- Email: [email protected]
Ultrasound contrast agents nucleate inertial cavitation in vitro
Ultrasound
in Med.
& Biol.,
Pergamon
Vol. 21. No. 8, pp. 10591065, 1995 Copyright 0 1995 Elsevier Science Inc. Printed in the USA. All rights reserved 0301.5629/95 $9.50 + .OO
0301-5629(95)0046-l
@Original
Contribution ULTRASOUND
CONTRAST AGENTS NUCLEATE CAVITATION IN VITRO
INERTIAL
L. MILLER and RONALD M. THOMAS Battelle Pacific Northwest Laboratories,Richland, WA, USA
DOUGLAS
(Received
14 Rebruary
1995;
in final form
11 April
1995)
Abstract-Some ultrasound contrast agents contain stable bodies of gas, and this study was undertaken to determine if these agents could provide nuclei for inertial cavitation. Inertial cavitation was detected and assessed by the measurement of the sonochemical hydrogen peroxide after exposure to 2.17-, 2.95 or 3% MHz ultrasound. A noncavitating system was obtained by removing cavitation nuclei from the rotating tube exposure chambers by vacuum degassing, and from the phosphate-buffered saline medium by filtering. Albunex@ added at lo-‘, 10e3 or lo-“ dilutions, or Levovist@ added at 2 mg mL-‘, 0.2 mg mL-’ or 0.02 mg mL-’ all initiated significant H,O, production for 2.17-MHz ultrasound at 0.41 MPa or higher spatial peak pressure amplitude for 5 min exposure gated at 0.25 s on and off with 60-rpm rotation. Not rotating the tube virtually eliminated H,OZ production. For 2.5-min continuous exposure, both agents initiated significant HZOZ production for 2.95-MHz exposure at 0.58 MPa or higher, but not for 3%MHz exposure up to 1.16 MPa. Bubble-based ultrasound contrast agents therefore appear to be able to provide nuclei for inertial cavitation in the rotating tube exposure s,ystem.
Key Words: Diagnostic ation, Hydrogen
ultrasound, peroxide.
Contrast
agents, Inertial
INTRODUCTION
Sonochemicals,
Cavitation
nucle-
gaseousinclusion in a liquid medium (AIUM 1993). For example, a form of cavitation called gas-body activation occurs when preexisting stable bodies of gas, such as those that exist in leaves of the aquatic plant Elodea, fruit fly larvae or pores in hydrophobic membranes, are insonated (Miller 1987). By this definition, bubble-based contrast agents for diagnostic ultrasound represent cavitation in viva whenever they are used clinically. Indeed, these agents function by the strong scattering typical of cavitation, which provides bright echoes for image enhancement (de Jong et al. 1992), or even differential contrast using second harmonic emissions(S&rope and Newhouse 1993). In the absence of stable gas bodies suitable for direct activation by ultrasound, cavitation may still occur, apparently through the modification of preexisting inhomogeneities in the medium, a process called cavitation nucleation. The nuclei normally are modeled as submicroscopic solid particles with gas entrapped in crevices (Atchley and Prosperetti 1989). However, free spherical bubbles also can nucleate cavitation (as is often assumedin theoretical calculations) (Flynn and Church 1988). Free spherical bubbles are unstable (due to gas diffusion) but may be stabilized by skins, which may allow them to serve as cavitation nuclei (Yount
Early in the development of imaging devices, the feasibility and desirability of contrast agents for diagnostic ultrasound was investigated, and gas-bubble based agents seemed to yield the best results (Ziskin et al. 1972). In the last few years, interest in contrast agents has accelerated and commercial products halve been created for clinical use (Golberg et al. 1994; Ophir and Parker 1989). Two types of bubble-based agents have received considerable attention: Albunex@ (Molecular Biosystems, San Diego, CA), a coated microbubble produced by sonication of an albumin solution (Winkelmann et al. 1994); and Levovist@ (Schering AG, Berlin), dry particles of galactose which form a microbubble suspensionwhen water is added (Schlief 1991). Since these ultrasound contrast agents essentially consist of gas bubbles, it is of interest to consider their relationship to the phenomenon of ultrasonic cavitation. Ultrasonic cavitation includes a variety of phenomena and processes.It may be considered generally as any interaction between an ultrasound field and any Battelle
cavitation,
Address correspondence to: Douglas Miller, Mail Stop W-53, Northwest, P.O. Box 999, Richland, WA 99352,, USA. IO59
1060
Ultrasound in Medicine and Biology
1979). The nuclei content of a medium can vary widely depending on its handling; for example, distilling or filtering tap water greatly increases its cavitation threshold. Normally, in the absenceof gas bodies suitable for direct activation, no cavitational bioeffects are possible without cavitation nucleation. In the megahertz frequency range, the nucleation process typically requires pressure amplitudes sufficient to produce inertial cavitation (also called transient cavitation) (Miller and Thomas 1993b). For inertial cavitation, bubbles collapse by fluid inertia on the compression phase to very small diameters (Flynn and Church 19SS), in contrast to noninertial cavitation, which involves only small radial oscillations. Above the inertial cavitation threshold, a suitable bubble collapseson each compressionphaseand produces exceptionally large gas pressures and temperatures. This phenomenon is evidenced experimentally by the generation of light, free radicals and sonochemicals during exposure of a liquid to ultrasound above the nucleation and inertial cavitation thresholds. The initiation of inertial cavitation is significant becauseit is a violent phenomenon in which cavities can proliferate and cause large-scale biological effects, such as the mechanical lysis of an entire cell suspension in vitro (Miller and Williams 1989). Inertial cavitation can also generate sufficient free radicals and sonochemicals that chemical mechanismsmust be considered to have a potential role in causing biological effects (Riesz and Kondo 1992). These effects include damage to cellular DNA, which could potentially be of greater biological consequence than purely mechanical damage from noninertial cavitation. Therefore, classification of cavitation phenomena as noninertial or inertial in nature is important to delineate the potential bioeffects mechanisms operating during ultrasound exposures. As noted above, cavitation occurs, by definition, whenever bubble-based contrast agents are exposed to ultrasound. The question arises as to whether these agents could serve to nucleate inertial cavitation. A few previous studies have examined cavitation resulting from bubble-basedcontrast agents. Holland and Apfel (1990) examined the occurrence of cavitation with Albunex@, and found that it could trigger acoustical scattering characteristic of cavitation. Thresholds as low as 0.52 MPa at 0.757 MHz were determined. Holland et al. (1992) used a 30-MHz active cavitation detector and found that a diagnostic ultrasound system produced results indicative of cavitation with [email protected] was indicated for peak negative pressures of 1.1 MPa or greater at 2.5 MHz. Williams et al. (1991) studied cavitation-induced lysis of erythrocytes at 0.75 MHz and found that addition of Echovist@ (another bubble-based contrast agent related to Levo-
Volume 21. Number 8. I995
vist@) enhanced the lysis for concentrations above about 5 mg mL-‘. However, lysis was suppressedby hematocrits higher than a few percent, with or without the added agent. Recently, Miller et al. (1995) observed the effect of added Albunex@ on hemolysis by l-MHz ultrasound in a rotating-tube exposure system. Hemolysis was enhanced by added Albunex@, but essentially ceasedto occur for hematocrits above 10%. Brayman et al. (1995) examined hemolysis for 1. l-MHz pulsed (1 ms repeated at 20 Hz) exposures in a 200-rpm rotating tube. Albunex@ added at 35 PL mL-’ allowed detection of hemolysis for 2.7-MPa peak negative pressure with a roughly constant amount (but decreasing percentage) of hemolysis as hematocrit was increased up to 40% (i.e., normal). These studies indicate that bubble-based contrast agents can enhance cavitation activity during ultrasound exposure. It seemslikely that the cavitation was at least partly inertial in nature; for example, based on the pressure amplitudes involved. However, this previous work lacks specific information on free radical and sonochemical production, which is a more direct indicator of inertial cavitation. The purpose of the present study was to test the potential for nucleation of inertial cavitation by Albunex@ and Levovist@ and the subsequent H202 production in the rotating tube system. This system is a convenient one for studying cavitation nucleation. Nuclei can be manipulated to some extent, and nucleation from nuclei in the tested medium can be evaluated separately from nuclei which may be present on the tube walls (Miller et al. 1991). In addition, only a few nuclei appear to be necessary for initiation of cavitation in the rotating-tube system (Miller and Williams 1992). The rotating-tube system is also convenient for studies of inertial cavitation, since free radical and sonochemical production can be observed at a range of frequencies (Miller and Thomas 1993a, 1993b). In this study, a procedure was developed to eliminate the nucleation of cavitation during exposure of a liquid-filled chamber. This allowed the detection of nuclei in test samples added to the medium. The contrast agents were added at various concentrations, and the results compared to the inertial cavitation produced with vigorously shaken water or fresh phosphate-buffered saline (PBS). In addition, tests were made with the tube stopped, and at three different frequencies (2.17 MHz, 2.95 MHz and 3.8 MHz). The results are discussed with regard to the nucleating ability of the ultrasound contrast agents. METHODS The rotating-tube exposure system has been described previously (Miller et al. 1991; Miller and
Ultrasound
contrast
agents 0 D. L. MILLER
Thomas 1993a, 1993b; Miller and Williams 19:89). The 5-mL chamber consisted of acoustically transparent plastic film formed into a 3.5-cm-long tube with plas,tic plugs in each end. One plug had a filling nec.k with a stopper, and the other had an imbedded magnet to help keep the tube centered (relative to a fixed magnet) during rotation. Normally, a dry chamber would be filled with fresh PBS, and cavitation would occur reliably upon exposure to ultrasound above about ‘0.4 MPa at 2.17 MHz for 60-rpm rotation. To prevent cavitation, potential cavitation nuclei must be removed from the chamber and the medium. To accomplish this, chambers were filled with PBS and placed in a vacuum for 30 min to remove or destroy nuclei on the tube walls. Tae chamber was then emptied, except for tlhe neck, by collapsing the tube, so that no air entered the chamber. Old PBS was denucleated by filtering through a 0.2-pm syringe filter (this filter had also been vacuum denucleated with the chambers) and carefully inserted icto the chamber neck to refill the tube with the filtered PBS, without introducing any air or visible bubbles irto the chamber. PBS at least 1 week old, ,which is known to contain fewer nuclei than fresh PBS (Miller et al. 1991), was used for filtering to minimize the original nuclei content. This procedure produced a denucleated chamber filled with medium of normal gas content (i.e., not degassed), which did not yield cavitation upon exposure up to the maximum pressure amplitudes available (e.g., 0.82 MPa at 2.17 MHz). For cavitation to occur, nuclei had to be added to the chamber. Thus, a medium could be tested for the presence of cavitation nuclei by pipetting small volumes of the test medium into the filling neck of the chamber just before refilling with the filtered PBS. The exposure system consisted of an oscillator (Hewlett Packard Model 3314A), an amplifier (Electronic Navigation Industries Model A-500), and airbacked transducers. The ultrasound fields generated were calibrated by scanning a bilaminar shielded membrane hydrophone (GEC-Marconi Ltd., Model Y343598) at the normal position of the tube. The beam characteristics are listed in Table 1. Exposures were carried out in a tank filled with degassed water at room
Table 1. Characteristics of the ultrasound be,ams. -Characteristic
Frequency (MHz) 2.17 2.95 3.8
Transducer diameter (cm) 1.9 1.9 1.3
Exposure distance (cm) 13 9 5
6-dB beam diameter (mm) 13.4 16.0 9.0
and R. M. THOMAS
1061
temperature, and the ultrasound beam was terminated at a block of sound absorbing rubber to simulate freefield conditions. Fresh PBS was made up each day of experimentation with premeasured dry salts (No. 1003, Sigma Chemical Co.) and purified water (Mini-Q UFplus, Millipore Corp.). Albunex@ was stored in a refrigerator at 4°C mixed by gently inverting the solution and used directly from the vial. Levovist@ was prepared for each exposure by weighing out 0.1 g into a 1.5mL microcentrifuge tube and adding 0.5 mL of purified water to yield a stock solution 200 mg mL’. This tube was vigorously shaken for 5 s, then allowed to rest for 1 min before sampling (the tube was discarded after sampling). Shaken water was prepared in the same way, but without the Levovist@. Polystyrene spheres (No. 111, 0.261-pm Duke Scientific Inc.) were sampled from the suspension supplied. After exposure, the hydrogen peroxide production was measured by the isoluminol method described previously (Miller and Thomas 1993a). For this study, the stock solution of isoluminol and microperoxidase was made up and allowed to stand overnight, when it was stable for at least 1 week. The assay was calibrated each day using a 0.5-PM concentration of H202. In addition, a ~-PM concentration was checked each day and these data, together with the 0.5-PM calibration values, were used to determine the exponent, which was 0.93, for the slightly nonlinear assay (Miller and Thomas 1993a). The measured data were converted to H202 concentration using each day’s calibration value, and are reported as the mean value of six measurements with standard deviation error bars. The first pressure amplitude in the increasing series (0.28 MPa, 0.41 MPa, 0.58 MPa, 0.82 MPa, and 1.16 MPa to yield results significantly larger than the sham-exposure result in an unpaired Student’s t test was taken to be the threshold for inertial cavitation. RESULTS Three series of tests were conducted to evaluate the test procedures and the nucleating ability of the contrast agents relative to PBS. The first series included: filtered PBS; filtered PBS plus 50 PL of fresh PBS; filtered PBS plus 50 PL of water which had been shaken and rested; filtered PBS plus 0.5 PL, 5.0 PL or 50 PL of Albunex@; filtered PBS plus 0.5 pL, 5.0 PL or 50 /IL of dissolved Levovist@; fresh PBS; or fresh PBS plus 50 PL of Albunex@. The Levovist@ samples gave final dilutions to concentrations of 0.02 mg mI-‘, 0.2 mg mL’, and 2 mg mL-‘. Tubes with each of these media were exposed for 5 min to 2.17MHz ultrasound gated for 250 ms on and off, at spatial
1062
Ultrasound
in Medicine
and Biology
peak pressure amplitudes of 0.28 MPa, 0.41 MPa, 0.58 MPa and 0.82 MPa (approximately equivalent to 2.8, 5.6, 11.2 and 22.4 W cm-’ SPTA). The on and off gating, combined with 60-rpm tube rotation, yielded a 0.25-s burst each half-revolution, which allows for bubble cycling back and forth across the exposure tube. Results are shown in Fig. 1 for the 0.X2-MPa exposures, with the 5-PL samples of contrast agents. As expected, the filtered PBS did not yield H202 concentrations higher than sham-exposed samples even at the highest pressure amplitude. Samples with the added 50 /JL of fresh PBS did not cavitate either. All the other samples produced significantly increased H,O,. There were some differences in the levels of HIOl produced. For example, the Levovist@ sample produced significantly more H202 than the sample with shaken water. Also, the fresh PBS plus 50 I.LL Albunex@ produced
significantly
more Hz02 than the sam-
ple with fresh PBS only. The results for different concentrations of the contrast agents are shown in Fig. 2, plotted against the pressure amplitude. The samples with shaken water, fresh PBS and fresh PBS plus Albunex@ gave similar results as a function of pressure amplitude (data not shown). The threshold for H202 production was 0.41 MPa for all samplesand dilutions. Although the higher dilutions of Levovist@ seemedto produce more H202, there were no significant differences between any of the dilutions for either agent. However, the Levovist”
7
BLK
+ PBS
+SW
t ABX
Nucleation
+LVT
PBS
PBS +ABx
Conditions
Fig. 1. Comparison of the hydrogen peroxide concentration producedby 2.17-MHz ultrasound exposure for 5 min, gated 250 ms on and 250 ms off, with 60-rpm rotation at 0.82MPa spatial peak pressure amplitude. The conditions were: BLK, filtered PBS; +PBS, filtered PBS plus 50 PL of fresh PBS; +SW, filtered PBS plus 50 PL of water which had been shaken and rested; +ABX. filtered PBS plus 5 PL Albunex@; +LVT, filtered PBS plus 5 ,LLLof dissolved Levovist@, PBS, fresh PBS; or PBS + ABX, fresh PBS plus 50 PL of Albunex@.
Volume
21, Number
8, 1995
ABX
0.0
0.4
Pressure Amplitude
0.6
0.8
1.0
(h4F’a)
Fig. 2. Hydrogen peroxide production with conditions as in Fig. 1. as a function of spatial peak pressure amplitude. Added volumes of different stock Albunex@ (open symbols) or Levovist@ (solid symbols) were: squares 50 p,L, diamonds 5 pL, and circles 0.5 /IL.
result was significantly greater than the Albunex@ result for all the dilutions. Cavitation activity appearsto be greatly enhanced when a tube is rotated relative to the stationary tube (Miller and Williams 1989). Since the contrast agents evaluated in the first seriesof tests seemedto be a rich source of cavitation nuclei, a second seriesof tests was conducted to determine if these agents could initiate rotating-tube-like cavitation activity in a nonrotating tube. Filtered PBS, filtered PBS plus 5 yL of Albunex@ or Levovist@ and fresh PBS were used in sham exposure, 0.82-MPa gated exposure for 5 min or 0.82-MPa continuous exposure for 2.5 min, all in a nonrotating tube. Results are shown in Fig. 3. One sample with Albunex@ exposed to the gated beam gave a result similar to that for a rotating tube, but after averaging, mean was not significantly different the six-exposure from the sham-exposuremean. Some exposed samples with nuclei added were slightly higher than the corresponding sham exposure, but the only statistically significant elevation of H202 concentration was for the continuous exposure of the sample with Levovist@. However, even this result was much less than for the rotating tube (see Fig. 2). Use of higher frequencies of ultrasound appears to increase the threshold for inertial cavitation and to reduce its effects (Miller and Thomas 1993b). A third series of tests was conducted to determine if the contrast agents might reduce the thresholds at 2.17 MHz, 2.95 MHz and 3.8 MHz for 2.5-min continuous exposure with 60-rpm rotation. Filtered PBS, filtered PBS plus 5 PL of Albunex@ or Levovist@ and fresh PBS were used in sham exposure and exposure at several pressure amplitudes. Results are shown in Fig. 4 for
Ultrasound
contrast
agents 0 D. L. MILLER
1.5
E 2
I
Sham
W
0.82 MPa Gated
SSS
0.82
MPa CW
r
1.0 1
Filtered
+mx Nucleation
+ LVT
Fresh
Condition
Fig. 3. Hydrogen peroxide results for exposureswith the tube stationaryfor sham, 5-min gated or 2.5-min continuous exposurewith conditionsotherwiseas in Fig. 1.
all these conditions. No cavitation was obtained for 3.8-MHz exposure of any of the media up to 1.16 MPa. At 2.95 MHz, the threshold was 0.58 MPa for the three nucleated media. At 2.17 MHz, thresholds were at 0.41 MPa except for the fresh PBS, which produced a signjficant increase in H202 only at 0.58 MPa. Small plastic spheresappear to be another means of adding nuclei to media (Holland and Apfel 1990; Holland et al. 1992). Polystyrene spheres of 0.26 pm in diameter were added to 5 mL of filtered PBS in 5-PL samples, as for the other tests. At 2.17-MHz continuous exposure, this additive produced significant ca.vitation at 0.82 MPa, but not at 0.41 MPa or 0.58 MPa. Results were of a hit-or-miss nature, and were comparable to those for fresh PBS in two of six tests at 0.58 MPa, and in four of six tests at 0.82 MPa. These solid particles were also tested at 3.8 MHz, but did not instigate significant H202 production. DISCUSSION
and
R. M.
1063
THOMAS
known, and can be used to estimate the concentration of nuclei (Iernetti 1971). This experiment therefore serves as a means of enumerating the nuclei in the medium (e.g., about 1 per milliliter for fresh PBS); although, of course, the nuclei counted are only those detectable by this specific system. The bubble-based contrast agents Albunex@ diluted by lo-‘, lo-’ and 10eJ and Levoviste at 2 mg mI-‘, 0.2 mg mI-’ and 0.02 mg mL-’ concentrations served to nucleate inertial cavitation in this system. The concentrations of these agents normally employed in diagnostic applications are near the concentrations used in this study. For example, a 5-mL injection might mix with 0.5 L of blood in several heart beats to give a lop2 dilution for enhancing contrast in the heart, or mix with 5 L of blood throughout the body to yield a lo-’ dilution. Dilution to lo-” with 0.5 yL of sample, the highest dilution we could obtain with a one-step dilution, also gave reliable nucleation. This implies that at least one nucleus was present in each 0.5..PL sample. For comparison, the concentration of microspheresin Albunex@ is known to be about 500 million per milliliter (Bamhart et al. 1990), which is about lo5 per 0.5 pL. Inertial cavitation was also nucleated by adding 50-PL samples(1 O-’ dilution) of water which had been vigorously shaken and then allowed to rest for 1 min (the same procedure as used for mixing Levovist@ samples). The shaking process apparently produces microbubbles which persist long enough to act as nuclei when exposed to ultrasound. The contrast agent dilutions gave remarkably similar results given the wide range of dilutions. However, this is consistent with the rotating-tube phenomenon, for which only
4k
AND CONCLUSION
Hydrogen peroxide generation was measured to assessinertial cavitation activity during ultrasound exposures in a rotating tube system. Using nuclei-free chambers and media, no inertial cavitation was detected unless nuclei were added. Nucleation was obttined at 2.17 MHz with a full 5 mL of fresh PBS, but not with 50 ,zL of fresh PBS added to denucleated PBS. This implies that there was at least one nucleus available for initiating inertial cavitation in 5 mL but none in six 50-PL samples of fresh PBS. The dependence of cavitation thresholds on volume is well
0.0
0.4
0.6
Pressure Amplitude
0.8 (MPa)
Fig. 4. Resultsfor H,O, concentrationfrom exposure to 2.17 MHz, 2.95 MHz or 3.8 MHz of ultrasoundat the indicated pressureamplitudesfor 2.5 min continuously. The different media were: circles, filtered PBS; diamonds, fresh PBS; squares, filtered PBS + 5 pL Albunex@; triangles, filtered PBS + 5 bL of Levovist@.
1064
Ultrasound
in Medicine
and Biology
one suitable nucleus appears to be needed to generate maximal results (Miller and Williams 1992). One intriguing finding was that, although adding various dilutions of contrast agent produced the about same cavitation activity (see Fig. 2) adding Levovist@ to shaken water or adding Albunex@ to fresh PBS seemed to produce an additive effect (see Fig. 1). Possibly this could be explained by a greater level of aeration, which might enhance cavitation activity, in the shaken water and fresh PBS. Polystyrene spheres were also briefly tried as a source of nuclei. These appeared to serve a nuclei, but on a hit-or-miss basis for lop3 dilution. This behavior indicates that only zero or one nucleus is present in the sample, even though many spheres are present. The stock concentration of spheres is about 2.10” mL’, or about lo9 microspheres in the 5-PL samples. It is uncertain how these solid plastic spheres nucleate inertial cavitation. One possibility is that an occasional sphere, or group of spheres, has some gas associated with it (e.g., trapped in a defect, or within a cluster of spheres). In a previous study, it was found that experimental thresholds for inertial cavitation appear to increase faster with frequency (pressure amplitude threshold proportional to frequency) (Miller and Thomas 1993b) than is expected from theory (pressure amplitude threshold proportional to square root of frequency) (Apfel and Holland 199 1). The rapid increase of experimental thresholds in the megahertz frequency range has been noted by other investigators and appears to depend, to some extent, on the availability of nuclei in the medium (Iemetti 1971). In this study, the contrast agents and fresh PBS were tested at 2.17 MHz, 2.95 MHz and 3.8 MHz, to determine if the addition of large numbers of potential nuclei would alter the dependence of the cavitation threshold on frequency. The results do not seem to support this hypothesis, and the thresholds obtained were about the same as those found in the previous study (Miller and Thomas 1993b), except at 3.8 MHz. At 3.8 MHz, no cavitation was obtained in this study under any of the conditions tested, including fresh PBS. However, cavitation was obtained, as in the previous study, when fresh PBS was used in a dry chamber,(i.e., one that had not been denucleated), which suggests that nuclei on the tube walls are important at this frequency. In conclusion, the bubble-based contrast agents Albunex@ and Levovist@ appeared to be able to provide nuclei for inertial cavitation in vitro. That is, addition of these agents allowed production of statistically significant hydrogen peroxide concentrations by ultrasound exposure under conditions which did not yield H,O, production without the additives. Shaken water,
Volume
21, Number
8, 1995
fresh PBS and polystyrene spheres were also found to add inertial cavitation nuclei to the medium. All the different samples appear to provide qualitatively similar nuclei: the thresholds and dependence of activity on pressure amplitude were similar for all the nucleating samples. None was found capable of consistently instigating cavitation, which simulated the vigorous activity in a rotating tube, in a nonrotating tube. In addition, the agents did not appear to alter the frequency dependence of the experimental inertial cavitation threshold. Acknowledgetnents~We are indebted to Dr. C. Church, Molecular Biosystems, Inc., San Diego, CA, and to Dr. V. Uhlendorf, Schering AG. Berlin, for their generous assistance in obtaining the ultrasound contrast agents. This work was supported by PHS Grant No. CA42947, awarded by the National Institutes of Health. DHHS (USA).
REFERENCES AIUM. Bioeffects & safety of diagnostic ultrasound. Bethesda, MD: American Institute of Ultrasound in Medicine; 1993. Apfel. R. E.; Holland, C. K. Gauging the likelihood of cavitation from short-pulse, low duty cycle diagnostic ultrasound. Ultrasound Med. Biol. 17: 170- 185; 1991. Atchley, A. A.; Prosperetti, A. The crevice model of bubble nucleation. .I. Acoust. Sot. Am. 86:1065-1084; 1989. Barnhart, J.; Levene, H.; Villapando, E.; Maniquis, J.; Femandez, .I.; Rice, S.; Jablonski, E.; Gjoen, T.; Tolleshaug. H. Characteristics of Albunex, air-filled albumin microspheres for echocardiography contrast enhancement. Invest. Radio]. 25:sl62-~164; 1990. Brayman, A. A.; Azadniv. M.; Makin, I. R. S.; Miller. M. W.; Carstensen, E. L.; Child, M. S.; Raeman, C. H.; Meltzer. R. S.; Everbach, E. C. Effect of a stabilized microbubble echo contrast agent on hemolysis of human erythrocytes exposed to high intensity pulsed ultrasound. Echocardiography 12: 13-21; 1995. Flynn, H. G.; Church, C. C. Transient pulsations of small gas bubbles in water. J. Acoust. Sot. Am. 84:1863-1876; 1988. Golberg, B. B.; Liu, J.; Forsberg, F. Ultrasound contrast agents: a review. Ultrasound in Med. Viol. 20:319-333; 1994. Holland, C. K.: Apfel, R. E. Thresholds for transient cavitation produced by p&d ultrasound in a controlled nuclei environment. J. Acoust. Sot. Am. 88:2059-2069; 1990. Holland, C. K.; Roy, R. A.: Apfel, R. E. In vitro detection of cavitation induced by a diagnostic ultrasound system. IEEE UFFC 39:95-101: 1992. Iernetti, G. Cavitation threshold dependence on volume. Acustica 24:191-196; 1971. de Jong, N.; Hoff, L.; Skotland, T.; Born, N. Absorption and scatter of encapsulated gas filled microspheres: theoretical considerations and some measurements. Ultrasonics 30:95-103; 1992. Miller, D. L. A review of the ultrasonic bioeffects of microsonation, gas-body activation, and related cavitation-like phenomena. Ultrasound Med. Biol. 13:443-470; 1987. Miller, D. L.: Thomas, R. M.; Williams, A. R. Mechanisms for hemolysis by ultrasonic cavitation in the rotating exposure systems. Ultrasound Med. Biol. 17:171-178; 1991. Miller, D. L.; Thomas, R. M. A comparison of hemolytic and sonochemical activity of ultrasonic cavitation in a rotating tube. Ultrasound Med. Biol. 19:83-90; 1993a. Miller, D. L.; Thomas, R. M. Frequency dependence of cavitation activity in a rotating tube exposure system compared to the mechanical index. J. Acoust. Sot. Am. 93:3475-3480; 1993b. Miller, D. L.; Williams, A. R. Bubble cycling as an explanation of the promotion of ultrasonic cavitation in a rotating tube exposure system. Ultrasound Med. Biol. 15:641-648; 1989. Miller, D. L.; Williams, A. R. Nucleation and evolution of ultrasonic cavitation in a rotating exposure chamber. J. Ultrasound Med. ll:407-412; 1992.
Ultrasound
contrast
agents 0 D. L. MILLER
Miller, M. W.; Azadniv, M.; Doida. Y.; Brayman, A. A. Effect of a stabilized microbubble contrast agent on CW ultrasound induced red blood cell lysis in ~~itro. Echocardiography 12: l-12; 1995. Ophir, .I.; Parker, K. J. Contrast agents in diagnostic ultrasound. Ultrasound Med. Biol. 15:319-333; 1989. Rresz, P.; Kondo, T. Free radical formation induced by ultrasound and its biological implications. Free Rad. Biol. Med. 13:247270; 1992. Schlief. R. Ultrasound contrast agents. Cum Opin. Radiol. 3: 19% 207; 1991. Schrope, B. A.; Newhouse. V. L. Second harmonic ultrasonic blood perfusion measurement. Ultrasound Med. Biol. 19:567-579; 1993.
and R. M. THOMAS
1065
Williams. A. R.; Kubowicz, G.: Cramer, E.; Schlief. R. ‘Ihe effects of the microbubble suspension SH U 454 (Echovist) on ultrasonically-induced cell lysis in a rotating tube exposure system. Echocardiography 8:423-433; 1991. Winkelmann, J. W.: Kenner, M. D.; Rave, R.; Chandwaney, R. H.; Feinstein. S. B. Contrast echocardiography. Ultrasound Med. Biol. 20:507-515; 1994. Yount, D. E. Skins of varying permeability: a stabilization mechanism for gas cavitation nuclei. J. Acoust. Sot. Am. 65:14291439; 1979. Ziskin, M. C.; Bonabarpour, A.; Weinstein, D. P.; Lynch, P. Contrast agents for diagnostic ultrasound. Investig. Radiol. 7:500-505; 1972.
in Med.
& Biol.,
Pergamon
Vol. 21. No. 8, pp. 10591065, 1995 Copyright 0 1995 Elsevier Science Inc. Printed in the USA. All rights reserved 0301.5629/95 $9.50 + .OO
0301-5629(95)0046-l
@Original
Contribution ULTRASOUND
CONTRAST AGENTS NUCLEATE CAVITATION IN VITRO
INERTIAL
L. MILLER and RONALD M. THOMAS Battelle Pacific Northwest Laboratories,Richland, WA, USA
DOUGLAS
(Received
14 Rebruary
1995;
in final form
11 April
1995)
Abstract-Some ultrasound contrast agents contain stable bodies of gas, and this study was undertaken to determine if these agents could provide nuclei for inertial cavitation. Inertial cavitation was detected and assessed by the measurement of the sonochemical hydrogen peroxide after exposure to 2.17-, 2.95 or 3% MHz ultrasound. A noncavitating system was obtained by removing cavitation nuclei from the rotating tube exposure chambers by vacuum degassing, and from the phosphate-buffered saline medium by filtering. Albunex@ added at lo-‘, 10e3 or lo-“ dilutions, or Levovist@ added at 2 mg mL-‘, 0.2 mg mL-’ or 0.02 mg mL-’ all initiated significant H,O, production for 2.17-MHz ultrasound at 0.41 MPa or higher spatial peak pressure amplitude for 5 min exposure gated at 0.25 s on and off with 60-rpm rotation. Not rotating the tube virtually eliminated H,OZ production. For 2.5-min continuous exposure, both agents initiated significant HZOZ production for 2.95-MHz exposure at 0.58 MPa or higher, but not for 3%MHz exposure up to 1.16 MPa. Bubble-based ultrasound contrast agents therefore appear to be able to provide nuclei for inertial cavitation in the rotating tube exposure s,ystem.
Key Words: Diagnostic ation, Hydrogen
ultrasound, peroxide.
Contrast
agents, Inertial
INTRODUCTION
Sonochemicals,
Cavitation
nucle-
gaseousinclusion in a liquid medium (AIUM 1993). For example, a form of cavitation called gas-body activation occurs when preexisting stable bodies of gas, such as those that exist in leaves of the aquatic plant Elodea, fruit fly larvae or pores in hydrophobic membranes, are insonated (Miller 1987). By this definition, bubble-based contrast agents for diagnostic ultrasound represent cavitation in viva whenever they are used clinically. Indeed, these agents function by the strong scattering typical of cavitation, which provides bright echoes for image enhancement (de Jong et al. 1992), or even differential contrast using second harmonic emissions(S&rope and Newhouse 1993). In the absence of stable gas bodies suitable for direct activation by ultrasound, cavitation may still occur, apparently through the modification of preexisting inhomogeneities in the medium, a process called cavitation nucleation. The nuclei normally are modeled as submicroscopic solid particles with gas entrapped in crevices (Atchley and Prosperetti 1989). However, free spherical bubbles also can nucleate cavitation (as is often assumedin theoretical calculations) (Flynn and Church 1988). Free spherical bubbles are unstable (due to gas diffusion) but may be stabilized by skins, which may allow them to serve as cavitation nuclei (Yount
Early in the development of imaging devices, the feasibility and desirability of contrast agents for diagnostic ultrasound was investigated, and gas-bubble based agents seemed to yield the best results (Ziskin et al. 1972). In the last few years, interest in contrast agents has accelerated and commercial products halve been created for clinical use (Golberg et al. 1994; Ophir and Parker 1989). Two types of bubble-based agents have received considerable attention: Albunex@ (Molecular Biosystems, San Diego, CA), a coated microbubble produced by sonication of an albumin solution (Winkelmann et al. 1994); and Levovist@ (Schering AG, Berlin), dry particles of galactose which form a microbubble suspensionwhen water is added (Schlief 1991). Since these ultrasound contrast agents essentially consist of gas bubbles, it is of interest to consider their relationship to the phenomenon of ultrasonic cavitation. Ultrasonic cavitation includes a variety of phenomena and processes.It may be considered generally as any interaction between an ultrasound field and any Battelle
cavitation,
Address correspondence to: Douglas Miller, Mail Stop W-53, Northwest, P.O. Box 999, Richland, WA 99352,, USA. IO59
1060
Ultrasound in Medicine and Biology
1979). The nuclei content of a medium can vary widely depending on its handling; for example, distilling or filtering tap water greatly increases its cavitation threshold. Normally, in the absenceof gas bodies suitable for direct activation, no cavitational bioeffects are possible without cavitation nucleation. In the megahertz frequency range, the nucleation process typically requires pressure amplitudes sufficient to produce inertial cavitation (also called transient cavitation) (Miller and Thomas 1993b). For inertial cavitation, bubbles collapse by fluid inertia on the compression phase to very small diameters (Flynn and Church 19SS), in contrast to noninertial cavitation, which involves only small radial oscillations. Above the inertial cavitation threshold, a suitable bubble collapseson each compressionphaseand produces exceptionally large gas pressures and temperatures. This phenomenon is evidenced experimentally by the generation of light, free radicals and sonochemicals during exposure of a liquid to ultrasound above the nucleation and inertial cavitation thresholds. The initiation of inertial cavitation is significant becauseit is a violent phenomenon in which cavities can proliferate and cause large-scale biological effects, such as the mechanical lysis of an entire cell suspension in vitro (Miller and Williams 1989). Inertial cavitation can also generate sufficient free radicals and sonochemicals that chemical mechanismsmust be considered to have a potential role in causing biological effects (Riesz and Kondo 1992). These effects include damage to cellular DNA, which could potentially be of greater biological consequence than purely mechanical damage from noninertial cavitation. Therefore, classification of cavitation phenomena as noninertial or inertial in nature is important to delineate the potential bioeffects mechanisms operating during ultrasound exposures. As noted above, cavitation occurs, by definition, whenever bubble-based contrast agents are exposed to ultrasound. The question arises as to whether these agents could serve to nucleate inertial cavitation. A few previous studies have examined cavitation resulting from bubble-basedcontrast agents. Holland and Apfel (1990) examined the occurrence of cavitation with Albunex@, and found that it could trigger acoustical scattering characteristic of cavitation. Thresholds as low as 0.52 MPa at 0.757 MHz were determined. Holland et al. (1992) used a 30-MHz active cavitation detector and found that a diagnostic ultrasound system produced results indicative of cavitation with [email protected] was indicated for peak negative pressures of 1.1 MPa or greater at 2.5 MHz. Williams et al. (1991) studied cavitation-induced lysis of erythrocytes at 0.75 MHz and found that addition of Echovist@ (another bubble-based contrast agent related to Levo-
Volume 21. Number 8. I995
vist@) enhanced the lysis for concentrations above about 5 mg mL-‘. However, lysis was suppressedby hematocrits higher than a few percent, with or without the added agent. Recently, Miller et al. (1995) observed the effect of added Albunex@ on hemolysis by l-MHz ultrasound in a rotating-tube exposure system. Hemolysis was enhanced by added Albunex@, but essentially ceasedto occur for hematocrits above 10%. Brayman et al. (1995) examined hemolysis for 1. l-MHz pulsed (1 ms repeated at 20 Hz) exposures in a 200-rpm rotating tube. Albunex@ added at 35 PL mL-’ allowed detection of hemolysis for 2.7-MPa peak negative pressure with a roughly constant amount (but decreasing percentage) of hemolysis as hematocrit was increased up to 40% (i.e., normal). These studies indicate that bubble-based contrast agents can enhance cavitation activity during ultrasound exposure. It seemslikely that the cavitation was at least partly inertial in nature; for example, based on the pressure amplitudes involved. However, this previous work lacks specific information on free radical and sonochemical production, which is a more direct indicator of inertial cavitation. The purpose of the present study was to test the potential for nucleation of inertial cavitation by Albunex@ and Levovist@ and the subsequent H202 production in the rotating tube system. This system is a convenient one for studying cavitation nucleation. Nuclei can be manipulated to some extent, and nucleation from nuclei in the tested medium can be evaluated separately from nuclei which may be present on the tube walls (Miller et al. 1991). In addition, only a few nuclei appear to be necessary for initiation of cavitation in the rotating-tube system (Miller and Williams 1992). The rotating-tube system is also convenient for studies of inertial cavitation, since free radical and sonochemical production can be observed at a range of frequencies (Miller and Thomas 1993a, 1993b). In this study, a procedure was developed to eliminate the nucleation of cavitation during exposure of a liquid-filled chamber. This allowed the detection of nuclei in test samples added to the medium. The contrast agents were added at various concentrations, and the results compared to the inertial cavitation produced with vigorously shaken water or fresh phosphate-buffered saline (PBS). In addition, tests were made with the tube stopped, and at three different frequencies (2.17 MHz, 2.95 MHz and 3.8 MHz). The results are discussed with regard to the nucleating ability of the ultrasound contrast agents. METHODS The rotating-tube exposure system has been described previously (Miller et al. 1991; Miller and
Ultrasound
contrast
agents 0 D. L. MILLER
Thomas 1993a, 1993b; Miller and Williams 19:89). The 5-mL chamber consisted of acoustically transparent plastic film formed into a 3.5-cm-long tube with plas,tic plugs in each end. One plug had a filling nec.k with a stopper, and the other had an imbedded magnet to help keep the tube centered (relative to a fixed magnet) during rotation. Normally, a dry chamber would be filled with fresh PBS, and cavitation would occur reliably upon exposure to ultrasound above about ‘0.4 MPa at 2.17 MHz for 60-rpm rotation. To prevent cavitation, potential cavitation nuclei must be removed from the chamber and the medium. To accomplish this, chambers were filled with PBS and placed in a vacuum for 30 min to remove or destroy nuclei on the tube walls. Tae chamber was then emptied, except for tlhe neck, by collapsing the tube, so that no air entered the chamber. Old PBS was denucleated by filtering through a 0.2-pm syringe filter (this filter had also been vacuum denucleated with the chambers) and carefully inserted icto the chamber neck to refill the tube with the filtered PBS, without introducing any air or visible bubbles irto the chamber. PBS at least 1 week old, ,which is known to contain fewer nuclei than fresh PBS (Miller et al. 1991), was used for filtering to minimize the original nuclei content. This procedure produced a denucleated chamber filled with medium of normal gas content (i.e., not degassed), which did not yield cavitation upon exposure up to the maximum pressure amplitudes available (e.g., 0.82 MPa at 2.17 MHz). For cavitation to occur, nuclei had to be added to the chamber. Thus, a medium could be tested for the presence of cavitation nuclei by pipetting small volumes of the test medium into the filling neck of the chamber just before refilling with the filtered PBS. The exposure system consisted of an oscillator (Hewlett Packard Model 3314A), an amplifier (Electronic Navigation Industries Model A-500), and airbacked transducers. The ultrasound fields generated were calibrated by scanning a bilaminar shielded membrane hydrophone (GEC-Marconi Ltd., Model Y343598) at the normal position of the tube. The beam characteristics are listed in Table 1. Exposures were carried out in a tank filled with degassed water at room
Table 1. Characteristics of the ultrasound be,ams. -Characteristic
Frequency (MHz) 2.17 2.95 3.8
Transducer diameter (cm) 1.9 1.9 1.3
Exposure distance (cm) 13 9 5
6-dB beam diameter (mm) 13.4 16.0 9.0
and R. M. THOMAS
1061
temperature, and the ultrasound beam was terminated at a block of sound absorbing rubber to simulate freefield conditions. Fresh PBS was made up each day of experimentation with premeasured dry salts (No. 1003, Sigma Chemical Co.) and purified water (Mini-Q UFplus, Millipore Corp.). Albunex@ was stored in a refrigerator at 4°C mixed by gently inverting the solution and used directly from the vial. Levovist@ was prepared for each exposure by weighing out 0.1 g into a 1.5mL microcentrifuge tube and adding 0.5 mL of purified water to yield a stock solution 200 mg mL’. This tube was vigorously shaken for 5 s, then allowed to rest for 1 min before sampling (the tube was discarded after sampling). Shaken water was prepared in the same way, but without the Levovist@. Polystyrene spheres (No. 111, 0.261-pm Duke Scientific Inc.) were sampled from the suspension supplied. After exposure, the hydrogen peroxide production was measured by the isoluminol method described previously (Miller and Thomas 1993a). For this study, the stock solution of isoluminol and microperoxidase was made up and allowed to stand overnight, when it was stable for at least 1 week. The assay was calibrated each day using a 0.5-PM concentration of H202. In addition, a ~-PM concentration was checked each day and these data, together with the 0.5-PM calibration values, were used to determine the exponent, which was 0.93, for the slightly nonlinear assay (Miller and Thomas 1993a). The measured data were converted to H202 concentration using each day’s calibration value, and are reported as the mean value of six measurements with standard deviation error bars. The first pressure amplitude in the increasing series (0.28 MPa, 0.41 MPa, 0.58 MPa, 0.82 MPa, and 1.16 MPa to yield results significantly larger than the sham-exposure result in an unpaired Student’s t test was taken to be the threshold for inertial cavitation. RESULTS Three series of tests were conducted to evaluate the test procedures and the nucleating ability of the contrast agents relative to PBS. The first series included: filtered PBS; filtered PBS plus 50 PL of fresh PBS; filtered PBS plus 50 PL of water which had been shaken and rested; filtered PBS plus 0.5 PL, 5.0 PL or 50 PL of Albunex@; filtered PBS plus 0.5 pL, 5.0 PL or 50 /IL of dissolved Levovist@; fresh PBS; or fresh PBS plus 50 PL of Albunex@. The Levovist@ samples gave final dilutions to concentrations of 0.02 mg mI-‘, 0.2 mg mL’, and 2 mg mL-‘. Tubes with each of these media were exposed for 5 min to 2.17MHz ultrasound gated for 250 ms on and off, at spatial
1062
Ultrasound
in Medicine
and Biology
peak pressure amplitudes of 0.28 MPa, 0.41 MPa, 0.58 MPa and 0.82 MPa (approximately equivalent to 2.8, 5.6, 11.2 and 22.4 W cm-’ SPTA). The on and off gating, combined with 60-rpm tube rotation, yielded a 0.25-s burst each half-revolution, which allows for bubble cycling back and forth across the exposure tube. Results are shown in Fig. 1 for the 0.X2-MPa exposures, with the 5-PL samples of contrast agents. As expected, the filtered PBS did not yield H202 concentrations higher than sham-exposed samples even at the highest pressure amplitude. Samples with the added 50 /JL of fresh PBS did not cavitate either. All the other samples produced significantly increased H,O,. There were some differences in the levels of HIOl produced. For example, the Levovist@ sample produced significantly more H202 than the sample with shaken water. Also, the fresh PBS plus 50 I.LL Albunex@ produced
significantly
more Hz02 than the sam-
ple with fresh PBS only. The results for different concentrations of the contrast agents are shown in Fig. 2, plotted against the pressure amplitude. The samples with shaken water, fresh PBS and fresh PBS plus Albunex@ gave similar results as a function of pressure amplitude (data not shown). The threshold for H202 production was 0.41 MPa for all samplesand dilutions. Although the higher dilutions of Levovist@ seemedto produce more H202, there were no significant differences between any of the dilutions for either agent. However, the Levovist”
7
BLK
+ PBS
+SW
t ABX
Nucleation
+LVT
PBS
PBS +ABx
Conditions
Fig. 1. Comparison of the hydrogen peroxide concentration producedby 2.17-MHz ultrasound exposure for 5 min, gated 250 ms on and 250 ms off, with 60-rpm rotation at 0.82MPa spatial peak pressure amplitude. The conditions were: BLK, filtered PBS; +PBS, filtered PBS plus 50 PL of fresh PBS; +SW, filtered PBS plus 50 PL of water which had been shaken and rested; +ABX. filtered PBS plus 5 PL Albunex@; +LVT, filtered PBS plus 5 ,LLLof dissolved Levovist@, PBS, fresh PBS; or PBS + ABX, fresh PBS plus 50 PL of Albunex@.
Volume
21, Number
8, 1995
ABX
0.0
0.4
Pressure Amplitude
0.6
0.8
1.0
(h4F’a)
Fig. 2. Hydrogen peroxide production with conditions as in Fig. 1. as a function of spatial peak pressure amplitude. Added volumes of different stock Albunex@ (open symbols) or Levovist@ (solid symbols) were: squares 50 p,L, diamonds 5 pL, and circles 0.5 /IL.
result was significantly greater than the Albunex@ result for all the dilutions. Cavitation activity appearsto be greatly enhanced when a tube is rotated relative to the stationary tube (Miller and Williams 1989). Since the contrast agents evaluated in the first seriesof tests seemedto be a rich source of cavitation nuclei, a second seriesof tests was conducted to determine if these agents could initiate rotating-tube-like cavitation activity in a nonrotating tube. Filtered PBS, filtered PBS plus 5 yL of Albunex@ or Levovist@ and fresh PBS were used in sham exposure, 0.82-MPa gated exposure for 5 min or 0.82-MPa continuous exposure for 2.5 min, all in a nonrotating tube. Results are shown in Fig. 3. One sample with Albunex@ exposed to the gated beam gave a result similar to that for a rotating tube, but after averaging, mean was not significantly different the six-exposure from the sham-exposuremean. Some exposed samples with nuclei added were slightly higher than the corresponding sham exposure, but the only statistically significant elevation of H202 concentration was for the continuous exposure of the sample with Levovist@. However, even this result was much less than for the rotating tube (see Fig. 2). Use of higher frequencies of ultrasound appears to increase the threshold for inertial cavitation and to reduce its effects (Miller and Thomas 1993b). A third series of tests was conducted to determine if the contrast agents might reduce the thresholds at 2.17 MHz, 2.95 MHz and 3.8 MHz for 2.5-min continuous exposure with 60-rpm rotation. Filtered PBS, filtered PBS plus 5 PL of Albunex@ or Levovist@ and fresh PBS were used in sham exposure and exposure at several pressure amplitudes. Results are shown in Fig. 4 for
Ultrasound
contrast
agents 0 D. L. MILLER
1.5
E 2
I
Sham
W
0.82 MPa Gated
SSS
0.82
MPa CW
r
1.0 1
Filtered
+mx Nucleation
+ LVT
Fresh
Condition
Fig. 3. Hydrogen peroxide results for exposureswith the tube stationaryfor sham, 5-min gated or 2.5-min continuous exposurewith conditionsotherwiseas in Fig. 1.
all these conditions. No cavitation was obtained for 3.8-MHz exposure of any of the media up to 1.16 MPa. At 2.95 MHz, the threshold was 0.58 MPa for the three nucleated media. At 2.17 MHz, thresholds were at 0.41 MPa except for the fresh PBS, which produced a signjficant increase in H202 only at 0.58 MPa. Small plastic spheresappear to be another means of adding nuclei to media (Holland and Apfel 1990; Holland et al. 1992). Polystyrene spheres of 0.26 pm in diameter were added to 5 mL of filtered PBS in 5-PL samples, as for the other tests. At 2.17-MHz continuous exposure, this additive produced significant ca.vitation at 0.82 MPa, but not at 0.41 MPa or 0.58 MPa. Results were of a hit-or-miss nature, and were comparable to those for fresh PBS in two of six tests at 0.58 MPa, and in four of six tests at 0.82 MPa. These solid particles were also tested at 3.8 MHz, but did not instigate significant H202 production. DISCUSSION
and
R. M.
1063
THOMAS
known, and can be used to estimate the concentration of nuclei (Iernetti 1971). This experiment therefore serves as a means of enumerating the nuclei in the medium (e.g., about 1 per milliliter for fresh PBS); although, of course, the nuclei counted are only those detectable by this specific system. The bubble-based contrast agents Albunex@ diluted by lo-‘, lo-’ and 10eJ and Levoviste at 2 mg mI-‘, 0.2 mg mI-’ and 0.02 mg mL-’ concentrations served to nucleate inertial cavitation in this system. The concentrations of these agents normally employed in diagnostic applications are near the concentrations used in this study. For example, a 5-mL injection might mix with 0.5 L of blood in several heart beats to give a lop2 dilution for enhancing contrast in the heart, or mix with 5 L of blood throughout the body to yield a lo-’ dilution. Dilution to lo-” with 0.5 yL of sample, the highest dilution we could obtain with a one-step dilution, also gave reliable nucleation. This implies that at least one nucleus was present in each 0.5..PL sample. For comparison, the concentration of microspheresin Albunex@ is known to be about 500 million per milliliter (Bamhart et al. 1990), which is about lo5 per 0.5 pL. Inertial cavitation was also nucleated by adding 50-PL samples(1 O-’ dilution) of water which had been vigorously shaken and then allowed to rest for 1 min (the same procedure as used for mixing Levovist@ samples). The shaking process apparently produces microbubbles which persist long enough to act as nuclei when exposed to ultrasound. The contrast agent dilutions gave remarkably similar results given the wide range of dilutions. However, this is consistent with the rotating-tube phenomenon, for which only
4k
AND CONCLUSION
Hydrogen peroxide generation was measured to assessinertial cavitation activity during ultrasound exposures in a rotating tube system. Using nuclei-free chambers and media, no inertial cavitation was detected unless nuclei were added. Nucleation was obttined at 2.17 MHz with a full 5 mL of fresh PBS, but not with 50 ,zL of fresh PBS added to denucleated PBS. This implies that there was at least one nucleus available for initiating inertial cavitation in 5 mL but none in six 50-PL samples of fresh PBS. The dependence of cavitation thresholds on volume is well
0.0
0.4
0.6
Pressure Amplitude
0.8 (MPa)
Fig. 4. Resultsfor H,O, concentrationfrom exposure to 2.17 MHz, 2.95 MHz or 3.8 MHz of ultrasoundat the indicated pressureamplitudesfor 2.5 min continuously. The different media were: circles, filtered PBS; diamonds, fresh PBS; squares, filtered PBS + 5 pL Albunex@; triangles, filtered PBS + 5 bL of Levovist@.
1064
Ultrasound
in Medicine
and Biology
one suitable nucleus appears to be needed to generate maximal results (Miller and Williams 1992). One intriguing finding was that, although adding various dilutions of contrast agent produced the about same cavitation activity (see Fig. 2) adding Levovist@ to shaken water or adding Albunex@ to fresh PBS seemed to produce an additive effect (see Fig. 1). Possibly this could be explained by a greater level of aeration, which might enhance cavitation activity, in the shaken water and fresh PBS. Polystyrene spheres were also briefly tried as a source of nuclei. These appeared to serve a nuclei, but on a hit-or-miss basis for lop3 dilution. This behavior indicates that only zero or one nucleus is present in the sample, even though many spheres are present. The stock concentration of spheres is about 2.10” mL’, or about lo9 microspheres in the 5-PL samples. It is uncertain how these solid plastic spheres nucleate inertial cavitation. One possibility is that an occasional sphere, or group of spheres, has some gas associated with it (e.g., trapped in a defect, or within a cluster of spheres). In a previous study, it was found that experimental thresholds for inertial cavitation appear to increase faster with frequency (pressure amplitude threshold proportional to frequency) (Miller and Thomas 1993b) than is expected from theory (pressure amplitude threshold proportional to square root of frequency) (Apfel and Holland 199 1). The rapid increase of experimental thresholds in the megahertz frequency range has been noted by other investigators and appears to depend, to some extent, on the availability of nuclei in the medium (Iemetti 1971). In this study, the contrast agents and fresh PBS were tested at 2.17 MHz, 2.95 MHz and 3.8 MHz, to determine if the addition of large numbers of potential nuclei would alter the dependence of the cavitation threshold on frequency. The results do not seem to support this hypothesis, and the thresholds obtained were about the same as those found in the previous study (Miller and Thomas 1993b), except at 3.8 MHz. At 3.8 MHz, no cavitation was obtained in this study under any of the conditions tested, including fresh PBS. However, cavitation was obtained, as in the previous study, when fresh PBS was used in a dry chamber,(i.e., one that had not been denucleated), which suggests that nuclei on the tube walls are important at this frequency. In conclusion, the bubble-based contrast agents Albunex@ and Levovist@ appeared to be able to provide nuclei for inertial cavitation in vitro. That is, addition of these agents allowed production of statistically significant hydrogen peroxide concentrations by ultrasound exposure under conditions which did not yield H,O, production without the additives. Shaken water,
Volume
21, Number
8, 1995
fresh PBS and polystyrene spheres were also found to add inertial cavitation nuclei to the medium. All the different samples appear to provide qualitatively similar nuclei: the thresholds and dependence of activity on pressure amplitude were similar for all the nucleating samples. None was found capable of consistently instigating cavitation, which simulated the vigorous activity in a rotating tube, in a nonrotating tube. In addition, the agents did not appear to alter the frequency dependence of the experimental inertial cavitation threshold. Acknowledgetnents~We are indebted to Dr. C. Church, Molecular Biosystems, Inc., San Diego, CA, and to Dr. V. Uhlendorf, Schering AG. Berlin, for their generous assistance in obtaining the ultrasound contrast agents. This work was supported by PHS Grant No. CA42947, awarded by the National Institutes of Health. DHHS (USA).
REFERENCES AIUM. Bioeffects & safety of diagnostic ultrasound. Bethesda, MD: American Institute of Ultrasound in Medicine; 1993. Apfel. R. E.; Holland, C. K. Gauging the likelihood of cavitation from short-pulse, low duty cycle diagnostic ultrasound. Ultrasound Med. Biol. 17: 170- 185; 1991. Atchley, A. A.; Prosperetti, A. The crevice model of bubble nucleation. .I. Acoust. Sot. Am. 86:1065-1084; 1989. Barnhart, J.; Levene, H.; Villapando, E.; Maniquis, J.; Femandez, .I.; Rice, S.; Jablonski, E.; Gjoen, T.; Tolleshaug. H. Characteristics of Albunex, air-filled albumin microspheres for echocardiography contrast enhancement. Invest. Radio]. 25:sl62-~164; 1990. Brayman, A. A.; Azadniv. M.; Makin, I. R. S.; Miller. M. W.; Carstensen, E. L.; Child, M. S.; Raeman, C. H.; Meltzer. R. S.; Everbach, E. C. Effect of a stabilized microbubble echo contrast agent on hemolysis of human erythrocytes exposed to high intensity pulsed ultrasound. Echocardiography 12: 13-21; 1995. Flynn, H. G.; Church, C. C. Transient pulsations of small gas bubbles in water. J. Acoust. Sot. Am. 84:1863-1876; 1988. Golberg, B. B.; Liu, J.; Forsberg, F. Ultrasound contrast agents: a review. Ultrasound in Med. Viol. 20:319-333; 1994. Holland, C. K.: Apfel, R. E. Thresholds for transient cavitation produced by p&d ultrasound in a controlled nuclei environment. J. Acoust. Sot. Am. 88:2059-2069; 1990. Holland, C. K.; Roy, R. A.: Apfel, R. E. In vitro detection of cavitation induced by a diagnostic ultrasound system. IEEE UFFC 39:95-101: 1992. Iernetti, G. Cavitation threshold dependence on volume. Acustica 24:191-196; 1971. de Jong, N.; Hoff, L.; Skotland, T.; Born, N. Absorption and scatter of encapsulated gas filled microspheres: theoretical considerations and some measurements. Ultrasonics 30:95-103; 1992. Miller, D. L. A review of the ultrasonic bioeffects of microsonation, gas-body activation, and related cavitation-like phenomena. Ultrasound Med. Biol. 13:443-470; 1987. Miller, D. L.: Thomas, R. M.; Williams, A. R. Mechanisms for hemolysis by ultrasonic cavitation in the rotating exposure systems. Ultrasound Med. Biol. 17:171-178; 1991. Miller, D. L.; Thomas, R. M. A comparison of hemolytic and sonochemical activity of ultrasonic cavitation in a rotating tube. Ultrasound Med. Biol. 19:83-90; 1993a. Miller, D. L.; Thomas, R. M. Frequency dependence of cavitation activity in a rotating tube exposure system compared to the mechanical index. J. Acoust. Sot. Am. 93:3475-3480; 1993b. Miller, D. L.; Williams, A. R. Bubble cycling as an explanation of the promotion of ultrasonic cavitation in a rotating tube exposure system. Ultrasound Med. Biol. 15:641-648; 1989. Miller, D. L.; Williams, A. R. Nucleation and evolution of ultrasonic cavitation in a rotating exposure chamber. J. Ultrasound Med. ll:407-412; 1992.
Ultrasound
contrast
agents 0 D. L. MILLER
Miller, M. W.; Azadniv, M.; Doida. Y.; Brayman, A. A. Effect of a stabilized microbubble contrast agent on CW ultrasound induced red blood cell lysis in ~~itro. Echocardiography 12: l-12; 1995. Ophir, .I.; Parker, K. J. Contrast agents in diagnostic ultrasound. Ultrasound Med. Biol. 15:319-333; 1989. Rresz, P.; Kondo, T. Free radical formation induced by ultrasound and its biological implications. Free Rad. Biol. Med. 13:247270; 1992. Schlief. R. Ultrasound contrast agents. Cum Opin. Radiol. 3: 19% 207; 1991. Schrope, B. A.; Newhouse. V. L. Second harmonic ultrasonic blood perfusion measurement. Ultrasound Med. Biol. 19:567-579; 1993.
and R. M. THOMAS
1065
Williams. A. R.; Kubowicz, G.: Cramer, E.; Schlief. R. ‘Ihe effects of the microbubble suspension SH U 454 (Echovist) on ultrasonically-induced cell lysis in a rotating tube exposure system. Echocardiography 8:423-433; 1991. Winkelmann, J. W.: Kenner, M. D.; Rave, R.; Chandwaney, R. H.; Feinstein. S. B. Contrast echocardiography. Ultrasound Med. Biol. 20:507-515; 1994. Yount, D. E. Skins of varying permeability: a stabilization mechanism for gas cavitation nuclei. J. Acoust. Sot. Am. 65:14291439; 1979. Ziskin, M. C.; Bonabarpour, A.; Weinstein, D. P.; Lynch, P. Contrast agents for diagnostic ultrasound. Investig. Radiol. 7:500-505; 1972.