(Daniels et al., 1980). The imaging system is capable of detecting gas bubbles with diameters > 10 ,im, although no accurate measurements of bubble diameter ...
Br. J. Cancer (1982) 45,
Suppl. V, 151
ULTRASONICALLY INDUCED CAVITATION IN VIVO G. TER HAAR,* S. DANIELS,t K. C. EASTAUGHt AND C. R. HILL* From the *Department of Physics, Institute of Cancer Research, Royal Marsden Hospital, Sutton, Surrey and the tDepartment of Pharmacology, South Parks Road, Oxford
IT HAS been established that cavitation is an important mechanism for producing effects in biological systems irradiated with ultrasound in vitro (Hill et al., 1969; Morton et al., 1982). However, evidence for cavitation activity in vivo as a result of ultrasound irradiation is sparse. In this paper, cavitation is defined as the formation, growth and activity of gas-filled cavities. Evidence is presented for the appearance of such cavities in vivo in the hind limb of anaesthetized guinea-pigs during irradiation with 0 75 MHz ultrasound. An 8 MHz ultrasonic imaging system (Daniels et al., 1979) has been used to detect these gas cavities and to study their distributions in space and time (Daniels et al., 1980). The imaging system is capable of detecting gas bubbles with diameters > 10 ,im, although no accurate measurements of bubble diameter can be made (Beck et al., 1978). METHODS
Animal preparation.-Male guinea-pigs (400-600 g bodyweight) were used for these experiments. Before each experiment the guinea-pigs were anaesthetized with urethane (1.5 g/kg-1 bodyweight, i.p.), shaved round the hindquarters, chest and neck, and all residual hair from the left hind limb removed using a proprietory depilatory cream. Electrocardiogram (e.c.g.) electrodes were attached, one positioned directly above the heart, the other at the back of the neck. Finally, the guinea-pigs were positioned in a Perspex holder with the left hind limb extended downwards away from the body and the right hind limb lying along the line of the body. Both limbs were securely taped in position, care being taken not to obstruct blood flow. Imaging system.-High resolution, bistable
B-scans were obtained using a mechanically swept, focused 5 mm diameter probe. The probe assembly was supported above a rectangular tank such that it could be positioned accurately in three dimensions. The Perspex holder containing the guineapig was fixed at the front of the tank and the ultrasound probe located centrally and 15 mm behind the extended left hind limb. Acoustic coupling was achieved with a salt solution (0.7% NaCl; 0 035% KCI; 0.03% MgSO4. 7H20: W/V). The gain, suppression and swept gain of the system were adjusted to give an image with a minimum of fine tissue detail. This facilitated subsequent image analysis. Alternate images were automatically recorded, at a rate of one every 2 sec, on 35 mm film (Kodak 2495 RAR). The entire scanning assembly, shown in Fig. 1, fitted inside a 36 litre pressure chamber. 0 75 MHz ultrasound irradiation.-A Rank Multiphon Mark II ultrasonic generator, fitted with a 2-5 cm diameter transducer, was used to irradiate the left hind leg of the guinea-pig with 0 * 75 MHz ultrasound. The intensity was continuously variable up to 3 Wcm-2 (spatial average). The transducer was positioned in the scanning tank perpendicular to the imaging probe (in the same horizontal plane) and 4 cm from the leg. The transducer was located vertically so that the imaged tissue plane fell in the centre of the therapy beam (Fig. 1). Control experiments established that no interaction between the two beams could be observ-ed for 0 75 MHz ultrasound intensities < 2 Wcm-2. The beam profile for the 0'75 MHz ultrasound is shown in Fig. 1. Image analysis.-Each film recorded from an experiment consisted of a 5 min (150 recorded images) control sequence followed by images recorded during 0-75 MHz ultrasound irradiation, with or without the application of increased ambient pressure. Analysis consisted of: (i) combination of all the images obtained during the control period, to produce
A
r) tA 4-i I
toc
4-C
I. ,s 1 2 3 4 5 6 7 8 /9, cm on ~~~~~axis ..
..
.
..
..
..
..
..
.
-o/
to
A
75cm .4-
6 5 cm
- 755cm 4 5cm distance from transducer I I AN Q c *a s. 11 IG. I.-(A) oeanming assembly showing relative orientation (not to scale) of 8 MHz imaging probe (DT) and 0-75 MHz irradiation probe (TT). Coil, shown at the bottom of the tank, thermostatically controls the temperature of the acoustic coupling medium to 370C. (B) Beam profile of 0O75 MHz
ultrasound. (C) Bistable ultrasound image recorded from guinea-pig leg; SK, skin surface; TB, tibula; FB, fibula; remaining internal echoes result from intermuscular interfaces. Direction of transducers indicated, DT, 8 MHz imaging probe; TT, 075 MHz irradiation probe. Time is indicated at the bottom of the image, displayed in min sec.
153
ULTRASONICALLY INDUCED CAVITATION
a "master control" which takes account of variations in the images due to slight movements of the leg; (ii) comparison of all successive images, in turn, with the master control, to identify all new echoes, and (iii) plotting of the location of the sites of appearance of new echoes, relative to the location of pre-existing tissue interfaces, together with the time and duration of their appearance. For those experiments in which the ambient pressure was raised to 5-5 bar gauge (0-55 MPa) during 0-75 MHz ultrasound irradiation a further 30 images, recorded from the start of compression, were included in the master control to account for additional movements of the leg caused by pressurization. RESULTS
It has previously been shown (ter Haar & Daniels, 1981) that the number of sites of appearance and the number of new echoes is proportional to the intensity and time of irradiation with 0-75 MHz ultrasound. The experiments described in this paper were designed to establish whether these new echoes are the result of cavitation. Raising the ambient pressure after a period of ultrasound irradiation will, if new echoes are due to gas bubbles, cause these echoes to disappear. Thus, these experiments consisted of image recording during: (a) 5 min control period, (b) 5 min irradiation with 0-75 MHz ultrasound at 680 m Wcm-2 (spatial average), (c) compression to and maintenance at either 4 or 5-5 bar gauge (0.4 or 0-55 MPa) for 5 min with continued 0-75 MHz ultrasound irradiation, (d) maintenance at pressure without 0-75MHz ultrasound for 5 min and (e) decompression to, and maintenance at, atmospheric pressure for i. 5mm. Initially, 3 experiments involving pressurization to 4 bar gauge were performed. However, due to incomplete recording of the images a full temporal analysis of these experiments was not possible. Spatial analysis was completed and is included in Fig. 2. The full results of the 3 experiments involving compression to 5-5 bar gauge are shown in Figs 2 and 3. An event
6Pr"ssure0 tatm.)
Intensity 1 (w cm-l
-
V
8060 X
tn-
._ 4020zi
E
is
nv-U- ol" c-f
.
.
-
-
.-
.-.,
1-
.
11
-
-
-
-
0 1 2 3 4 5 6 7 8 910 11121314151617151920 Time (min) FIG. 2.-Rate of accumulation of sites of bubble formation. 0-5 min, 0-75 MHz ultrasound, 680 mWcm-2 (7 animals); 5-10 min, 0-75 MHz ultrasound, 680 mWcm-2, 5-5 bar gauge (6 animals); 10-15 min, Ultrasound off, 5-5 bar gauge (6 animals); 15-20 min, Ultrasound off, atmospheric pressure -
.
-
-
-
(5 animals).
Pressure 6(atm.) 0
Intensity 01 (Wcmr') 80 0
6
60 4
-4-0J 0 a)
4-
E
20-
E
U
f
i\\\''
+
X
'
'
b'
'
0 1 2 3 4 5 6 7 8 9 1011 121314151617181920 Time (min) FIG. 3.-Histograms showing the total number of events in each 60 sec interval for one animal. Shaded areas of the histogram represent events that are "persistent" (see text). See legend of Fig. 2 for experimental timing.
is defined as any record of a new echo at a single site. The duration of events ranges from those present on only a single image (single frame events) to those recorded on several successive images (persistent events). Events are equated with individual gas bubbles (Daniels et al., 1980). Single frame events are believed to represent intravascular bubbles passing through the plane of scan. Persistent events may represent either intra- or extravascular gas bubbles. The spatial analysis (Fig. 2) shows that the recruitment of sites at which events are seen is rapid, > 50% of the final
154
G. TER HAAR, S. DANIELS, K. C. EASTAUGH AND C. R. HILL
mWcm-2 and that a majority of these new echoes are caused to disappear by increasing the ambient pressure. This susceptibility to pressure strongly suggests that these echoes are due to gas bubbles within the leg. Furthermore, the temporal distribution of these echoes, leading to the differentiation of single frame and persistent events, suggests a mixture of intravascular and extravascular bubbles. The rapidity of onset of activity, on the time scale of observation, is consistent with the idea that the 0 75 MHz ultrasound is the causal agent. It is perhaps surprising that in some experiments (Fig. 2) evidence is seen for a continuation of recruitment of sites throughout the 5 min period of irradiation. This suggests a number of different sites for cavitation, each having a different activation energy. If this is so, it is not known whether the same initiation mechanism operates at each site. The bubbles observed may constitute a biological hazard either directly or indirectly. Extravascular bubbles may cause tissue disruption and damage, and stationary intravascular bubbles may cause local ischaemia and activation of blood clotting mechanisms. There have been some indications in the early literature (e.g. Grabar, 1953) that ultrasonic irradiation of tumours may sometimes promote metastasis and it seems possible that in vivo cavitation could have the effect of dispersing malignant cells into blood. Restricted, as in these experiments, to a peripheral region these effects may not prove to represent any significant hazard. However, mobile intravascular bubbles are potentially a more serious hazard. Arriving centrally they may disturb cardiac or respiratory function or, if introduced into the blood supply of nervous tissue, they may cause ischaemic damage and consequent neurological sympDISCUSSION toms. From these experiments it is posThese results show that a considerable sible to estimate the volume of free gas number of new echoes arise during irradia- represented by the observed single frame tion with 0 75 MHz ultrasound at 680 events. The total mean number of single
number of sites being seen within the first minute. The total number of sites recruited during 5 min irradiation with 0 75MHz ultrasound at 680 mWcm-2 ranges from 10 to 65 (mean 36-7, standard deviation 22.3). This variation may reflect the variation in leg size (related to bodyweight) and in the exact vertical location of the imaging probe. Recruitment of sites ceased with the application of pressure. The temporal analysis (Fig. 3) shows that the total number of events, recorded during 1 min periods of irradiation, was roughly constant over the 5 min irradiation period. The total number of events recorded during the first minute was within 15% of the mean total for the whole period. The application of pressure caused an immediate reduction in the number of events (mean total during irradiation 41-5 events; mean total during irradiation plus pressure 6.8). Finally, a change in the type of events recorded was observed. During irradiation with 0 75 MHz ultrasound approximately 50% of all events were single frame events, both with and without pressure. However, in the absence of irradiation with 0 75 MHz ultrasound the proportion of single frame events fell to 15%, only beginning to rise after decompression. This may suggest that the most susceptible area for cavitation is intravascular and that even at pressure a small number of intravascular bubbles can be generated. The residual "events" during the period at pressure without 0*75 MHz ultrasound irradiation may represent the "noise" level. Alternatively, it is also possible that even these events, which persist after pressurization and in the absence of 0 75 MHz ultrasound irradiation, still represent stationary gas bubbles stabilized by some structural mechanism (e.g. Philp et al., 1971; Harvey, 1951).
ULTRASONICALLY INDUCED CAVITATION
frame events over the 5 min irradiation period was 108. They all fell into a category, defined elsewhere (Daniels et al., 1980), that represents bubbles with diameters between 10 and 100 yim. Assuming an average diameter of 50 ,um the free volume of gas would be 7 x 10-3 d or, if the maximum radius was assumed, 5-7 x 10-2 ,l. Even if the true number of single frame events were 2 orders of magnitude higher, from a consideration of the intermittent nature of the observations (Daniels et al., 1980), the free volumes would still be 0 7 pl or 5-7 pl respectively. This correction factor is unrealistically high, but even so as an upper limit, suggests that the volume of free gas involved is very small. In conclusion we have demonstrated that free gas bubbles are induced within living mammalian tissue by 0 75 MHz ultrasound irradiation at 680 mWcm-2. For 5 min irradiations we have observed no specific indications ofassociated hazard, but some possibily hazardous outcomes of the phenomenon can be envisaged and may justify further investigation.
155
REFERENCES BECK, T. W., DANIELS, D., PATON, W. D. M. & SMITH, E. B. (1978) The detection of bubbles in decompression sickness. Nature, 276, 173. DANIELS, S., PATON, W. D. M. & SMITH, E. B. (1979) An ultrasonic imaging system for the study of decompression induced gas bubbles. Undersea Biomed. Res., 6, 197. DANIELS, D., DAVIES, J. M., PATON, W. D. M. & SMITH, E. B. (1980) The detection of gas bubbl4 in guinea-pigs after decompression from air saturation dives using ultrasonic imaging. J. Physiol., 308, 369. GRABAR, P. (1953) The biological action of ultrasonic waves. Adv. Bio. Med. Physics, 3, 191. HARVEY, E. N. (1951). Physical factors in bubble formation. In Decompression Sickness. Ed. J. F. Falton. Philadelphia: W. B. Saunders Co. p. 90. HILL, C. R., CLARKE, P. R., CROWE, M. R. & HAMMICK, J. W. (1969) Biophysical effects of cavitation in a 1 MHz ultrasonic beam. Proc. Conf. Ultrasonics for Industry. London: Illiffe. p. 26. MORTON, K., TER HAAR, G., STRATFORD, I. J. & HILL, C. R. (1982) The role of cavitation in the interaction of ultrasound with V79 Chinese hamster cells in vitro. Br. J. Cancer, 45, Suppl. V,147. PHILP, B. R., SCHACHAM, P. & GOWDEY, C. W. (1971) Involvement of platelets and microthrombi in experimental decompression sickness: similarities with disseminated intravascular coagulation. Aerospace Med., 42, 494. TER HAAR, G. & DANIELS, S. (1981) Evidence for ultrasonically induced cavitation in vivo. Phys. Med. and Biol., 26, No. 6 (Nov.).
