c) Calculated images of FITC dextran, which was transported through a vessel wall opening, along with representative optical images. Inner vessel diameter is 10 µm, wall thickness is 1 µm, and vascular pressure is 35 mm Hg. Square containing each subimage indicates computational interstitial domain (400 x 400 µm2). Solid rectangle = vessel segment (10 x 80 µm2). Open rectangle = scale bar (100 µm). Inner propagating front is 90% of the dextran concentration; outer propagating front is 10% of the dextran concentration. t = Time from start of dextran extravasation.

View larger version (15K): [in this window] [in a new window] [Download PPT slide]   Figure 2b: (a) Schematic of parabolic blood flow in vessel. d = Opening in vessel wall, Di = inner diameter of vessel (10 µm), Do = outer diameter of vessel (12 µm), Pi = intravascular pressure (35 mm Hg), umax = maximum flow velocity (3 mm/sec). (b) Graph shows mean velocity of dextran traveling through 0.5- and 1-µm openings of vessel. * = Mean velocity could not be calculated any further because extravasation traveled beyond the computational domain. (c) Calculated images of FITC dextran, which was transported through a vessel wall opening, along with representative optical images. Inner vessel diameter is 10 µm, wall thickness is 1 µm, and vascular pressure is 35 mm Hg. Square containing each subimage indicates computational interstitial domain (400 x 400 µm2). Solid rectangle = vessel segment (10 x 80 µm2). Open rectangle = scale bar (100 µm). Inner propagating front is 90% of the dextran concentration; outer propagating front is 10% of the dextran concentration. t = Time from start of dextran extravasation.

View larger version (33K): [in this window] [in a new window] [Download PPT slide]   Figure 2c: (a) Schematic of parabolic blood flow in vessel. d = Opening in vessel wall, Di = inner diameter of vessel (10 µm), Do = outer diameter of vessel (12 µm), Pi = intravascular pressure (35 mm Hg), umax = maximum flow velocity (3 mm/sec). (b) Graph shows mean velocity of dextran traveling through 0.5- and 1-µm openings of vessel. * = Mean velocity could not be calculated any further because extravasation traveled beyond the computational domain. (c) Calculated images of FITC dextran, which was transported through a vessel wall opening, along with representative optical images. Inner vessel diameter is 10 µm, wall thickness is 1 µm, and vascular pressure is 35 mm Hg. Square containing each subimage indicates computational interstitial domain (400 x 400 µm2). Solid rectangle = vessel segment (10 x 80 µm2). Open rectangle = scale bar (100 µm). Inner propagating front is 90% of the dextran concentration; outer propagating front is 10% of the dextran concentration. t = Time from start of dextran extravasation.

During the design of the study, an initial power analysis was performed to estimate sample size on the basis of a two-sample t test by using software (Power and sample size calculations program, version 2.1.31; W.D. Du Pont and W.D. Plummer, http://www.mc.vanderbilt.edu/prevmed/ps/index.htm). In our initial study design, we considered two groups: a control group and a group exposed to low-frequency contrast-enhanced ultrasound. With an value (the probability of a type I error) of .05, a power of .8, and an initial estimate of an observed effect in 20% of the individuals in the control group and in 50% of the individuals exposed to low-frequency contrast-enhanced ultrasound, we calculated a sample size of 40 for each group. In a pilot study (data not included here) we evaluated 15 CAMs: 10 were exposed to 1.00 MHz at 2.3 MPa, and five received the US contrast agents without exposure to ultrasound (control group). We observed extravasation in all animals in the first group but in no animals in the control group; therefore, in designing the present study, intermediate pressure values were included to enable us to better characterize the effect threshold, and the sample size was decreased.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
On the basis of the optical observations, the threshold pressure for extravasation was 0.5 MPa for a transmitted center frequency of 1.00 MHz (corresponding to an MI of 0.5) and 1.6 MPa for a transmission center frequency of 2.25 MHz (corresponding to an MI of 1.06); thus, the observed frequency dependence of the observed events cannot be summarized simply by the MI.

Optical Observations and Finite-Element Analysis
Optical studies allowed real-time observation of the effect of ultrasound on the CAM vasculature. Extravasation of FITC-labeled dextran was not observed in the control groups. Small, discrete regions of extravasated dextran were observed with low-frequency contrast-enhanced ultrasound. Flow continued within vessels from which extravasation occurred and within vessels in close proximity to the source of the extravasating dextran.

In 81 of 85 observations after insonation in the low-stress class, when extravasation occurred, the leading edge of the dextran traveled at a velocity of up to 250 µm/sec (mean velocity, 188.6 µm/sec), spreading nearly spherically from a single point (Fig 1). Alternatively, in the high-stress groups, extravasation increased in velocity to a peak near 700 µm/sec (mean velocity, 362.5 µm/sec), with the maximum velocity oriented perpendicular to the vessel axis (Fig 1). The flow of fluorescent dextran emanated from a single point, traveling away from the vessel while spreading more conically. On the basis of the two-dimensional image and flow rate, the extravasated volume from a single event was on the order of (0.2–12.0) x 10^–3 and (0.3–3.0) x 10^–4 µL within 1 second in the high- and low-stress classes, respectively.

With a vascular gap diameter of 0.5 µm (as predicted for a low-stress insonation), extravasated fluid is predicted to travel spherically, according to finite-element analysis. With a larger vascular opening (diameter of 1 µm), as expected for a high-stress insonation, the peak velocity oriented perpendicular to the vessel axis is larger, and extravasation is predicted to travel conically. In each case, the mean velocity of the fluid front, as estimated with finite-element analysis (Fig 2), is similar to the experimentally observed mean velocity (Table 2). The effect of diffusion in the fluid transport is very small, as illustrated by the increased distance between the 10% and 90% dextran concentration contours over time (approximately 5 µm/sec); thus, convection is the dominant transport mechanism accounting for these findings.

View larger version (14K): [in this window] [in a new window] [Download PPT slide]   Figure 3a: (a) Bar graph shows occurrence of extravasation in each pressure group as a function of PNP and corresponding number of vessels affected in each field of view. (b) Bar graph shows average size of vessels affected in each field of view and area over which extravasation was observed as a percentage of each field of view. (c) Bar graph shows rate of extravasation from each vessel, calculated from the time (in seconds) required for fluid to travel 100 µm2 from the affected vessel. (d) Bar graph shows rate of fluid extravasating from openings in walls of vessels with an initial diameter between 10 and 15 µm. Dotted lines with arrows indicate low-stress group. Error bars indicate standard deviations.

View larger version (16K): [in this window] [in a new window] [Download PPT slide]   Figure 3b: (a) Bar graph shows occurrence of extravasation in each pressure group as a function of PNP and corresponding number of vessels affected in each field of view. (b) Bar graph shows average size of vessels affected in each field of view and area over which extravasation was observed as a percentage of each field of view. (c) Bar graph shows rate of extravasation from each vessel, calculated from the time (in seconds) required for fluid to travel 100 µm2 from the affected vessel. (d) Bar graph shows rate of fluid extravasating from openings in walls of vessels with an initial diameter between 10 and 15 µm. Dotted lines with arrows indicate low-stress group. Error bars indicate standard deviations.

View larger version (11K): [in this window] [in a new window] [Download PPT slide]   Figure 3c: (a) Bar graph shows occurrence of extravasation in each pressure group as a function of PNP and corresponding number of vessels affected in each field of view. (b) Bar graph shows average size of vessels affected in each field of view and area over which extravasation was observed as a percentage of each field of view. (c) Bar graph shows rate of extravasation from each vessel, calculated from the time (in seconds) required for fluid to travel 100 µm2 from the affected vessel. (d) Bar graph shows rate of fluid extravasating from openings in walls of vessels with an initial diameter between 10 and 15 µm. Dotted lines with arrows indicate low-stress group. Error bars indicate standard deviations.

View larger version (10K): [in this window] [in a new window] [Download PPT slide]   Figure 3d: (a) Bar graph shows occurrence of extravasation in each pressure group as a function of PNP and corresponding number of vessels affected in each field of view. (b) Bar graph shows average size of vessels affected in each field of view and area over which extravasation was observed as a percentage of each field of view. (c) Bar graph shows rate of extravasation from each vessel, calculated from the time (in seconds) required for fluid to travel 100 µm2 from the affected vessel. (d) Bar graph shows rate of fluid extravasating from openings in walls of vessels with an initial diameter between 10 and 15 µm. Dotted lines with arrows indicate low-stress group. Error bars indicate standard deviations.

The two-dimensional area over which dextran extravasated during optical observation (the extravasated area) was estimated (Fig 3), in addition to the mean pixel intensity in the field. A significant difference in extravasated area was observed between the low- and high-stress classes (P < .001) and between the 1.00- and 2.25-MHz center frequencies (P < .001). The number of vessels showing extravasation was correlated with the area over which dextran was detected (Pearson correlation coefficient = 0.64; P < .001). With 1.00-MHz transmission, the extravasated area and mean pixel intensity increased significantly between the low-stress group insonated at 1.00 MHz and 0.7 MPa and the high-stress group insonated at 1.00 MHz and 1.3 MPa (P < .01), with the extravasated area increasing from 4.8% to 26% of the field of view (Fig 3). The mean extravasated area and pixel intensity did not increase further as transmission pressure increased above 1.3 MPa.

The time to fill an area of 100 µm² averaged 0.7 second ± 1.1. The rate of extravasation observed optically increased with increasing ultrasound pressure. In the 1.00-MHz groups, the fill time decreased from 2.6 seconds to 0.8 second with increasing pressure; however, the change in the observed fill time was not significant (F = 0.9, df = 4, P = .44) (Fig 3). Given that larger vessels were affected by low-frequency contrast-enhanced ultrasound as the pressure was increased, the extravasation rate was also evaluated for size-matched vessels. For vessels with a diameter between 10 and 15 µm, the extravasation rates between the high- and low-stress classes were significantly different (P < .001) (Fig 3).

Effect of Ultrasound on Microvascular Integrity Evaluated with Electron Microscopy
On the basis of results of electron microscopy, the size of the affected region increased and the qualitative nature of the effect changed between the low- and high-stress groups. In the low-stress groups, the extent of observed events was not different from that in the noninsonated control groups (P = .649) (Fig 4). These events were principally small membrane blebs, unusual vacuoles, or multiple filopodia and probably represented handling and fixation artifacts. In the higher stress class, changes in the wall perimeter included small membrane blebs, unusual vacuoles or multiple filopodia, small gaps in the endothelial layer (Fig 4), regions of disrupted or missing endothelium (Fig 4), and gaps penetrating both endothelium and basement membrane that permitted extravasation of blood to the interstitium (Fig 4). The diameter of these gaps ranged from hundreds of nanometers to several microns.

View larger version (218K): [in this window] [in a new window] [Download PPT slide]   Figure 4: Transmission electron micrographs show range of effects of ultrasound treatment in CAM model. Slices were stained as described in Appendix E1 (http://radiology.rsnajnls.org/cgi/content/full/243/1/112/DC1). A, Image in control group with injection of dextran only shows subchorionic blood capillaries immediately next to chorionic epithelium and overlying interstitial space. B, Low-magnification view of venular vessel in control group shows chorionic epithelium and associated blood capillaries at right and allantoic epithelium at left. C, High-magnification view of wall of venular blood vessel in control group shows highly attenuated vascular endothelium. D, High-magnification view in control group shows intact junction of endothelial cells and thin basement membrane between endothelium and pericyte. E, Detailed view of F, which was obtained in high-stress group insonated at 1.00 MHz and 1.3 MPa. Gap in endothelial layer is associated with extravasation of blood (arrows). F, Overview of region near venular vessel. Extravasated erythrocytes and plasma are visible. G, Detailed view of F shows erythrocyte impacted in wall of vessel where endothelial cells are missing (arrow). H–K, Images in high-stress group insonated at 1.00 MHz and 2.3 MPa. H, Chorionic capillary. Remnants of endothelial cells remain attached to basement membrane over much of perimeter (arrows). Erythrocyte is impacted in gap in endothelial layer. I, There is a small gap in endothelial layer, though basement membrane of capillary remains intact. J, Venular microvessel shows gap in endothelial layer and separation from underlying intact pericyte (arrows). K, There are disrupted endothelial cells with plasma in space separating endothelium and pericyte, yet basement membrane on abluminal side of pericyte remains intact. One erythrocyte is impacted on basement membrane where endothelium is missing. Extravasated erythrocytes from region out of the plane of section are in interstitium. Density possibly indicating fibrin is seen adjacent to site where blood plasma is in contact with basement membrane. a = Allantoic epithelium; bm = basement membrane; C = capillary; Ch = chorionic epithelium; ec = endothelial cell; Er, er, er* = erythrocyte; ex = extravasated erythrocytes; f = fibrin; g = gap; L = lumen; N = nucleus; P = pericyte; tj = tight junction.

View larger version (25K): [in this window] [in a new window] [Download PPT slide]   Figure 5a: (a) Bar graph shows percentage of vessel wall perimeter altered by insonation. Increasing the pressure produced an effect in a larger fraction of the vessel perimeter, particularly in 25–60-µm vessels. (b) Bar graph shows length of affected perimeter normalized by vessel area to account for the higher concentration of contrast agents in larger vessels. After correction for the relative frequency of microbubbles as a function of vessel diameter, spatial extent of the effect in 7–25-µm vessels was similar to that in 25–60-µm vessels. Error bars indicate standard deviations.

View larger version (27K): [in this window] [in a new window] [Download PPT slide]   Figure 5b: (a) Bar graph shows percentage of vessel wall perimeter altered by insonation. Increasing the pressure produced an effect in a larger fraction of the vessel perimeter, particularly in 25–60-µm vessels. (b) Bar graph shows length of affected perimeter normalized by vessel area to account for the higher concentration of contrast agents in larger vessels. After correction for the relative frequency of microbubbles as a function of vessel diameter, spatial extent of the effect in 7–25-µm vessels was similar to that in 25–60-µm vessels. Error bars indicate standard deviations.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

Gaps and Transport
On the electron microscopy images, small intracellular or intercellular gaps were observed in approximately 10% of the high-stress cases. On the basis of the optical observations, convective transport of dextran occurs through these small gaps. The observed transport through small gaps and the observation of extravasation from vessels as large as 55 µm suggest that the mechanism producing these events includes a liquid jet producing a very small opening in the endothelium rather than an elongated rupture resulting from an enlarged microbubble. Diffuse observations of blebs, unusual vacuoles, or multiple filopodia are hypothesized to result from a shock or pressure wave propagating through the fluid.

In our study, the estimated volume of extravasated dextran observed per event increased by approximately an order of magnitude with the higher stress parameters. In the high-stress class, we observed extravasation in the form of directional fluid jets traveling at hundreds of microns per second away from the original site, as predicted for convective transport from a small arteriole.

Results of finite-element analysis indicated that an extravasating directional fluid stream can be produced as the result of blood traveling through a small vascular gap into the interstitium. Although the finite-element analysis included here considered only the size of the vascular opening, the extravasation rate and volume will also depend on the diameter and pressure in the affected vessel and the interstitial pressure. Increasing the transmission pressure or decreasing the frequency can increase both the rate and the volume of transport to the interstitium.

When extravasation occurred in a small vessel, the event occurred immediately distal to a bifurcation from a large to a small vessel. On the basis of results of in vitro studies (23), it was anticipated that microbubble oscillation would decrease in a small vessel when the microbubble was far from a fluid reservoir (ie, large vessel). The viscous and resistive effects of the fluid within a capillary can restrict microbubble expansion, reducing the extravasation events recorded in our study.

Parameters and Predictions
In our study, MI did not predict the threshold for events as the transmission frequency was changed. Instead, the increase in center frequency from 1.00 to 2.25 MHz required a larger increase in pressure (0.5 to 1.6 MPa) to create an effect than would be predicted by MI. Given that the MI was not initially derived with the intention of predicting the likelihood of vascular events with contrast-enhanced ultrasound, evaluation of the threshold for such events as a function of frequency, pressure, and tissue type may be warranted.

It was anticipated that at the lower stress values in our study, a preferential effect would be observed on capillary-sized vessels. This expectation was based on results of optical studies of microbubble expansion (23), in which a 2-µm bubble insonated with a PNP near 0.8 MPa could expand to a diameter on the order of 10 µm. In the CAM studies reported here, the extravasation of fluid after insonation in the lower stress class was observed more frequently in capillary-sized vessels, with a small amount of dextran extravasation observed at each site. For a drug delivery strategy, producing a widespread effect in capillaries by using the lower stress parameters may result in a lower but more consistent concentration of the drug in the interstitium.

Although extravasation was observed in nearly all of the optical fields in the high-stress group insonated at 1.00 MHz and 1.3 MPa, the extravasation area, rate of extravasation, and the length of damage of the perimeter of the vessel wall did not increase substantially in the high-stress group insonated at 1.00 MHz and 2.3 MPa. Therefore, increasing transmission pressure may not increase drug delivery without bound. Results of in vitro optical studies of contrast agent expansion (15,23) indicated that insonation at 1.00 MHz and 1.3 MPa resulted in microbubble expansion to approximately 20 µm, which is sufficient to affect the entire boundary of a vessel with a diameter as large as 54 µm. Images of extravasation from the CAM similarly showed extravasation from vessels ranging up to a maximum diameter of 55 µm in the high-stress group.

CAM Model
A major advantage and central limitation of this work was the evaluation of these vascular effects in the optically accessible CAM model. The optical imaging technique in the CAM model provides great sensitivity to a small number of events that can be difficult to detect with electron microscopy and other techniques; however, there are always inherent difficulties extrapolating from experimental to human studies. The optical observations demonstrated that extravasation of dextran could occur in the low mechanical stress groups with a transmission pressure as low as 0.5 MPa at a center frequency of 1.00 MHz. At this low pressure, changes in the endothelium observed at electron microscopy were not significantly different from those observed in control groups, likely due to the difficulty of finding a small number of sites with electron microscopy.

While the threshold for vascular events may be different in various models, the observations and parameters found by studying the CAM model fit well with previous observations of other models. Although a small number of events occurred in the low-stress group insonated at 1.00 MHz and 0.5 MPa, the volume of material extravasated increased substantially as the transmission pressure was increased from 0.5 to 1.3 MPa. In mammalian cardiac models, the threshold for petechia and Evans Blue delivery with a center frequency of 1.7 MHz was 0.54 MPa (18), and marked microbubble-augmented gene delivery was observed with a center frequency of 1.3 MHz and a transmission pressure of 1.8 MPa (4). Price et al (2) have demonstrated extravasation with similar parameters in a rat model.

Practical applications: Low-frequency contrast-enhanced ultrasound produces extravasation of a model drug through small endothelial gaps, with convective transport observed to result in fluid velocities on the order of tens to hundreds of microns per second. As the transmission pressure increased or the center frequency decreased, the rate and volume of extravasated material increased, and the vascular events observed at electron microscopy increased in extent and severity. The potential for such effects during contrast-enhanced US imaging at 1.00 MHz should be acknowledged. The major application of this work is in the optimization of enhanced local drug delivery.

	ADVANCES IN KNOWLEDGE

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 

• Vascular permeability is increased in vessels with a diameter of up to 55 µm with intravascular injection of a US contrast agent and in response to 1.00-MHz ultrasound with a peak transmission pressure of 1.3 MPa; US contrast agents can expand up to 20 µm with these ultrasound parameters.
  
• Electron microscopy demonstrated micron-sized focal endothelial gaps and irregularly distributed blebs, vacuoles, and filopodia extending across 10s of microns.
  
• The mechanical index does not predict the frequency dependence of vascular effects as a function of frequency; increasing the transmission pressure and decreasing the center frequency are observed to increase the rate and volume of extravasated material.
  

	ACKNOWLEDGMENTS

 
We thank Tyler Kitano for technical assistance.

	FOOTNOTES

 

Abbreviations: CAM = chorioallantoic membrane • FITC = fluorescein isothiocyanate • MI = mechanical index • PNP = peak negative pressure

See Materials and Methods for pertinent disclosures.

Author contributions: Guarantors of integrity of entire study, S.M.S., F.E.C., E.R.W., K.W.F.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, S.M.S., C.F.C., R.H.A., F.E.C., K.W.F.; experimental studies, S.M.S., C.F.C., R.H.A., E.R.W., K.W.F.; statistical analysis, S.M.S., C.F.C., R.H.A., S.Q., F.E.C., K.W.F.; and manuscript editing, S.M.S., C.F.C., R.H.A., S.Q., F.E.C., E.R.W., K.W.F.

	References

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 

1. Sheikov N, McDannold N, Vykhodtseva N, Jolesz F, Hynynen K. Cellular mechanisms of the blood-brain barrier opening induced by ultrasound in presence of microbubbles. Ultrasound Med Biol 2004;30:979–989.[CrossRef][Medline]
2. Price RJ, Skyba DM, Kaul S, Skalak TC. Delivery of colloidal particles and red blood cells to tissue through microvessel ruptures created by targeted microbubble destruction with ultrasound. Circulation 1998;98:1264–1267.[Abstract/Free Full Text]
3. Skyba DM, Price RJ, Linka AZ, Skalak TC, Kaul S. Direct in vivo visualization of intravascular destruction of microbubbles by ultrasound and its local effects on tissue. Circulation 1998;98:290–293.[Abstract/Free Full Text]
4. Korpanty G, Chen S, Shohet RV, et al. Targeting of VEGF-mediated angiogenesis to rat myocardium using ultrasonic destruction of microbubbles. Gene Ther 2005;12:1305–1312.[CrossRef][Medline]
5. Bekeredjian R, Chen S, Frenkel PA, Grayburn PA, Shohet RV. Ultrasound-targeted microbubble destruction can repeatedly direct highly specific plasmid expression to the heart. Circulation 2003;108:1022–1026.[Abstract/Free Full Text]
6. Lawrie A, Brisken AF, Francis SE, Cumberland DC, Crossman DC, Newman CM. Microbubble-enhanced ultrasound for vascular gene delivery. Gene Ther 2000;7:2023–2027.[CrossRef][Medline]
7. Miller DL, Song J. Lithotripter shock waves with cavitation nucleation agents produce tumor growth reduction and gene transfer in vivo. Ultrasound Med Biol 2002;28:1343–1348.[CrossRef][Medline]
8. Ng KY, Liu Y. Therapeutic ultrasound: its application in drug delivery. Med Res Rev 2002;22:204–223.[CrossRef][Medline]
9. Unger EC, Hersh E, Vannan M, Matsunaga TO, McCreery T. Local drug and gene delivery through microbubbles. Prog Cardiovasc Dis 2001;44:45–54.[CrossRef][Medline]
10. Vannan M, McCreery T, Li P, et al. Ultrasound-mediated transfection of canine myocardium by intravenous administration of cationic microbubble-linked plasmid DNA. J Am Soc Echocardiogr 2002;15:214–218.[CrossRef][Medline]
11. Yuh EL, Shulman SG, Mehta SA, et al. Delivery of systemic chemotherapeutic agent to tumors by using focused ultrasound: study in a murine model. Radiology 2005;234:431–437.[Abstract/Free Full Text]
12. Apfel RE, Holland CK. Gauging the likelihood of cavitation from short-pulse, low-duty cycle diagnostic ultrasound. Ultrasound Med Biol 1991;17:179–185.[CrossRef][Medline]
13. Postema M, van Wamel A, Lancee CT, de Jong N. Ultrasound-induced encapsulated microbubble phenomena. Ultrasound Med Biol 2004;30:827–840.[CrossRef][Medline]
14. Zhong P, Zhou Y, Zhu S. Dynamics of bubble oscillation in constrained media and mechanisms of vessel rupture in SWL. Ultrasound Med Biol 2001;27:119–134.[CrossRef][Medline]
15. Chomas JE, Dayton P, May D, Ferrara KW. Threshold of fragmentation for ultrasonic contrast agents. J Biomed Opt 2001;6:141–150.[CrossRef][Medline]
16. Philipp A, Lauterborn W. Cavitation erosion by single laser-produced bubbles. J Fluid Mech 1998;361:75–116.[CrossRef]
17. Song J, Chappell JC, Qi M, VanGieson EJ, Kaul S, Price RJ. Influence of injection site, microvascular pressure and ultrasound variables on microbubble-mediated delivery of microspheres to muscle. J Am Coll Cardiol 2002;39:726–731.[Abstract/Free Full Text]
18. Li P, Cao LQ, Dou CY, Armstrong WF, Miller D. Impact of myocardial contrast echocardiography on vascular permeability: an in vivo dose response study of delivery mode, pressure amplitude and contrast dose. Ultrasound Med Biol 2003;29:1341–1349.[CrossRef][Medline]
19. Auerbach R, Kubai L, Knighton D, Folkman J. A simple procedure for the long-term cultivation of chicken embryos. Dev Biol 1974;41:391–394.[CrossRef][Medline]
20. Hu YT, Qin S, Hu T, Ferrara KW, Jian Q. Asymmetric oscillation of cavitation bubbles in a microvessel and its implications upon mechanisms of clinical vessel injury in shock-wave lithotripsy. Int J Nonlinear Mech 2005;40:341–350.[CrossRef]
21. Zweifach BW, Lipowsky HH. Quantitative studies of microcirculatory structure and function. III. Microvascular hemodynamics of cat mesentery and rabbit omentum. Circ Res 1977;41:380–390.
22. Dimitropoulou C, Malkusch W, Fait E, Maragoudakis ME, Konerding MA. The vascular architecture of the chick chorioallantoic membrane: sequential quantitative evaluation using corrosion casting. Angiogenesis 1998;2:255–263.[CrossRef][Medline]
23. Caskey CF, Kruse DE, Dayton PA, Kitano TK, Ferrara KW. Microbubble oscillation in tubes with diameters of 12, 25, and 195 microns. Appl Phys Lett 2006;88:033902.[CrossRef]

This Article Abstract Figures Only Full Text (PDF) Appendix Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Stieger, S. M. Articles by Ferrara, K. W. Search for Related Content PubMed PubMed Citation Articles by Stieger, S. M. Articles by Ferrara, K. W.