- Email: [email protected]
Exposure-time dependence of the threshold for ultrasonically induced murine lung hemorrhage
Copyright
Ultrasound in Med. & Biol., Vol. 22, No. I. pp. 13% 141. 1996 0 1996 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 030 I -5629/96 S IS.tXl + .OO
0301-5629(95)02036-5
ELSEVIER
@Technical Note EXPOSURE-TIME ULTRASONICALLY
DEPENDENCE OF THE THRESHOLD FOR INDUCED MURINE LUNG HEMORRHAGE
C. H. RAEMAN,* S. Z. CHILD, $ * D. DALECKI, C. Coxt * and E. L. CARSTENSEN~ *
$*
Departments of -‘Biostatistics, *Electrical Engineering and the *Rochester Center for Biomedical Ultrasound, The University of Rochester, Rochester, NY, USA (Received
7 July
1994; in jinal form
25 July
199.5)
Abstract-Although the extent of suprathreshold damage to murine lung that results from exposure to pulsed ultrasound increases with time, the threshold level for lung hemorrhage is relatively insensitive to total exposure time. Adult mice were exposed for 20 s and 3 min to 2.3-MHz ultrasound (lO-Jo pulses, NO-HZ pulse repetition frequency) at peak positive pressures ranging up to 3 MPa. Threshold pressures for the two exposure times, 1.6 MPa and 1.4 MPa, respectively, are the same within the statistical significance of the measurements. Key Words: Lung hemorrhage, Pulsed ultrasound, Mice, Bioeffects.
The mice were treated with a 2.3-MHz unfocused transducer ( lO+s pulses; lOO-Hz pulse repetition frequency) as used in our earlier study of lung hemorrhage thresholds (Child et al. 1990). The coordinates of the axis of the sound field were determined with a Marconi (Marconi Research Center, Chelmsford, England) hydrophone and these data were used to position the surface of the animals on axis, at a distance of 6 cm from the source. No part of the holder was exposed to the sound beam. The dorsal surface of the animal faced the source transducer. Ultrasound exposures were given to the center of each lung of each animal. For a given animal, each lung received the same exposure. The center of the lung was taken to be the center of the rib cage and 2 mm from the left or right edge of the spinal column of the mouse. Immediately following treatment, the animals were euthanized by cervical dislocation, the lungs were removed, fixed by immersion in phosphate-buffered 10% formalin for 24 h and then inspected visually for evidence of superficial hemorrhage and for penetration of the hemorrhage through the lung. Scoring was blind. The general experimental procedures used here have been described elsewhere (Child et al. 1990). The results are presented in terms of the peak positive pressureat the surface of the animal. Approximate values for the maximum pressuresat the surface of the lung can be obtained by multiplying the pres-
INTRODUCTION There is little question that the extent of damage to lung tissue exposed to suprathreshold pressuresof ultrasound depends on the length of the exposure time. In fact, it turns out that, even with a constant total ontime, greater damage occurs if the exposure takes place over a period of 3 min than if it is applied in one tenth of that time (Raeman et al. 1993). But, the extreme insensitivity of the threshold pressure to pulse repetition frequency (Child et al. 1990) suggests that the threshold itself may not be critically dependent on the exposure time. We report here a direct test of this hypothesis using adult mice.
EXPERIMENTAL
METHODS
As in Child et al. (1990), male, C3H mice, 7-8 weeks old, were anesthetized with ketamine (200 mg/ kg) and rompun ( 10 mg/kg), their backs were shaved and depilated (Neet @, Whitehall Labs, NY) and they were positioned by limb restraint on a small animal holder. The holder, in turn, was supported on a threeway positioner. Protocols for all experimental procedures were approved by the University of Rochester’s Committee on Animal Resources. Address correspondence ter for Biomedical Ultrasound, ter, NY 14627, USA.
to: Carol H. Raemen, Rochester CenThe University of Rochester, Roches-
139
140
Ultrasound
in Medicine
sures in water by 0.84, the minimum the rib cage (Child et al. 1990).
and Biology
Volume
22, Number
1. I996
attenuation of
RESULTS To determine the effects of exposure duration on the threshold for lung hemorrhage, mice were exposed for 20 s and 3 min with a IO-/.LS pulse, a duty cycle of 1: 1000 and peak positive pressures ranging down to 1 MPa in water at the surface of the animal following the procedures described in Experimental Methods. The results are recorded in Table 1 and are plotted in Fig. 1. Although the probability of producing damage at high pressures is greater with the longer exposures, the thresholds are remarkably similar with 20 s and 3 min exposure durations. (Threshold is defined as the intersection of the rate of hemorrhage curve versus acoustic pressure with the background rate.) The peak positive pressure threshold estimates (standard error) were 1.6 (0.4) MPa for the 20-s data, and 1.4 (0.2) MPa for the 3-min data, using the linear plateau model (Baggs et al. 1995). These two estimates were not significantly different (p = 0.66). Note that these are the peak positive pressures at the surface of the animal. Pressures at the surface of the lung are smaller by a factor of 0.84. Because the two thresholds were not significantly different, the data in Table 1 were reanalyzed with the assumption of a single threshold for both exposures. The estimated peak positive pressure threshold from the combined data was 1.5 (0.3 ) MPa with a lower 95% confidence limit of 1.21 by the method of profile likelihood (McCullagh and Nelder 1989; see also Baggs et al. 1995) described below. The curves in Fig. 1 are based on this reanalysis. The threshold obtained in this study is consistent with the estimate in our initial investigation of lung hemorrhage
Table 1. Exposure duration study with mice. Numbers of
lungs with lesionsflungs exposed. Data from this study for 3-mitt exposures are combined with observations under identical conditions in Child et al. (1990) and plotted in Fig. 1. 3 min (&a) Sham 1 1.3 I .5 1.75 2.0 2.5 2.8 3.0 4.9
Child
et al. (1990)
o/2 o/2 214 518
20 s This study
This study
4116 408 308 4/18
0120
7118
2/20 4119
414
6/16
718 2l2
J 0
t
2
3
4
5
Incident Pressure (MPa)
Fig. 1. Comparison of threshold levels for hemorrhage in adult murine lung for total exposure times of 20 s and 3 min. Exposure conditions: 2.3-MHz carrier frequency, lops pulse length, lOO-Hz repetition frequency. Exposure levels are given in terms of the peak positive acoustic pressure at the skin. (0, 0) 3-min exposures; (+ ) 20-s exposures.
Separate datapoints are shown for the 1990 study (a, Child et al. 1990) and tbe present study (0). The background (spontaneous hemorrhage) level in animals used in the second 3-mitt study was atypically large. The composite background level for the 3-min exposure is weighted by the total number of animals. Note that, for the particular analysis presented here, both curves were constrained to have the same threshold as explained in the text.
in mice for comparable exposure conditions (Child et al. 1990). DISCUSSION The value of 1.4 MPa for the peak positive pressure at the threshold for a 3-min exposure compares well with the value of 1.2 MPa reported earlier for the same field conditions (Child et al. 1990). As reported in the earlier study, other descriptors of the threshold wave that correspond to 1.2-MPa peak positive pressure are: 0.6 MPa, negative pressure; 0.8 MPa, fundamental pressure; 0.4, mechanical index; 20 W cme2, spatial peak, pulse average intensity; 20 mW cm-‘, spatial peak, temporal average intensity. It appears that the threshold level for lung hemorrhage is relatively insensitive to total exposure time even though the extent and degree of suprathreshold damage increases with time. This was suggested in the first reports of lung hemorrhage in the mouse by the observation that thresholds were affected little by order of magnitude changes in pulse repetition frequency (Child et al. 1990) and from the observation that the pressure threshold for lung hemorrhage in mice from only 20 individual spark-generated shock waves (Hart-
Exposure-time
dependence
in murine
man et al. 1990) was approximately the same as for 3 min of pulsed ultrasound. Finally, in the tests reported here, we were unable to demonstrate that the threshold for a 20-s exposure was different than that for a 3-min exposure (Raeman et al. 1993). It must be that hemorrhage occurs very rapidly when a suprathreshold ultrasound field interacts with a susceptible region of lung tissue. This is in contrast with thermal effects of ultrasound where there is a trade-off between intensity and time for damage thresholds (AIUM 1993). If one of the criteria for routine applications of ultrasound is to produce no adverse biological effects, it may be important not to exceed threshold for lung hemorrhage, even briefly. Acknowledgements-This Grant No. DK39796.
work
was supported
in part by USPHS
lung hemorrhage
0 C. H. RAEMEN
et al.
141
REFERENCES American Institute of Ultrasound in Medicine (AIUM). Bioeffects and safety of diagnostic ultrasound. Rockville. MD: AIUM. 1993. Baggs R, Penney DP, Cox C, Child SZ, Raeman CH, et al. Thresholds for ultrasonically induced lung hemorrhage in neonatal swine. Ultrasound Med Biol 1996; 22: I2 I - 130. Child SZ, Hartman CL, McHale LA, Carstensen EL. Lung damage from exposure to pulsed ultrasound. Ultrasound Med Biol 1990; 16:817-825. Hartman C, Child SZ, Mayer R, Schenk E, Carstensen EL. Lung damage from exposure to the fields of an electrohydraulic lithotripter. Ultrasound Med Biol 1990; 16:675 -679. McCullagh P, Nelder JA. Generalized linear models. 2nd ed. London: Chapman and Hall, 1989. Raeman CH, Child SZ, Carstensen EL. Timing of exposures in ultrasonic hemorrhage of murine lung. Ultrasound Med Biol 1993; 19:507-512.
Ultrasound in Med. & Biol., Vol. 22, No. I. pp. 13% 141. 1996 0 1996 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 030 I -5629/96 S IS.tXl + .OO
0301-5629(95)02036-5
ELSEVIER
@Technical Note EXPOSURE-TIME ULTRASONICALLY
DEPENDENCE OF THE THRESHOLD FOR INDUCED MURINE LUNG HEMORRHAGE
C. H. RAEMAN,* S. Z. CHILD, $ * D. DALECKI, C. Coxt * and E. L. CARSTENSEN~ *
$*
Departments of -‘Biostatistics, *Electrical Engineering and the *Rochester Center for Biomedical Ultrasound, The University of Rochester, Rochester, NY, USA (Received
7 July
1994; in jinal form
25 July
199.5)
Abstract-Although the extent of suprathreshold damage to murine lung that results from exposure to pulsed ultrasound increases with time, the threshold level for lung hemorrhage is relatively insensitive to total exposure time. Adult mice were exposed for 20 s and 3 min to 2.3-MHz ultrasound (lO-Jo pulses, NO-HZ pulse repetition frequency) at peak positive pressures ranging up to 3 MPa. Threshold pressures for the two exposure times, 1.6 MPa and 1.4 MPa, respectively, are the same within the statistical significance of the measurements. Key Words: Lung hemorrhage, Pulsed ultrasound, Mice, Bioeffects.
The mice were treated with a 2.3-MHz unfocused transducer ( lO+s pulses; lOO-Hz pulse repetition frequency) as used in our earlier study of lung hemorrhage thresholds (Child et al. 1990). The coordinates of the axis of the sound field were determined with a Marconi (Marconi Research Center, Chelmsford, England) hydrophone and these data were used to position the surface of the animals on axis, at a distance of 6 cm from the source. No part of the holder was exposed to the sound beam. The dorsal surface of the animal faced the source transducer. Ultrasound exposures were given to the center of each lung of each animal. For a given animal, each lung received the same exposure. The center of the lung was taken to be the center of the rib cage and 2 mm from the left or right edge of the spinal column of the mouse. Immediately following treatment, the animals were euthanized by cervical dislocation, the lungs were removed, fixed by immersion in phosphate-buffered 10% formalin for 24 h and then inspected visually for evidence of superficial hemorrhage and for penetration of the hemorrhage through the lung. Scoring was blind. The general experimental procedures used here have been described elsewhere (Child et al. 1990). The results are presented in terms of the peak positive pressureat the surface of the animal. Approximate values for the maximum pressuresat the surface of the lung can be obtained by multiplying the pres-
INTRODUCTION There is little question that the extent of damage to lung tissue exposed to suprathreshold pressuresof ultrasound depends on the length of the exposure time. In fact, it turns out that, even with a constant total ontime, greater damage occurs if the exposure takes place over a period of 3 min than if it is applied in one tenth of that time (Raeman et al. 1993). But, the extreme insensitivity of the threshold pressure to pulse repetition frequency (Child et al. 1990) suggests that the threshold itself may not be critically dependent on the exposure time. We report here a direct test of this hypothesis using adult mice.
EXPERIMENTAL
METHODS
As in Child et al. (1990), male, C3H mice, 7-8 weeks old, were anesthetized with ketamine (200 mg/ kg) and rompun ( 10 mg/kg), their backs were shaved and depilated (Neet @, Whitehall Labs, NY) and they were positioned by limb restraint on a small animal holder. The holder, in turn, was supported on a threeway positioner. Protocols for all experimental procedures were approved by the University of Rochester’s Committee on Animal Resources. Address correspondence ter for Biomedical Ultrasound, ter, NY 14627, USA.
to: Carol H. Raemen, Rochester CenThe University of Rochester, Roches-
139
140
Ultrasound
in Medicine
sures in water by 0.84, the minimum the rib cage (Child et al. 1990).
and Biology
Volume
22, Number
1. I996
attenuation of
RESULTS To determine the effects of exposure duration on the threshold for lung hemorrhage, mice were exposed for 20 s and 3 min with a IO-/.LS pulse, a duty cycle of 1: 1000 and peak positive pressures ranging down to 1 MPa in water at the surface of the animal following the procedures described in Experimental Methods. The results are recorded in Table 1 and are plotted in Fig. 1. Although the probability of producing damage at high pressures is greater with the longer exposures, the thresholds are remarkably similar with 20 s and 3 min exposure durations. (Threshold is defined as the intersection of the rate of hemorrhage curve versus acoustic pressure with the background rate.) The peak positive pressure threshold estimates (standard error) were 1.6 (0.4) MPa for the 20-s data, and 1.4 (0.2) MPa for the 3-min data, using the linear plateau model (Baggs et al. 1995). These two estimates were not significantly different (p = 0.66). Note that these are the peak positive pressures at the surface of the animal. Pressures at the surface of the lung are smaller by a factor of 0.84. Because the two thresholds were not significantly different, the data in Table 1 were reanalyzed with the assumption of a single threshold for both exposures. The estimated peak positive pressure threshold from the combined data was 1.5 (0.3 ) MPa with a lower 95% confidence limit of 1.21 by the method of profile likelihood (McCullagh and Nelder 1989; see also Baggs et al. 1995) described below. The curves in Fig. 1 are based on this reanalysis. The threshold obtained in this study is consistent with the estimate in our initial investigation of lung hemorrhage
Table 1. Exposure duration study with mice. Numbers of
lungs with lesionsflungs exposed. Data from this study for 3-mitt exposures are combined with observations under identical conditions in Child et al. (1990) and plotted in Fig. 1. 3 min (&a) Sham 1 1.3 I .5 1.75 2.0 2.5 2.8 3.0 4.9
Child
et al. (1990)
o/2 o/2 214 518
20 s This study
This study
4116 408 308 4/18
0120
7118
2/20 4119
414
6/16
718 2l2
J 0
t
2
3
4
5
Incident Pressure (MPa)
Fig. 1. Comparison of threshold levels for hemorrhage in adult murine lung for total exposure times of 20 s and 3 min. Exposure conditions: 2.3-MHz carrier frequency, lops pulse length, lOO-Hz repetition frequency. Exposure levels are given in terms of the peak positive acoustic pressure at the skin. (0, 0) 3-min exposures; (+ ) 20-s exposures.
Separate datapoints are shown for the 1990 study (a, Child et al. 1990) and tbe present study (0). The background (spontaneous hemorrhage) level in animals used in the second 3-mitt study was atypically large. The composite background level for the 3-min exposure is weighted by the total number of animals. Note that, for the particular analysis presented here, both curves were constrained to have the same threshold as explained in the text.
in mice for comparable exposure conditions (Child et al. 1990). DISCUSSION The value of 1.4 MPa for the peak positive pressure at the threshold for a 3-min exposure compares well with the value of 1.2 MPa reported earlier for the same field conditions (Child et al. 1990). As reported in the earlier study, other descriptors of the threshold wave that correspond to 1.2-MPa peak positive pressure are: 0.6 MPa, negative pressure; 0.8 MPa, fundamental pressure; 0.4, mechanical index; 20 W cme2, spatial peak, pulse average intensity; 20 mW cm-‘, spatial peak, temporal average intensity. It appears that the threshold level for lung hemorrhage is relatively insensitive to total exposure time even though the extent and degree of suprathreshold damage increases with time. This was suggested in the first reports of lung hemorrhage in the mouse by the observation that thresholds were affected little by order of magnitude changes in pulse repetition frequency (Child et al. 1990) and from the observation that the pressure threshold for lung hemorrhage in mice from only 20 individual spark-generated shock waves (Hart-
Exposure-time
dependence
in murine
man et al. 1990) was approximately the same as for 3 min of pulsed ultrasound. Finally, in the tests reported here, we were unable to demonstrate that the threshold for a 20-s exposure was different than that for a 3-min exposure (Raeman et al. 1993). It must be that hemorrhage occurs very rapidly when a suprathreshold ultrasound field interacts with a susceptible region of lung tissue. This is in contrast with thermal effects of ultrasound where there is a trade-off between intensity and time for damage thresholds (AIUM 1993). If one of the criteria for routine applications of ultrasound is to produce no adverse biological effects, it may be important not to exceed threshold for lung hemorrhage, even briefly. Acknowledgements-This Grant No. DK39796.
work
was supported
in part by USPHS
lung hemorrhage
0 C. H. RAEMEN
et al.
141
REFERENCES American Institute of Ultrasound in Medicine (AIUM). Bioeffects and safety of diagnostic ultrasound. Rockville. MD: AIUM. 1993. Baggs R, Penney DP, Cox C, Child SZ, Raeman CH, et al. Thresholds for ultrasonically induced lung hemorrhage in neonatal swine. Ultrasound Med Biol 1996; 22: I2 I - 130. Child SZ, Hartman CL, McHale LA, Carstensen EL. Lung damage from exposure to pulsed ultrasound. Ultrasound Med Biol 1990; 16:817-825. Hartman C, Child SZ, Mayer R, Schenk E, Carstensen EL. Lung damage from exposure to the fields of an electrohydraulic lithotripter. Ultrasound Med Biol 1990; 16:675 -679. McCullagh P, Nelder JA. Generalized linear models. 2nd ed. London: Chapman and Hall, 1989. Raeman CH, Child SZ, Carstensen EL. Timing of exposures in ultrasonic hemorrhage of murine lung. Ultrasound Med Biol 1993; 19:507-512.