Delivery of Colloidal Particles and Red Blood Cells to Tissue Through Microvessel Ruptures Created by Targeted Microbubble Destruction With Ultrasound
Background—We have previously shown that the application of ultrasound to thin-shelled microbubbles flowing through small microvessels (<7 μm in diameter) produces vessel wall ruptures in vivo. Because many intravascular drug- and gene-delivery vehicles are limited by the endothelial barrier, we hypothesized that this phenomenon could be used to deliver drug-bearing vehicles to tissue.
Methods and Results—An exteriorized rat spinotrapezius muscle preparation was used. Intravascular fluorescent red blood cells and polymer microspheres (PM) (205 and 503 nm in diameter) were delivered to the interstitium of rat skeletal muscle through microvessel ruptures created by insonifying microbubbles in vivo. On intravital microscopy, mean dispersion areas per rupture for red blood cells, 503-nm PM, and 205-nm PM were 14.5×103 μm2, 24.2×103 μm2, and 27.2×103 μm2, respectively. PM dispersion areas were significantly larger than the mean dispersion area for red blood cells (P<0.05).
Conclusions—Microvessel ruptures caused by insonification of microbubbles in vivo may provide a minimally invasive means for delivering colloidal particles and engineered red blood cells across the endothelial lining of a targeted tissue region.
There is intense interest in developing drug-delivery systems that deliver drugs and genes to diseased tissues and organs without producing detrimental side effects in healthy tissues and organs.1 2 3 4 A number of these promising intravascular drug-delivery systems, such as microspheres, liposomes, and surface-modified ceramics, are particulate in nature and are limited by their inability to easily diffuse across the endothelial lining.1 To date, there is no minimally invasive method for guiding intravascular particulate and cell drug carriers across the endothelial lining of a target tissue.
In this article, we present data that support a novel approach for overcoming the technical challenge of delivering particles (>205 nm in diameter) and engineered red blood cells (RBCs) across the endothelial lining of a target region. The basis for the technique is the interaction between gas-filled microbubbles and ultrasound. Both direct and indirect methods have shown that microbubbles are destroyed by ultrasound in vitro,5 6 and microbubble destruction has been indirectly observed in vivo as well.6 7 We8 have recently reported specific bioeffects associated with the insonification of gas-filled microbubbles in rat skeletal muscle microvessels in vivo. When microbubbles were exposed to ultrasound, their destruction created microvessel ruptures that were large enough to permit the extravasation of RBCs, yet cell and tissue damage were limited to the rupture site itself. The aim of the current study was to test the efficacy of delivering polymer microspheres (PMs) and RBCs across the endothelium through distinct microvessel ruptures.
The study was approved by the Animal Research Committee at the University of Virginia and conformed to the “Position of the American Heart Association on Research Animal Use.” Nine female Sprague-Dawley rats were anesthetized by an intramuscular injection (0.6 mL/kg body weight) of a 1% α-chloralose and 13.3% urethane solution. The left femoral vein and right carotid artery were cannulated. The right spinotrapezius muscle was exteriorized for intravital microscopy and arranged in a rectangular chamber that contained an ultrasound transducer aligned with the muscle thickness as previously described.8 For experiments intended to simulate the delivery of ex vivo engineered RBCs to the interstitium, 5 mL of whole blood was drawn from a donor rat and centrifuged. The RBCs were labeled with 10 μg/mL 1,1′-dioctadecyl-2,3,3′,3′-tetramethylindocarbo- cyanine perchlorate [DiIC18(3)].9
Before microbubbles were infused, several frames of ultrasound were applied to the muscle, which was then scanned to ensure that no microvessel ruptures were present. Microbubbles (Optison, Molecular Biosystems) were infused through the venous catheter at a rate of 0.24 mL/min. After 1 minute of infusion, a 0.25-mL bolus of red fluorescent PMs (205 nm in diameter, n=3; 503 nm in diameter, n=3) or a 0.4-mL bolus of heparinized DiIC18(3)-labeled RBCs (n=3) at 20% hematocrit was injected through the carotid artery catheter. PM concentrations were 2.1×1012 · mL−1 and 1.4×1011 · mL−1 for the 205- and 503-nm diameters, respectively. Three seconds after initiation of the bolus injection, which was the time of peak fluorescence in the capillaries, a single sweep of ultrasound (containing 128 lines and forming a 90° sector) was delivered over 12.8 ms (HDI 3000cv, Advanced Technologies Laboratory). Each line of ultrasound consisted of a 4-cycle pulse delivered over 0.1 ms. The mean transmission frequency was 2.3 MHz, and the mechanical index was set to 0.7.
After application of ultrasound to the muscle, it was scanned with a ×20 objective. Each field of view was first examined under transillumination to locate sites of microvessel rupture, which were identified by extravasation of RBCs. These were videotaped under transillumination and epifluorescent illumination with a red fluorescence filter used to detect labeled RBCs and PMs. In preliminary studies, ultrasound pulsing of muscles containing red fluorescent PMs but lacking microbubbles exhibited no microvessel ruptures.
Rupture site images from each muscle were digitized from videotape on a personal computer. Areas of PM and RBC dispersion were determined by planimetry. Because a low magnification objective with a long depth of focus was used, the entire depth of each dispersion volume was visible. Thus, measured dispersion areas corresponded to the planar area circumscribed by the entire dispersion volume. After ANOVA, statistical comparisons of mean dispersion areas per rupture site between the groups were made with Bonferroni t tests (P<0.05).
Ultrasound-induced microbubble destruction resulted in the rupture of small (<≈7-μm diameter) microvessels. Fluorescent PMs and RBCs were seen moving into the interstitium through the ruptures sites for ≈1 to 5 seconds after ultrasound application. Figure 1⇓, which consists of composite transilluminated and epifluorescent images, depicts regions from 3 spinotrapezius muscles in vivo after the delivery of 205-nm PMs (A), 503-nm PMs (B), and labeled RBCs (C) across the microvessel wall. Each delivery site was easily visualized because of RBC extravasation. PMs were typically dispersed beyond the volume defined by RBC extravasation, and in some cases, they extended as far as ≈200 μm beyond the RBCs. Fluorescent RBCs from donor rats were confined to the same zone as native RBCs. Application of ultrasound to the muscle when no microbubbles were present or when only PMs were present caused no vessel ruptures.
Figure 2⇓ is a histogram of the measured dispersion areas for RBCs and PMs. The 205- and 503-nm-diameter PM data represent 66 measurements each. RBC dispersion data comprise 113 measurements. Mean dispersion areas (±SD) for RBCs, 503-nm PMs, and 205-nm PMs were 14.5×103±870 μm2, 24.2×103±1580 μm2, and 27.2×103±1370 μm2, respectively. There was no significant difference between mean dispersion areas for 205- and 503-nm PMs, but mean dispersion area for labeled RBCs was significantly less (P<0.05) than the mean for PMs. Ranges for the dispersion areas of 205-nm PMs, 503-nm PMs, and RBCs were 10.79 to 67.8×103 μm2, 4.38 to 66.4×103 μm2, and 1.47 to 42.9×103 μm2, respectively.
The aim of this investigation was to test the hypothesis that large particles (205 to 503 nm in diameter) and RBCs can be delivered effectively to the interstitium of rat skeletal muscle through microvessel ruptures created by insonification of microbubbles. The results indicate that an individual rupture event causes the dispersion of PMs into a tissue area of ≈25×103 μm2. Assuming a circular dispersion region, the dispersion area has a radius of 89 μm. For RBCs, an area of ≈15×103 μm2 is typically covered, corresponding to a radius of 69 μm. The extent to which RBCs and PMs are driven into the interstitium is likely influenced by several factors, which include resistance of the tissue interstitium and microvessel rupture geometry, the pressure gradient across the rupture, and the local convective forces produced by the physics of microbubble destruction.
Resistance to extravasation is determined by the hydraulic resistance of the rupture site, the size of the rupture with respect to the size of the particle, the geometric configuration of the extracellular space, and the material properties of the delivery vehicle. Rupture size is likely dependent on several properties that influence microbubble destruction, including microbubble size and shell composition, delivered acoustic pressure, and intravascular pressure. Vessel wall structure also influences the degree of rupture. Clearly, the ruptures created in the vessel wall by the protocol described herein are large enough to allow the passage of nondeformable particles up to and quite possibly beyond 0.5 μm in diameter. These ruptures also allow passage of the highly deformable RBC. Extravasation of RBCs after ultrasound contrast application has been reported recently by others as well.10 11 Because RBC dispersion was significantly less than for PMs, it indicates that particle size is one important determinant of dispersion area. Additional work is needed to determine the maximum particle size that can move through these rupture sites and whether the size of the rupture can eventually be controlled by modulation of the various factors listed above. PMs with mean diameters of 205 and 503 nm were chosen for the present study because they represent the upper range of sizes for liposomes commonly used for drug and gene delivery. Extracellular matrix density and composition vary widely in different organs, implying that the relative effectiveness of this technique will be organ specific.
A second factor influencing dispersion area is the pressure gradient across the wall, which we postulate to be the primary force responsible for driving the PMs and RBCs through the rupture. This notion is supported by our observation that PMs and RBCs move through ruptures for ≈1 to 5 seconds after ultrasound application. This finding indicates that RBCs and PMs were flowing along a pressure gradient that steadily decreased until interstitial pressure had equilibrated with intraluminal pressure or until hemostasis occurred. Capillary pressure is heterogeneous in skeletal muscle, and this heterogeneity may contribute to the large variability in the dispersion areas for PMs and for RBCs.
Finally, although we hypothesize that it is the pressure gradient across the wall that primarily drives the motion of PMs and RBCs, local forces produced by the unique physical phenomena associated with microbubble destruction may also be important. Depending on the applied mechanical index and the condition and dimensions of the shell, microbubble destruction may consist of a single period of gas ejection from the encapsulating albumin shell or the partitioning of the gas bubble into numerous smaller, shell-free bubbles.5 It is conceivable that both modes of destruction occurred in the present study and that each mode may have significantly different effects on rupture size (as mentioned above) or on local convective forces that may help to propel delivery agents into the muscle. In addition, at least 4 other ultrasound contrast agents of differing shell materials produce skeletal muscle microvessel ruptures on exposure to ultrasound in our preparation.8
The ability of this technique to deliver large particles or cells across the endothelial lining of blood vessels indicates that it may be useful as a minimally invasive method for concentrating drugs in vascularized tumors that are identified and then targeted by ultrasound. In addition, this technique evokes the concept of using the RBC as a drug-delivery vehicle to tissue. The obvious advantage to using host RBCs as drug-delivery vehicles is that they are not targeted as foreign entities by the immune system. For instance, it has been shown that RBCs can be effectively used to remove pathogens from the bloodstream via bispecific constructs linked to complement receptor-1 on the RBC membrane.12 A similar strategy may be used to link neutralizing antibodies or cytokines to RBCs for the purpose of blocking or stimulating angiogenesis in a target zone.
Clearly, for this technique to be effective, the benefits of delivering therapy through capillary ruptures must outweigh the costs of damaging tissue. By our calculations, if 100 ultrasound frames are applied at time intervals chosen to restore microbubble concentration in the muscle between frames, only 1.5% of all capillaries will be ruptured and PM coverage will be 50% of total muscle area. This calculation indicates that broad PM delivery may be achieved with minimal capillary damage or muscle flow deficit.
In summary, colloidal particles and engineered RBCs can be delivered to the interstitium through endothelial ruptures created by the destruction of microbubbles by ultrasound. Although much work remains to be done before this technique may be used clinically, it may eventually represent a minimally invasive means for treating pathologies via a targeted drug-delivery strategy.
Dr Price is supported in part by a Scientist Development Grant from the American Heart Association (No. 9730025N), Dallas, Tex, and Dr Skyba is the recipient of a postdoctoral fellowship grant from the National Institutes of Health (No. F32-HL09540) Bethesda, Md. This study was also supported in part by grants (R01-HL48890 to Dr Kaul and R01-HL52309 to Dr Skalak) from the National Institutes of Health, Bethesda, Md, and Molecular Biosystems, Inc, San Diego, Calif, and by an equipment grant from Advanced Technologies Laboratories, Bothell, Wash.
- Received May 28, 1998.
- Revision received August 19, 1998.
- Accepted August 19, 1998.
- Copyright © 1998 by American Heart Association
Hnatyszyn HJ, Kossovsky N, Gelman A, Sponsler E. Drug delivery systems for the future. PDA J Pharm Sci Technol. 1994;48:247–254.
Dayton P, Morgan K, Allietta M, Klibanov A, Brandenburger G, Ferrara K. Simultaneous optical and acoustical observations of contrast agents. IEEE Ultrasonics Symp. 1997;1583–1591.
Wei K, Jayaweera AR, Firoozan S, Linka AZ, Skyba DM, Kaul S. Ultrasound-induced destruction of intravenously administered microbubbles: a novel method for the quantification of myocardial blood flow with echocardiography. Circulation. 1998;97:473–483.
Skyba DM, Price RJ, Linka AZ, Skalak TC, Kaul S. In vivo destruction of microbubbles by ultrasound: effects on local tissues and capillaries. Circulation. 1998;98:290–293.
Taylor RP, Martin EN, Reinagel ML, Nardin A, Craig M, Choice Q, Schlimgen R, Greenbaum S, Incardona NL, Ochs HD. Bispecific monoclonal antibody complexes facilitate erythrocyte binding and liver clearance of a prototype pathogen in a monkey model. J Immunol. 1997;159:4035–4044.